RPL1116 (MSO1000Z) and PLA2216 (MSO5000) Active Logic Probe teardown

[Vtech]

Here are some photos of RPL1116 Active Logic Probe pod used in MSO1000Z series scope.

Scope has 16 channel logic probe. The probe pod uses 8x LMH7322 (Dual 700ps High Speed Comparator with RSPECL Outputs). 4 comparators on each side. The pod is connected to the scope using 68 pin connector and flat cable. Reference voltage is generated inside the scope.
Cable connector has:
16x2 pins for PECL signals,
22x GND between diff pairs + some extra,
6 pins for VCC
2x2 pins for +/- supply voltage
Some 4 more signals (reference voltage??)


Most likely, the same design is used in new pod for MSO5000 series - PLA2216 Active logic probe. Here are some x-rays of the new pod: https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2055715/#msg2055715

[wulfman]

get me some dimensions and i can reproduce the board for everybody.
If i make some preliminary schematics, maybe you can fill in the blanks with a ohm meter ?

[oliv3r]

Here's some links to the PLA2216 X-Ray's. Attached is the close-up from the second link

https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2055715/#msg2055715
https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2055778/#msg2055778
https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2057401/#msg2057401

So I reccon to rename the topic to add the PLA2216 for the MSO[5789]000 series

Was looking at pricing of the LMH73222's and TI says they are 4.35 USD per 1 Ku! 
Digikey had them for 8 something a piece!
So with the need of 8, that's 35 to 65 USD! just for the chips

[TK]

Was looking at pricing of the LMH73222's and TI says they are 4.35 USD per 1 Ku! 
Digikey had them for 8 something a piece!
So with the need of 8, that's 35 to 65 USD! just for the chips

You can use 4 LMH7324 instead and save some $$

[TopLoser]

i think it will be useless without the LA daughter board inside, thats where the dso recognize the LA existant, or am i missing something on the newer rigol model?

The board just contains the comparators, scope is ready and waiting for the digital outputs from the comparators and outputs threshold voltages for the comparators as selected on the scope. I think there’s an input to be pulled low to indicate to the scope that the pod is plugged in. That’s certainly the case with the 5000 series and this looks exactly the same but different pinout and connector.

[Vtech]

I've changed the topic to include PLA2216 as this is now more desired version. I'm going to do full reverse of RPL1116 including measurement of passives. I hope we can make DIY logic pod for all Rigols.

The board is 4 layer so not exactly cheap to make. To cut the price we can switch to LMH7324 as Oliv3r said (4 would be reqired). Mayby order samples from TI

[TopLoser]

madmac posted this about the 50 pin 0.1 pitch good old cheap and cheerful IDC connector:

Had a quick look at the 50 way logic connector while having coffee

0V     X    X    D7P
D7N   X    X    D6P
D6N   X    X    D15P
D15N  X    X    D14P
D14N  X    X    D0V

0V     X    X    D5P
D5N   X    X    D4P
D4N   X    X    D13P
D13N  X    X    D12P
D12N  X    X    0V

0V     X    X    D3P
D3N   X    X    D2P
D2N   X    X    D11P
D11N  X    X    D10P
D10N  X    X    0V

0V     X    X    D1P
D1N   X    X    D0P
D0N   X    X    D9P
D9N   X    X    D8P
D8N   X    X    0V

4V0    X    X    0V
4V0    X    X    2V4
D0-7V X    X    D8-15 VREF   10:1 INPUT  +/- 1V5
-2V5   X    X    0V
0V     X    X    DETECT  LOW FOR PROBE ATTACHED

Input range is +/- 15 volts.

Top of table is power switch on/off side and   lower pin  X   X  upper pin

https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2057401/#msg2057401

[joeyjoejoe]

Another nice option is that those needing fewer channels can just populate fewer comparators! With that in mind, it might also be possible to create a board with 2 layers only.

Another option for 2 layers would be to increase the size of the PCB if that eases the routing.

[oliv3r]

i think it will be useless without the LA daughter board inside, thats where the dso recognize the LA existant, or am i missing something on the newer rigol model?

For the DS1000Z series; yeah it's not that usefull ...

[dren.dk]

The reason this subject is interesting is that the MSO5000 series is trivially upgradeable and even the cheapest one comes with all of the hardware, except for probes:
https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2057338/

I think a lot of us that have an MSO5072 in the mail without the LA option would really like to be able to roll our own LA pod cheaper than what Rigol wants for one, even if performance is a bit more modest.

[Vtech]

Reverse engineering of RPL116 completed:D

I've drawn schematic of pod's single channel. Every channel is identical. I've desoldered elements and measured them using RLC meter. The measurements should be quite accurate.

The probe has 3 supply voltages:
VCCO = 2.4 V under load (when pod is connected, no load 2.5 V)
VCCI = 4.0 V (no load 4.1 V)
VEE = -2.4 V (no load -2.5 V)

IN- of channels 1-8 and 9-16 are connected together to common reference voltages delivered by the scope. Reference voltages from the scope are divided by 10 since input attenuator also divides the signal by 10.

There is one mysterious pin which I've named CONF.  It is connected to resistor divider between VCCI and GND (division factor 3/4). It is either some configuration pin so the scope knows that the pod is connected or it is simply measurement of VCCI voltage (now when I think about it, this is more likely the case).

Here is the pinout of the 68 pin connector:

1   GND      35   VEE
2   VEE      36   VCCI
3   VCCI   37   REF9-16
4   NC      38   REF1-8
5   VCCO   39   VCCO
6   VCCO   40   VCCO
7   VCCO   41   VCCO
8   D16Q+   42   D16Q-
9   GND      43   D15Q+
10   D15Q-   44   GND
11   D14Q+   45   D14Q-
12   GND      46   D13Q+
13   D13Q-   47   GND
14   D12Q+   48   D12Q-
15   GND      49   D11Q+
16   D11Q-   50   GND
17   D10Q+   51   D10Q-
18   GND      52   D9Q+
19   D9Q-   53   GND
20   GND      54   GND
21   CONF   55   GND
22   GND      56   GND
23   D8Q+   57   D8Q-
24   GND      58   D7Q+
25   D7Q-   59   GND
26   D6Q+   60   D6Q-
27   GND      61   D5Q+
28   D5Q-   62   GND
29   D4Q+   63   D4Q-
30   GND      64   D3Q+
31   D3Q-   65   GND
32   D2Q+   66   D2Q-
33   GND      67   D1Q+
34   D1Q-   68   GND

Comparing this with TopLoser's findings, looks like the RPL1116 and PLA2216 pods are nearly identical. It should be easy to make a converter to attach RPL1116 to MSO5000 (that's what I'm interested in). Probably, PLA2216 has slightly different input attenuator to allow 200 MHz signals (RPL1116 is rated for 100 MHz).

One more thing - the LMH7332 comparators get pretty hot  - I've measured over 60 C when the pod case was open.

[madmac]

The LMH7324 is only spec'ed down to 5V VCC. The scope supplies 4V

The other difference is hysteresis is adjustable on the LMH7322 and set to around 15mV. On the LMH7324 it is fixed at 25mV. Should not make much difference unless trying to probe very low logic swings.

[Vtech]

Supply voltage should not be a problem. The scope supplies 4V on VCCI and -2.4V on VEE so total supply voltage is 6.4V for VCCI. Even for VCCO it is 2.4V to ground + 2.4V to VEE so 4.8V in total which should be close enough .
It would be also possible to use some other, cheaper RSPECL comparators (are there any? Maxim has something) to build lower performance logic probe. These RPL1116 logis probes are a bit over engineered in my opinion. LMH7322 are specified up to 4GHz so quite an overkill for 100MHz probe.

According to datasheet, sampling rate of digital channels on MSO5000 is 1Gsps. 700ps propagation delay on LMH7322 ensures that the sampling time lags no more than 1 clock behind actual data. But this is only for the comparator itself. What about input attenuator and propagation through connector cable and scope's internal circuitry?

MSO5000 datasheet says:
Min. Detectable Pulse Width: 5 ns
Maximum Input Frequency: 200 MHz
Inter-channel Time Delay: 2 ns (typical), 5 ns (maximum)

Am I right that any LVDS comparator should work with RSPECL inputs at this voltage levels (2.4V)?
From LMH7322 datasheet: "When used in combination with a VCCO supply voltage of 2.5V the outputs have LVDS compatible levels."

[oliv3r]

So I took your pics and overlayed them; just for fun and got the following
I also added the X-Ray from ian, but it was immediatly obvious, while similar (the chips on the same spots) the traces and mounting holes where fastly different.



As for the 'overratedness' I do think they used the same probe cable in all of their earlier MSO's and use the PLA2216 for all the new high-end scopes. So you would have to look at the specs of the MSO7000 or MS9000 when it arrives what the logic probes will have to endure. And then; it was probably cheaper to do one for all is my guess.

[Vtech]

As for the 'overratedness' I do think they used the same probe cable in all of their earlier MSO's and use the PLA2216 for all the new high-end scopes. So you would have to look at the specs of the MSO7000 or MS9000 when it arrives what the logic probes will have to endure. And then; it was probably cheaper to do one for all is my guess.

MSO7000 uses its own logic probe RPL2316. I can't find exact bandwidth spec of this logic pod but it is probably identical to RPL2216 except for the connector.

[suj]

MSO7000 uses its own logic probe RPL2316...

My MSO2072A also uses RPL2316 probe. If I remember correctly, they are of the passive type - the comparators are on the oscilloscope PCB.

[helmy]

what would happen if we feed a 5v p-p signal using a homemade passive probe into the MSO5000 without the comparators?

also is there anyone working on making their own active probe to use with the MSO5000, and getting it made in china to be sold for cheap?

[Vtech]

what would happen if we feed a 5v p-p signal using a homemade passive probe into the MSO5000 without the comparators?

The scope would blow up  At least if there was enough current provided.

Logic pod inputs in MSO5000 accept RSPECL differential signals. Typical levels are Vo-1.1V for high and Vo-1.5V for low line. Since Vo is about 2.4V, H=1.3V, L=0.9V (remember this is differential signal so one line is L and the other is H at the same time).
At this voltage levels it is compatible with LVDS levels so it is possible to use "something" with that type of outputs but you would lose ability to watch negative voltages (not very useful anyway, is anyone still using ECL logic?)

[dren.dk]

I'm in the process of drawing up a LA pod using LMH7324 ICs, I hope to be able to whack it on a 2 layer board and make a couple of units for me and a friend.

I don't have the time to stick it on gitlab right now, but it will end up there when it's a little more complete.

[Noy]

I got 10 samples from ti and will also start of i have a bit more time .
I don't think that a 2 layer board is a good Idea for differential rspecl signals.

[dren.dk]

Yeah, 2-layer is not the best, but the differential signals are destined for a long flat ribbon cable anyway, so I doubt a couple of mm on a two layer board will mean anything in comparison.

[dren.dk]

Well, I just took at look at the pinout of LMH7324 and it's a huge bitch to lay out on a two-layer PCB, so I have re-drawn the entire pod using LMH7322.

The complete schematic is now on gitlab, feel free to critique and/or steal it outright:
https://gitlab.com/dren.dk/mso5k-la-pod/tree/master

[oliv3r]

That's looking great! And on gitlab as well yay!

One question though, Rigol splits each of those comparators 'far' apart, e.g. on chip1, it's ch0 and 8, on chip2, it's ch1 and 9, right?

Is there any downside to having it like ch0 and ch4; ch1 and ch5. That would indeed allow your plan, to have them split in 2 pods, which has many advantages (cheap bastards can only get the 1

But what does that mean to the backend? Do they have more memory/bandwith if you only use chn 0 - 7. Further more, if you do ch0 - 4 on chip 1, pin-out wise, that would mean that on the UI, it would still be chan 0 - 8; so that could be quite confusing?

[dren.dk]

I had actually not considered the distribution of channels vs. comparators, I have no idea if it matters, but my guess is that it's only a matter of layout.

[Noy]

I think its better to build a small but long pod. Put female 50pin straight  connectors in the pcb edge (Pad 1 top, Pad 2 bot, ..).
Then you can Plug it directly in the scope and If you put a second connector in the Others side  you can use jumperwires ...
And If you want, you can use a ribbon cable instead If you use a male header in both pcb sides.
For the cheap guys: Just solder the amount of chips you have/want on the pcb.. You only have to watch out for the correct pin.

Edit:
Why aren't there any protection diodes or others stuff?
Maybe we should add some usb3 (High speed, low capacity) tvs/esd protection diodes?

[joeyjoejoe]

I can get the 7324 symbol in the KiCad library - should help clean up the project a bit if it's a standard symbol.

[dren.dk]

According to the datasheet there's plenty of protection diodes built into the LMH7322, so no external diodes should be needed.

The presence of the input network should be plenty to limit the current in the built-in protection diodes, to the point where you'd have to go seriously nuts to damage anything through a 90k resistor.

Putting the pod directly on the scope is a terrible idea, the whole point of putting the input comparators in the pod is to keep the sensitive single-ended input leads short and get the signals converted to robust LVDS for the long journey through the ribbon cable to the scope.

[wraper]

According to the datasheet there's plenty of protection diodes built into the LMH7322, so no external diodes should be needed.

ESD protection is not better than average MCU has.

The presence of the input network should be plenty to limit the current in the built-in protection diodes, to the point where you'd have to go seriously nuts to damage anything through a 90k resistor.

Not nuts. Spark can easily jump over small resistor. If there are spark gaps on PCB, then it would not happen, though.

[dren.dk]

Hmm, iIt would not be too hard to model the input network in LTSPICE along with the 100 pF + 1k5 resistor needed to simulate a HBM discharge.
See: https://en.wikipedia.org/wiki/Human-body_model

I have not done any simulation or calculations, but my gut feeling is that the energy that makes it into the device is on the very low end that can easily be handled by the built-in diodes.

Do you want to do the simulation?

[wraper]

HBM is something so low in energy that not even worth considering for finished device IMO. Real discharge often will have over 20kV and higher capacitance.

[TK]

Remember this is a digital POD, it is not going to take high voltage in.  I have several logic analyzers (HP, Agilent, Zeroplus, saleae) and I have also analyzed the digital input on Agilent 54622D... the only protection they have is the resistor divider (I think it is 10:1, if I remember correctly).  Even the very expensive high end Agilent Logic Analyzers had resistor network at the input of the comparators.

[dren.dk]

How large discharges are logic analyzers typically designed to survive?

I failed to find a rating of the internal protection diodes of the LMH73222 in the datasheet, the closed I got is that the device survives 2.5 kV HBM and 250 V MM, but perhaps a simulation results of those discharges can be used as a bench mark for discharges though the network.

My guess is that someone experienced in the art of test and consumer gear has already done those simulations and possibly even real-world tests and concluded that nothing reasonable would harm the inputs.

[Vtech]

How large discharges are logic analyzers typically designed to survive?

I think you shouldn't plug it directly into mains socket

Logic analyzers are used in different situations than analog scope probes. You probe some CPU and interface signals not ~kV levels on IGBTs of your 10kW converter. From my experience, input divider of logic pod is enough protection. I've blown several scope probes including high voltage one but never hurt a logic analyzer.

For proper input protection you would need some bigger size resistors (like 1206 at least) and some TVS with breakdown voltage above +-15V and that would be bad for fast signals and signal integrity.

[SilverSolder]

How large discharges are logic analyzers typically designed to survive?

Funnily enough, I was just looking at the specs of the 54622D for other reasons...  -   the digital inputs are rated at max +/-40V peak.

The destruction level is probably quite a bit higher than that.

[dren.dk]

I'm starting to lay out the board and I've used relatively agricultural 0603 passives for the layout, because I suspect I'll have to hand-assemble it:)

I'm having a hard time making it work on two layers, so I think I'm about to buckle and go 4-layer, which is a first for me:)

Getting the pitch of ICs down to 10 mm should be easy enough with 4 layers, even when dedicating a whole layer just for ground.

I think the stack should be something like this:

TOP: Signals  (and almost all components)
1: Ground plane
2: Power and jumpers
Bottom: Power, jumpers and decoupling caps.

I like having the ground plane and uninteresting stuff on the inner layers for debugging, but I worry that sticking the ground plane in the middle means that it's much less effective as a heat sink, is that a valid concern?

[texaspyro]

Bottom: Power, jumpers and decoupling caps.

Try to keep all parts on the top layer... it makes it a LOT easier for most people to build.  Double sided components tend not to work well with a home reflow oven.

[Rerouter]

for the lmh7322, its easy to pull off 2 layers, It will just be a little wider than the original, Have fun, For the second block, I just duplicated and renamed the references, would probably recommend similar for the other, equally without the silk, you can pack them very close, may just need to nudge the matching networks a tad if you want it as small as possible.

[Vtech]

This layout looks teally nice but I think there will be problems with overheating. Thermal pad is not connected to large plane.
Look at the attached thermal image of Rigol's RPL1116 pad without the case. Chip temperature already above 60degC!
This design (I'm taking about Rerouter's layout) would requires some heatsink glued to the chips.

[Rerouter]

That was just to give you an idea on how to do a 2 sided, PCB with all components on 1 side,

The traces can be nudged to get some good via coupling to the VEE pins (-2.5V on the schematic), To the same extent the -2.5V connection to the capactors and similar could be made much thicker to pull some heat with them

At 38C/W, and what reads as about 0.25W of dissipation from the datasheet, I cant see where there is much need for over the top PCB cooling.

Edit:
Part of the extra heating your seeing there is from the ones on the back, Ironically by making it on the same side, the density of the heat sources would be lower.

Edit2: Here is what you could do if you wanted some extra heat sinking.

Edit3: Just to make clear, at this point I have put 0 effort into impedance matching, If you guys need help with it, I would just need to know what board thickness your working towards.

[ebclr]

What about to make the board for only one channel, a very tiny board, and use 16 boards connected to one big adaptor board. Will this improve signal integrity since the wire will be differential and less susceptible to noise?

Do this make any sense?

[Rerouter]

Single channel PCB's would be a bit of a pain due to the common wires between all the channels, meaning at the wiring loom side you would have a proper mess.

8 channel podlets is easy enough, and 4 channel I would probably call the cut-off for wiring mess issues.

Also Third go at it, each channel pair is 12.5 x 22.8mm in size, so 8 channels can be made into a board about 50mm in size, I suppose if you rotated the resistor networks you could scrape back enough room to try and cram both podlets onto a 50 x 50 mm pcb order. but I don't see the point seeing as you generally get 5 or 10 boards per order.

[joeyjoejoe]

What size components are these? 0805?

[Rerouter]

0603 and a 0.5mm pitch QFN, If your not used to Kicad, the Red is the Copper pads / traces, the Purple is the solder mask (currently at the default, can easily be trimmed down), and the White is the "courtyard" essentially a spacing guide for components so they are easy for most pick/place machines to use. In general I have found kicads default courtyards are relaxed enough that most people should be able to hand solder them.

[ebclr]

Interesting tool to make a team PCB, porting to this tool

https://upverter.com/3DgeekStore/c728af4c317fb73e/lmh7322/

[Rerouter]

Link seems to be broken?

[ebclr]

https://upverter.com/3DgeekStore/   then find for lmh7322/

[ebclr]

need one email to be invite to the project collaboration, Tell and I will add

[Rerouter]

Probably my last run for a while on this, All that's left is adding the 50 pin connector, and routing for it. as you only need to route half the channels, it should be easy. I just can't be bothered making the footprint.

[joeyjoejoe]

I can help with the footprint. Do we know what sort of connector it is?

I've got a pull request in for the 7324 symbol.

[Rerouter]

2x25 1.27mm pitch, intended for mounting on the edge of the PCB, so 1 row of pins per side of the board

Don't know the exact plug part number.

I suspect this means the PCB is 1.0mm or 0.8mm thick.

[joeyjoejoe]

It might be best to look at options that can work with 1.6mm boards as these are cheap for fab houses.

If we start getting off the beaten path, things get expensive pretty fast...

I'll take a look.

[texaspyro]

It might be best to look at options that can work with 1.6mm boards as these are cheap for fab houses.

Almost all the cheap fab houses off 0.8-1mm boards for the same price as the 1.6 mm boards.

[TopLoser]

2x25 1.27mm pitch, intended for mounting on the edge of the PCB, so 1 row of pins per side of the board

Don't know the exact plug part number.

I suspect this means the PCB is 1.0mm or 0.8mm thick.

Are you sure about that? Looks like a good old fashioned 2x25 0.1” pitch connector on the one I’ve got.

[joeyjoejoe]

2x25 1.27mm pitch, intended for mounting on the edge of the PCB, so 1 row of pins per side of the board

Don't know the exact plug part number.

I suspect this means the PCB is 1.0mm or 0.8mm thick.

Are you sure about that? Looks like a good old fashioned 2x25 0.1” pitch connector on the one I’ve got.

Oh crap, you're right! Good to know 0.6-1.6mm at least for PCBWay is the same price.

[Rerouter]

Could well be. I was working off the xray image. (Dont have the hardware myself)

[TopLoser]

It’s 2x25 0.1” pitch on the scope, interface lead and pcb take it down to smaller pitch cable so I guess it may well be a 0.05” pitch connector on the active probe pcb...

[Rerouter]

I would base it on the smaller the ribbon the less cumbersome it is to use. A 0.05" ribbon cable for 50 wires is about the size of an IDE hard drive ribbon.

Edit:
Did some checking based on images, its quite clearly using 0.025" or 0.635mm pitch ribbon wire between the scope and pods. Nice and standard, and also reduces the amount of connectors it could be.

So for each pod of 8 channels, there is (ground, 4v, +2.5, -2.5, Vref, 8 differential pairs) at minimum 21 signals, do we want to make these devices purely compatible with what is already in place, or make it designed around 2 seperate 8 channel pods? e.g. have a little breakout board at the scope end to 2 plugs for the pods?

I was thinking something like the Hirose ST40 series connectors, most likely the 24 pin one, but as they don't have prefab cables, Its probably not a good option for a projec tlike this. the scope end is 0.1" 2x25, shall we just make both ends the same? or do we get creative with smaller ribbons or more flexible wiring? by using a little adaptop board that plugs in at the scope?

[madmac]

In testing the LA input with a LVDS driver I noticed that channel 6 has a lot of crosstalk when unterminated. The remaining lines are all quiet.
It would be good if some else could confirm that it is a layout issue inside the scope not a fault on my unit. To get live update on the LA short pins 1 and 2 (the right most pins) on the 50 way connector and press the LA button. With nothing connected I get this issue.

These pins seem to have a little bit of protection and go directly to one of the ASIC's which I guess is why they are not hot pluggable.

With the SN75LVDS389 connected it works correctly.

[Rerouter]

At a glance, some of the differential pairs in the scope are far closer than others,
https://www.flickr.com/photos/eevblog/44893153125

So i could imagine some pairs being more noise effected than others. Not to mention once they disappear into internal layers Its hard to say what happens.

[dren.dk]

I just finished my first stab at the 4-layer + single sided version of the board.

All components are on the top layer along with almost all of the signals.
Layer 2 is a solid ground plane.
Layer 3 is a mostly solid -2.5 V plane for heat sinking, because somone noted that the chips were about 60 degrees in the Rigol pod.
Layer 4 busses the power and other common signals between channels.

I have not yet done anything about the silkscreen, though I might as well just hide all the references and values because there's no room for them:)

The channels will need to be labeled though.

I'm still not entirely happy with the crampedness of parts around the ICs I could only make around 1 mm worth of space there, so any rework is going to be a bitch.

I slapped a 50 pin 0.1" IDC connector on the board because that's what's in the scope and with the size the parts take up there's really no reason to save space there.
The board ended up at 86x62 mm.

I used through hole connectors because I like robustness.

Everything is online here: https://gitlab.com/dren.dk/mso5k-la-pod

[joeyjoejoe]

I would base it on the smaller the ribbon the less cumbersome it is to use. A 0.05" ribbon cable for 50 wires is about the size of an IDE hard drive ribbon.

Edit:
Did some checking based on images, its quite clearly using 0.025" or 0.635mm pitch ribbon wire between the scope and pods. Nice and standard, and also reduces the amount of connectors it could be.

So for each pod of 8 channels, there is (ground, 4v, +2.5, -2.5, Vref, 8 differential pairs) at minimum 21 signals, do we want to make these devices purely compatible with what is already in place, or make it designed around 2 seperate 8 channel pods? e.g. have a little breakout board at the scope end to 2 plugs for the pods?

I was thinking something like the Hirose ST40 series connectors, most likely the 24 pin one, but as they don't have prefab cables, Its probably not a good option for a projec tlike this. the scope end is 0.1" 2x25, shall we just make both ends the same? or do we get creative with smaller ribbons or more flexible wiring? by using a little adaptop board that plugs in at the scope?

I would keep it simple with a 0.1" 50 pin cable. H3DDH-5006G has a polarizing key and a strain relief and is cheap. Then slap a header on the PCB and call it a day. Could even look for shorter versions as well - I've seen some other companies that have them even shorter.

[Rerouter]

Well if thats the case. 50 pin scsi cables should fit the bill and come in various lengths.

[wulfman]

It looks nice but the output of the Comparators need to be treated as a diff pair all equal lengths. Also on the rigol pod the inputs all show a ground with the signal so at least 32 pins to the board under test. I do like the 50 pin connector, cheap and simple I was however thinking about the use of an 68 Pin Ultra320 SCSI cable that uses twisted pairs. I am in my spare time creating a board like this for my 5000 series scope.  TI makes a 4 channel version of that IC  LMH7324SQ/NOPB  it may make the board a lot easier to manufacture. at 13~ a lot cheaper too. 

[TK]

TI makes a 4 channel version of that IC  LMH7324SQ/NOPB  it may make the board a lot easier to manufacture. at 13~ a lot cheaper too. 

I proposed the LMH7324 but it seems to not have all the required features.

https://www.eevblog.com/forum/testgear/rpl1116-active-logic-probe-pod-for-1000z-series-teardown/msg2090230/#msg2090230

[Rerouter]

Its more as I was trying 2 layers it would be an utter pain to layout. If 4 layers are in the mix I can give it another spin with the lmh7324's

[wulfman]

I been reading the data sheet. It looks like i can make it work. What have i got to loose ?
Boards are cheap, parts are cheap, I have a large order i need to send to my Chinese board house so i might as well send in this one,

[dren.dk]

I initially wanted to use the LMH7324 too, but the pin-out does make it a bit harder to use, because the outputs sit on opposite sides and it's not really that much cheaper at the sub-10 pieces end, because with the LMH7322 you get into price breaks twice as fast.

The layout of the 16 channel pod is somewhat cumbersome due to the ordering of signals in the 50 pin connector and the resulting scramble of inputs.

As far as I can tell the pin-out of the 50 pin connector was all dictated by having a dual-sided load in the Rigol pod.

I think I want to do a new design with a 3 board construction:
* Two identical 8 channel pods, with a sane straight-through layout.
* A simple two-sided converter board which unscrambles the 50 pin ordering and provides two 26 pin connectors for twisted pair ribbon cable such as this: https://www.mouser.dk/ProductDetail/Amphenol-Spectra-Strip/132-2801-026?qs=sGAEpiMZZMsJiFh04Lj2rnHfTmy1h2ypJt6bk8ifWw4%3d


There are a couple of nice properties of the split approach:
* It's the way the adults do it, see Tektronix P6418: https://www.google.com/search?q=Tektronix+P6418&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiOuZ2a9IDgAhUBjSwKHVVECf8Q_AUIDigB&biw=1920&bih=1077 or Agilent 54620-61601: https://www.google.com/search?q=Agilent+54620-61601&client=firefox-b-ab&tbm=isch&tbo=u&source=univ&sa=X&ved=2ahUKEwiz2rL784DgAhWBOSwKHX-sCVIQsAR6BAgGEAE&biw=1920&bih=1077
* The layout of the pod becomes nicer because the ordering of inputs and outputs can be the same.
* Hand-assembly, rework and debugging is easier with two separate pods, I think.
* The two smaller pods with smaller cables are easier and more flexible to use than one 16 channel pod with a large cable.
* If someone decides to make a probe for another scope, then it's very cheap and easy to produce a new converter.

I wanted to use a female 50 pin board mounted connector on the converter board (like this, but ideally edge-mounted: https://www.mouser.dk/ProductDetail/3M-Electronic-Solutions-Division/6850-4500PL?qs=sGAEpiMZZMs%252bGHln7q6pm1D%2fu5fPR%252bY0EqgFxb8dXZE%3d), now I'm thinking that having a 10 cm bit of 50 lead ribbon going to the scope might be better due to it possibly being lower physical stress on the scope connector and probably cheaper to make too.

[Rerouter]

Well for an 8 channel pod. Feel free to use my 2 layer pcb design. If you have a case in mind i can tweak it.

edit:
Your linked 50 pin connector calls for a 1.6mm PCB,

Edit2: sorry didnt see you where planning on speccing your own loom ribbons, instead of premade assemblies.

Edit3: If your going with custom ribbons to the pods, why not consider the smaller pitch ribbons? really make it a clone of the tek and agilent? It ends up slightly cheaper per foot.
https://au.mouser.com/ProductDetail/3M-Electronic-Solutions-Division/3754-26-300?qs=sGAEpiMZZMsJiFh04Lj2rnGVP%252b060guwueW9iVoIWzc%3d

[dren.dk]

I certainly do want to use a narrower pitch ribbon cable, the only trouble with that is finding connectors that aren't hugely expensive.

The 3M tripolarized series looks nice but it's spendy.

I wonder if anyone makes ribbon to board transition connectors in 0.025" pitch like these for 0.1": https://katalog.we-online.de/en/em/BHD_2_54_BOARD_TO_CAB_TRANSITION_CONNECTOR_6120XX22323

[wulfman]

cable  https://www.cablewholesale.com/specs/10p2-39103.php
connector https://www.digikey.com/product-detail/en/te-connectivity-amp-connectors/8-5174682-7/8-5174682-7-ND/957333

twisted pair high speed. I am sure better than any flat cable

[Noy]

Do we really need such cables?
I think Rigol knows what they are doing...
I thought the main goal was to get a working  equivalent good pod like the original but much more cheaper... Not 350€...

[wulfman]

Well its not much different really about 17$ more
for something that IMHO would work better
but please go on with your design with the 50 pin stuff
I am just playing around with my design.
I have never used the tuned line feature in Orcad and want to explore the
feature. This is an excellent opportunity for me to do so.

connectors https://www.ebay.com/itm/6-Pcs-IDC-Cable-Connector-FC-50P-50Pin-Female-Header-2-54mm-Pitch/311008750065?epid=1339185073&hash=item486990e9f1:rk:4:pf:0
   
cable        https://www.ebay.com/itm/1M-3-3Ft-1-27mm-Pitch-50-Pin-Wire-Gray-Flat-Ribbon-Cable-For-2-54mm-FC-Connector-/172341711789

board connectors  https://www.ebay.com/itm/6-50-Pin-IDC-Box-Header-Straight-Angled-Flat-Ribbon-Cable-PCB-FC-Connector/253520275026?hash=item3b06fc6652:m:m-nK827yuu_e05v_eSwju6g:rk:3:pf:0


so 15 ~ bux appx for the 50 pin stuff

17$ for the 68 pin scsi cable 20 bux for the board connectors.

[Noy]

With the lvds scsi cable you need an additional converter pcb added to the BOM.

[ebclr]

What about change Ground plane to top and make GND a heat dissipation plane, those ic's like to be hot

[TK]

Well its not much different really about 17$ more
for something that IMHO would work better
but please go on with your design with the 50 pin stuff
I am just playing around with my design.
I have never used the tuned line feature in Orcad and want to explore the
feature. This is an excellent opportunity for me to do so.

connectors https://www.ebay.com/itm/6-Pcs-IDC-Cable-Connector-FC-50P-50Pin-Female-Header-2-54mm-Pitch/311008750065?epid=1339185073&hash=item486990e9f1:rk:4:pf:0
   
cable        https://www.ebay.com/itm/1M-3-3Ft-1-27mm-Pitch-50-Pin-Wire-Gray-Flat-Ribbon-Cable-For-2-54mm-FC-Connector-/172341711789

board connectors  https://www.ebay.com/itm/6-50-Pin-IDC-Box-Header-Straight-Angled-Flat-Ribbon-Cable-PCB-FC-Connector/253520275026?hash=item3b06fc6652:m:m-nK827yuu_e05v_eSwju6g:rk:3:pf:0


so 15 ~ bux appx for the 50 pin stuff

17$ for the 68 pin scsi cable 20 bux for the board connectors.

The IDC connector might be too shallow to fit inside the MSO connector

[dren.dk]

You can't have the ground plane on top, that's where your components live, so very little copper would be left at these densities.

The holes poked in the power planes for vias alone takes out a lot of copper, even with 0.4mm clearance, poking holes for components would make it useless.

You can't connect the heat spreading copper layer to GND, because the bottom pad of the LMH7322 is connected to VEE (-2.5V), but with LMH7324 it's unconnected AFAIR, so you could if you used that in stead.

In my layout I have GND on Layer 2 and -2.5 V on layer 3, so I expect to get much better heat spreading than the original.

The yellow in this screen shot is the GND pour, the -2.5V heat spreading leayer is under that, but it looks almost the same: https://gitlab.com/dren.dk/mso5k-la-pod/blob/master/pcb-complete-in-pcbnew.png

[2N3055]

SCSI cables are as thick as your thumb, stiff, not bendable, weigh a kilogram, and are pain.
Without being screwed in or locked , they will fall out from connector under it's own weight.
There is a reason why every manufacturer uses flat cables for this. And they are usually very soft and pliable ones.

It would make sense to make it compatible with used LA cables that are available on E-bay for older LA and MSO scopes.

Or just make a flat cable from scope to a small board on user side, that has some kind of transition to hooks, grabbers or whatever you want to use for DUT connection.

[Rerouter]

2N3055, care to give some example links,

Its easy to route for whatever connectors are needed, I'm thinking 0.635 / 0.25" pitch ribbon if its possible to the pods, 2x13 is a very nice option for the pods 8 channels with a good number of grounds, or 2x15 if you want every pair separated by a ground. (I personally am leaning towards 2x13, as the buffered differential logic should be the safest signals on the board from crosstalk.

[2N3055]

2N3055, care to give some example links,

Its easy to route for whatever connectors are needed, I'm thinking 0.635 / 0.25" pitch ribbon if its possible to the pods, 2x13 is a very nice option for the pods 8 channels with a good number of grounds, or 2x15 if you want every pair separated by a ground. (I personally am leaning towards 2x13, as the buffered differential logic should be the safest signals on the board from crosstalk.

Sure, for instance: https://www.ebay.com/itm/Agilent-HP-10089A-Logic-Woven-Cable-w-Flying-Leads-for-54622D-MSOX2000-MSOX3000/253907595253?_trksid=p2485497.m4902.l9144

With logic pod cables, most pain is on the user side. Many people buy original MSO cables because they  come with nice hooks or grabbers.

Terminating into a small PCB on user side, that has a breakout pins, would make it easy to just get 10 PCBs and keep them permanently connected to DUT. That would make switching scope from DUT to DUT easy, and connecting DUT even easier. You could solder to it, making PCB disposable. You just order 10 more.
Or you can connect PCB to grabbers of your choice (or just arduino style pin connectors) and connect normally.

[dren.dk]

Does anyone have a good idea for the scope connector?

I just measured my scope and the depth of the cavity from the front of the scope to the bottom of the connector is just under 15 mm and the front cavity wraps quite tightly around the connector, so I doubt a standard ribbon IDC connector will be able to fit in there.

I'm currently thinking that a 2x25 pin socket with wirewrap pins (or those reinforced long pins as used for pc104) could be a solution.
With the tall socket the board that it connects to would sit flush with the front of the scope and have a couple of SMD connectors for the cables that run to the 8 channel pods, it would probably have to shaped to fit with the contour of the scope, to provide keying, but with a nice black board that would not look all that silly, I think.

Another alternative would be to get a proper keyed connector such as this and edge mount it: https://uk.rs-online.com/web/p/pcb-sockets/6081437/ that way the pcb will be sticking out horizontally, but I can't find a connector that's supposed to be mounted like that.

[Rerouter]

Wasn't exactly what I was thinking, could you explain that detail in more detail. As if we can find a source for those grabbers, buy those and terminate them to a 9 pin 0.1" header?  meaning you could have many of those 9 pin plugs read to swap to and from the pods at will??

I should point out the PCB's we are routing are the 8 or 16 channel pods with the ECL comparitors, I can make them pretty small if needed, but they need to be close to the grabbers,
If dual loaded 4 layer was in the mix, then an 8 channel pod pcb with a 0.05" 2x13 header / 0.635mm ribbon to the scope could be shrunk all the way down to 25x30mm, Small is not that hard, But the comparitors are not the cheapest parts to say you want 10 sets, the looms should be.

The other thing with getting this small is finding a good enclosure. I've come across just that many times in the past

[Rerouter]

If you have the exact opening size, we could also shim the PCB a little to ensure the pivot is away from the connector,

Equally that would let us know what we can get away with.
I do feel it will have to be inline soldered, but 50 solder joints is quite a bit of strength, considering that a 1.6mm board has barely any gap,

the shorter we make the breakout PCB, the less torque it will experience, so perhaps we attack it that way, almost 0 length from the first plug to the other 2? 1 plug in, 2 plugs out.

[dren.dk]

Sorry I didn't think to measure the opening width and height, but from memory it was just about as tight as the inside of the connector.

BTW. I think our venerable NPN power tranny friend was talking about doing away with probes and in stead connecting the pod to a breakout board with a cheap connector that allows you to solder the breakout board into the DUT you're working on and leave it there when switching to another project.

If we wanted to make the probe breakout boards super cheap, then they use a PCIe card edge connector, though I'd be wary of having that connector directly in the pod as that would be somewhat of a dangerous leaver to have on the presumably short and fragile probe wires that are soldered directly to the DUT.

I doubt PCIe connectors come in IDC ribbon variants, but this looks nice: https://www.te.com/usa-en/product-5-1734857-1.html

[Rerouter]

Looking into the LMH7324, It doesn't look like its normally stocked, For a 4 layer single side loaded PCB its actually pretty easy to route, more comes down to can anyone even get those chips?

[dren.dk]

Huh?
According to octopart lots of places have stock:  https://octopart.com/search?q=LMH7324
Mouser certainly has them: https://www.mouser.dk/Search/Refine?Keyword=LMH7324

Am I missing something?

Oh some other nice things about the LMH7324 is:
* The belly pad is unconnected so it can be tied to the GND plane, which means that I would not need to dedicate a layer to -2.5V
* It doesn't have settable hysteresis, so those two resistors go away, the original hysteresis is about 10 mV and the fixed one of the LMH7324 is about 21-24 mV), I can't see it making any difference.
* It doesn't have the latch feature, so the bus to supply those two logic levels and the resistors to pull them up and down go away too.

[madmac]

Do we really need such cables?
I think Rigol knows what they are doing...
I thought the main goal was to get a working  equivalent good pod like the original but much more cheaper... Not 350€...

A low cost solution if you are happy just looking at 3V3 and 5V0 signals is to use a single ended to LVDS driver and an inline resistor to offer a bit of protection.
75LVDS387 for 16 ways (or 8 way 75LVDS389, or 4 way 75LVDS391. The trigger level would not be used. A low drop out regulator can supply the 3V3 rail from the 4V0 supply.
Not for high performance.

[joeyjoejoe]

Do we really need such cables?
I think Rigol knows what they are doing...
I thought the main goal was to get a working  equivalent good pod like the original but much more cheaper... Not 350€...

I agree with this.

What is the design goal here?

The nice part, if we are dealing with KiCad, is anyone can design as they wish With the open source schematic that has been published, it's now just a few hours of figuring out routing.

For me, it's the price point. Cost of BOM and board (2 layer) would be above anything else. After all, if we approach anywhere near the retail price, you can buy a supported, engineered product direct from Rigol.

Using a cable from DSO to probe lets the board grow in size. It also presents the least risk to the DSO from torque - accidents happen.

[dren.dk]

I've just measured the cavity the connector on the scope sits in and it's at least 11 mm high, so it's not a problem to fit a regular 50 pin 0.1" pitch IDC ribbon connector in there, if that's your thing.


I had another thought about connecting the pods to the converter board and it struck me that HDMI cables is available everywhere, really cheap, exists in a lot of lengths and there are very thin and flexible models available too.

HDMI has 4 shielded twisted pairs and a bunch of extra wires, so it's perfect for an 4 channel pod, which is the minimum you'd get with a single LMH7324

The overhead of having 4 pods does seem a bit over the top, but getting to use commonly available, cheap and robust cables would be super awesome.

[Rerouter]

Ok, so looks like Ribbon Cable Assemblies are not that expensive for the wiring to the Pods,

https://au.element14.com/samtec/ffsd-13-d-12-00-01-n/cable-assy-26p-idc-rcpt-rcpt-305mm/dp/2856836

Now that leaves 2 options for the breakout board,
1 - Find a cheap source for a 2x25 0.1" ribbon cable about 200mm long, and plug that into the breakout board
2 - make the breakout board plug in directly, but shim the shape so its supported. e.g. 3D printed housing.

[joeyjoejoe]

https://www.digikey.com/product-detail/en/assmann-wsw-components/H3DDH-5006G/H3DDH-5006G-ND/1218621

$2.60

I think this is best - minimize risk of damage to scope, logic analyzer pod, or both.

[dren.dk]

Yes, it's certainly hard to beat a pre-fab ribbon, just for my own benefit these are the options I see for the scope to-pod-connection:

0: 50 lead 0.1" IDC ribbon connector plug into the scope and run all the way to the 16 ch pod (that's what my already completed 16 ch pod does).

1: 50 lead 0.1" IDC ribbon connector plug into the scope, with a very short ribbon that plugs into the breakout board that allows two tighter (0.025" pitch) 26 lead ribbons to go to two 8 channel pods.

2: Straddle mount a 50 pin female connector on the edge of the breakout board so it can be plugged horizontally into the scope.

3: Use an elevated 50 pin connector to allow the breakout board to sit flush against the front of the scope.


0: Easy to source, not that cheap, but a bit unwieldy with that huge ribbon.
1: Easy to source, but it just feels wrong to have a dongle on the scope like that, it's not an Apple product after all.
2: Somewhat easy to source, could be very cheap if a proper straddle mount connector can be sourced, but a bit harder to do because it requires a collar/case around the board to support it against the fascia of the scope, otherwise it more closely matches Rigols solution.
3: Could be harder to find the exact part, but super compact, cheap and robust with zero risk of putting stress on the scope connector, the board should be designed to fit flush against the front of the scope to key the orientation and the same board could be used for the #1 solution if unable to source the connector.

Samtec makes the connector: ESQ-125-23-L-D: https://www.samtec.com/products/esq and it's in stock in at least a few places: https://octopart.com/search?q=ESQ-125-23-L-D
I like Mouser and though they don't have stock, the lead time is only 5 days, so it's not too bad.

Mpe also makes it, but it seems to be a custom part: 100-2-050-0-NFX-YS0-1844:
https://www.mpe-connector.de/index.php?lang=en&menu=16&id_product=1002&order_code=100-2-050-0-NTX-YS0-2324

[wulfman]

Well it can all be done, all 16 channels in 2.3"x2" of space using the 4 channel devices.
4 layers is mandatory

[dren.dk]

Looks nice, though I think you need some thermal vias to tie the bottom pad into some heat spreading zones.

[Rerouter]

wulfman, that is with 0403 passive isn't it? If you even hopped back to a more normal 2x25 0.1" header, that would only make your board a sliver wider, at 2.5" wide.

[madmac]

wulfman, that is with 0403 passive isn't it? If you even hopped back to a more normal 2x25 0.1" header, that would only make your board a sliver wider, at 2.5" wide.

Wulfman's layout is for the  RPL1116 Active Logic Probe 68 pin connector.

[wulfman]

Oh yeah no way finished with it loads more to do   yes thermal vias etc...  but its going well.
I finished the routing but not the entire design. I was tired last night 

is the pinout for the 5000 series the same as for the earlier scopes ? if not the small board that plugs into
the scope would be the only thing different from this board to any scopes.

[dren.dk]

No, the mso5k uses a 2x50 0.1" pitch header, the pinout can be seen on the first page of my schematic: https://gitlab.com/dren.dk/mso5k-la-pod

The deci-inch connector pretty much demands a converter board, IMHO.

[wulfman]

Ok thats the pinout i used on my converter board. I like the 68 pin connectors they use on SCSI III drives as they are easy to get
and reasonable. you can use the flat internal or round external and the diff pairs use twisted pair in the cable.
I still want to use the diff pair equal length routing but setup is not easy since i have never done it before in Orcad.

[Rerouter]

What might make things easier is you have control of the order of the channels. And you only have to length match input to output. Whether its trimmed on the single ended or differential you can match the skew.

[wulfman]

Well diff routing was not hard. here is adapter board to 68 pin cable.

[ebclr]

Will HDMI or CAT 7 connectors be a good choice to break in 4 channels pods?

[dren.dk]

Regarding pod-cabling:
* RJ45 doesn't really work, it's only 4 pairs and that's too little for even a 4 channel pod.
* HDMI could work for a 4 channel pod as there are 4 pairs and a number of extra conductors, but you'd still have a problem if children start plugging monitors into the breakout board, though.

Regarding length matching the LVDS pairs, I don't buy that length matching matters at all, because at 1 GSa/s nobody is going to be able to be able to measure a sub-cm length difference in signal paths.

Looking at the traced near the LA connector in the scope does not reveal any meandering to equalize the quite uneven lengths of traces going to the LA connector, sure, they could do that on an inner layer, but doing that near the connector where there's plenty of room would make more sense to me, so I suspect that it wasn't a consideration at all.

[wulfman]

Length matching was not my concern as much as impedance matching with the distance between dif traces.

[dren.dk]

Ok, it seems like you're aiming for a straddle mount socket for the scope connector, what kind are you using?

[Muza]

Regarding pod-cabling:
* RJ45 doesn't really work, it's only 4 pairs and that's too little for even a 4 channel pod.
* HDMI could work for a 4 channel pod as there are 4 pairs and a number of extra conductors, but you'd still have a problem if children start plugging monitors into the breakout board, though.

What about full-feature USB 3.1 Type-C cables?
Plenty of shielded data lines and an abundance of cables...

Hopefully some one chimes in on the pin configuration, so that when accidentally plugged into a computer the device just gets ignored.

I kinda like the idea of sleek pods with low profile connectors such as Type-C...

[Muza]

A quick'n'dirty render of what I meant. Last number is of course 15, not 16...
Essentially, it does not matter if we decide to use any interim connectors between the scope and the pod, in any case I was planning to 3D print a case and then make a silicone mold of out of it.
Afterwards, casing can be poured in PUR or Polyester molding the idc connector with pcb or idc connector with the flat ribbon cable.

[wulfman]

Ok, it seems like you're aiming for a straddle mount socket for the scope connector, what kind are you using?

Like this  https://www.mouser.com/ProductDetail/3M-Electronic-Solutions-Division/9150-4500PL?qs=sGAEpiMZZMs%252bGHln7q6pmzuniGC9vGvgdUx8jl58Y%252bI%3d

I am sure they are found cheaper on ebay

[Muza]

That's the 3M 9150-4500 Series i used the CAD from

[dren.dk]

I did look into USB-C, it does have 4 pairs, but it's not keyed and supporting rotation of the connector would be somewhat hard to support the reversible connector.

USB-C also has built in chips that make it harder to abuse for something like this.

Natually it's possible to use the 4 pairs for the data and then implement a proper +5V power supply that generates the +4V, +2V5 and -2V5 rails and to feed the vref through one of the connections, but it's also somewhat hard and not all that elegant.

I think someone who wanted to (ab)use a consumer cable for a 4 channel pod would be best served with HDMI, though pure mini and micro variants don't seem to be very common...

[Muza]

That's a valid point regarding the polarity with USB-C... which with correct layout would just mean that two channels would be swapped if inserted "upside-down". Not an ideal case.

Going back to the 50-wire flat ribbon cable: maybe can just stack two narrower ribbons on top of each other to make the connection between the scope and the pod less bulky just like the Rigol RPL2316

[wulfman]

The cable issue is why i like the SCSI III cabling

https://www.ebay.com/itm/Internal-SCSI-Ribbon-Cable-2-Drive-with-LVD-SE-Terminator-32-Inch/161643732466?hash=item25a2b79df2:g:HCkAAOSw34FVCz4Z

cut off the extra connector and terminator. 3' cable

[Rerouter]

While I try and figure out how I am going to route the 1.27mm header, here is an 8 Channel pod using LMH7324's,

Not the smallest of things at 41x33mm, and the ribbon is most of that width, but the smaller header / ribbon version is on the todo list.

at a glance it should be possible to make this a 2 layer PCB, but I'm not up for that challenge just yet.

[Rerouter]

And the 1.27mm header version, for a 0.635mm pitch ribbon. This requires a 0.6mm thick PCB to inline the header.

If you want to go more compact than this, it would need to use smaller components, At present I have been working to keep the 0603 minimum as set earlier.

Edit: missed a via, board files have been updated.

[oliv3r]

For the cheap guys: Just solder the amount of chips you have/want on the pcb.. You only have to watch out for the correct pin.

I think  you are hitting a good point. I understand where everybody wants to optimizethe design for their own use case. But going with 4 or 8 channel pods for the 'cheap' crowd, makes little sense, as it's far easier to just not get the LMH chips.

Even with the 4 channel chips the same goes Cheaper = don't solder the chips.

I think Rigol knows what they are doing...

Do we really think that? I mean that Rigol is knowing what they are doing


In any case, good progress is being made.

[Rerouter]

I was working on 8 channel pods, mainly for ease of routing, they have a wrapped around pin order for a dual side loaded PCB, so if stuck to single side loaded PCB's, its easy to break into 2 logical units, but I am including all the design files. so people are free to edit them to there own requirements.

[dren.dk]

Nice, but 0.6 mm 4 layer boards aren't all that cheap:) Perhaps it's ok to accept normal through hole mount, it's not like the straddle mount is that much smaller.

For me it's more important to be able to hand assemble the first 10 units and rework mistakes than it is to the get to the smallest possible size.     

I've been considering using a 2x8 female socket for the input with half of the pins dedicated to gnd. Probe wires with male ends are easy to find, but the really neat way to interface with that would be a single-use board with male headers that can be soldered to the DUT and left there.

I find I often have to solder wires to the DUT to be able to clip test leads or a scope probe on, so having the interface to the LA be soldering to a board seems like it would cut out the middle man.

[dren.dk]

bwt. rerouter: Nice looking layouts there.

I just did the core of my own 16 channel pod layout with lmh7324 yesterday, everything except connectors fitted nicely in 32x54 mm, even with 4mm of space around the ICs for rework, it's funny how easy it is to route on 4 layers, even when dedicating a whole layer to the ground plane.

I don't think I like the 16 channel pod as an idea at all, for three reasons:
* Pulling all the single-ended input connections together into one location might be tricky.
* Pulling the big fat uplink cable from scope to pod is going to be clunky, even if it is 0.025" pitch.
* Aside from the 68 wire scsi cable idea, sourcing a small amount of 50 wire ribbon cables and connectors is quite hard, while 26 pin 0.025" ribbons are much more accessible.

[dren.dk]

rerouter: I see what you were saying about routing the header for the 0.635 mm pitch ribbon, it's annoying as all hell not being able to put a wire between the pads.

[Rerouter]

As per the 0.6mm 4 layer pcb. I may be able to walk that back to a 2 layer 0.6mm pcb. As the thickness is low enough that the impedance widths are easier.

And yeah differential wiring the header was my nightmare when it wasnt inline. It can be done. But its not that easy and definatly seperates the differential pairs for a bit to ensure they are side by side on the ribbon

[wulfman]

Just placed my prototype order b board 2 layer was 50 cents each quantity 10, a board 4 layer was 2.90$ each  quantity 10

IF these work as advertised i will upload all the project files to gitlab for all to use. Or i will have 9 extra boards available 

[dren.dk]

Here's my take on an 8 channel lmh7324 based pod, highlights:

* Still single-sided load, for ease of assembly with a super-generous courtyard around the QFNs for ease of rework, should it be needed, but at 40x42 mm I don't really think it needs to be any smaller.
* I went with a 26 lead, 1 mm pitch ribbon and IDC connectors towards the scope, because I don't want to deal with the nightmare of laying out the 1.27mm pitch header and 26mm is narrow enough for me:)
* The inputs are via a 2x8 2.54 mm pitch female socket, with half of the pins being ground, that should make it easy to plug in a single-use adaptor board or individual probes, I want to plug in my inputs from the top, but it's just as easy to use the same foot print for an angled header or socket if front-entry is desired.
* I want to laser cut an acrylic sandwich case, which basically requires a hole for a screw in each corner, so the board grew 10 mm in width because of that.
* To use the space between the mounting holes for something I stuck a crude voltmeter on the board in the form of an ATtiny85 and 10 charlieplexed LEDs, so it's fast and easy to verify that the threshold voltage is set correctly.
* The Kicad project is on gitlab here: https://gitlab.com/dren.dk/mso5k-la-pod/tree/master/lmh7324-8ch

Roast my layout:)

[Rerouter]

Ok, 1.27mm non-inline header done, I'll make up my ribbon breakout board probably tonight.

[oliv3r]

Beautiful! and really good use of that attiny!! Just 2 question I have.

Why not make it a nice 50 x 50 mm, gives you more room to play with bigger ground cooling plane, and allt hese cheap PCB services houses have their minimum size set to 50 x 50 anyway. Size is probably smallest concern anyway, right?

One question remains, I do not recall the pinout from the scope-end; but do remember the original 2 channel LM's they 'split' the inputs between channel 0 and 8 for the first LM. The second LM goes to 1 and 9 and so on. Now if that's also how the signal enters, then it does not matter I think how you do it.

But with this new redisgn, you will end up with a really odd UI don't you think? So bit 0 - 8 on your probe, will correspond to the odd bits, and 9 - 16 to the even?

[Rerouter]

That's why the breakout board idea, to remap the channels back to the correct order, 2 layers is a pain for routing crossing differential pairs with decent impedance. So at this point I've drifted to 4 layer

[dren.dk]

I started out making my breakout board in 2 layers, but then I noticed that jlcpcb allows me to panelize two pods and a breakout for 15 usd, which means that I can get 4 layer boards for 20 pods and 10 breakout boards in black for 60 USD, so I'm going with 4 layers for my breakout too.

I've been trying to hit the correct impedance with the stackups offered by jlcpcb and I've taken the time to length match the wires of each pair, but not the inter-pair length.

Seeing that it was possible to route the 1.27mm pitch header, I switched to a 1.27 mm pitch header in smd and that freed up an enormous amount of space which made it much easier to route.

The idea with the breakout is to reorder the pin outs so the channel numbering on the scope fits with the numbering of the pods and channel 0-7 goes on one pod and 8-15 goes on the other

My breakout is designed to sit flat against the front of the scope with a tall female socket on the bottom, but it can also be used with a short cable to the scope.

[oliv3r]

ah, right; ok so I really like the breakout board; what connector are you intended to use? I remember there was talk about HDMI cables earlier which make a lot of sense I suppose, but then your short.

I wonder if it would make sense to 'downgrade' to 4 bit pods then. Seeed studio's boards are super cheap even for 4 layers, and if you have 4 bit pods, you still would get 10-ish pods per order (if you use the full 50 x 50 for a pod. If you'd go with 2.5 x 2.5 per pod (very small though) you could put 4 pods on a pcb and thus you'd get 40-ish pods per order!

[Rerouter]

a 8 channel 4 layer pod is only 34x33mm in size, and thats with single layer loading and 0603 components. so you really don't gain anything by going to 4 channel pods.

the breakout board i am currently working on is 65x15mm for reference so far.

there is enough room for 4 pods and 2 breakout boards on a 100x100mm pcb, but that leaves the end user the guillotine them out.

so yep, if you wanted, 20 pcb sets could be had for about $37 posted.

The current ribbon me and dren are working towards are 0.025" or 0.635mm pitch ribbon cable, the same stuff they use on the more professional gear, (very flexible)

[oliv3r]

Going to 4 bit pods could not be so much as for size, but price as well of course. And cables. If a HDMI type A connector is used as a cable for example, which has 4 differential pairs if i'm not mistaken. USB-C connector is 24 pins, but unless you put reversing logic in it, I think you effectivily end up with the same amount, 4 differential channels (the other pins would be unused to allow for connector reversal.

If a 4 bit pod would even be smaller (due to the single LM) and cheaper, it's up to members to decide if they want 1 or all 4. Also scale of economics may be achiveable by group-bying the parts 'kickstarter' like or something?
Just a thought is all...

[Rerouter]

All i can say oliv3r is, I have posted my current design files only a few posts ago, the software is free (Kicad) and the layout I have used can be cut in half or doubled, the only thing they would need to do it is fiddle with the output connector, and run there own breakout board for whichever esoteric connectors they please.

[Noy]

Which connectors/cables do you want to use on the DUT side?
I was thinking about using DSlogic shielded fly wires from my DSLogic but there are some passives used in them.
So i have to buy another set of cables and dismount/mount 0Ohms.

Do you have better ideas?
I think the shielding is good for highspeed signals?

[wulfman]

I doubt shielded will make a diff most of the wires going to the DUT will not be tightly grouped, just left flying.
Testing will show the truth.

[stj]

this is all good, but:
does anybody have a link to images of the internal board in the 1000z series scopes?
that really needs to be cloned.

all the teardowns i'v seen where on the 1054z that does not have the function.

[joeyjoejoe]

For those working with it, the LMH7324 is now in the official KiCad repository - if you're working off GitHub, just do a git pull to get the latest. If you're not working off GitHub, you should be

[Noy]

Any News here?

[dren.dk]

Yep, I received my boards, stencil and all the jellybean parts from jlcpcb on Monday and ordered the expensive parts from Mouser on Tuesday:
https://photos.app.goo.gl/MEetU1SV5LrNyh2t5

One of the parts in my Mouser order is backordered and it's the one part that I have not managed to find anywhere else:
https://www.mouser.dk/ProductDetail/200-ESQ12523LD

BTW, talking over the design with a friend, I ended up sticking a murder of de-coupling 0402 caps on the bottom of the board, because it would be a shame if I left it out and it turned out to be needed.

As soon as I have all the bits I'll fry up a batch and let you know about it.

[Noy]

Hi, can you please provide your mouser project cart?
Which tolerance class 1% or even better for the resistors?
Which condensator C0G or X7R?
Some of the measured values in the schematic aren't E series..?

[dren.dk]

I ordered almost all of the parts from https://lcsc.com so they could be shipped with my pcbs from https://jlcpcb.com.

The Mouser order contained:
https://www.mouser.dk/ProductDetail/926-LMH7324SQ-NOPB
https://www.mouser.dk/ProductDetail/649-221121-00026T4LF
https://www.mouser.dk/ProductDetail/855-M50-3301342
https://www.mouser.dk/ProductDetail/200-ESQ12523LD

I picked X7R caps and 1% resistors.

I've attached the BOM export from LCSC.

I see now I ordered 200 extra LEDs by mistake, good thing they were so cheap:)

[nimish]

I ordered almost all of the parts from https://lcsc.com so they could be shipped with my pcbs from https://jlcpcb.com.

The Mouser order contained:
https://www.mouser.dk/ProductDetail/926-LMH7324SQ-NOPB
https://www.mouser.dk/ProductDetail/649-221121-00026T4LF
https://www.mouser.dk/ProductDetail/855-M50-3301342
https://www.mouser.dk/ProductDetail/200-ESQ12523LD

I picked X7R caps and 1% resistors.

I've attached the BOM export from LCSC.

I see now I ordered 200 extra LEDs by mistake, good thing they were so cheap:)

Assuming this works do you plan on assembling these for sale?

[dren.dk]

That's a hard question.

The raw BOM cost at 10-off numbers is around 100 EUR and that's without hooks and cases.

The sale price a PLA2216 is 300 EUR, so there's certainly room for saving a bit of money, but at low numbers the cost of assembly, shipping and other overhead quickly eats up any savings.

For me it looks as though it's viable to build my own as long as I count the time spent as hobby-time.

If I were to count the time spent as work-time I'd probably have to jump to a production run of at least 1000 units assembled professionally, possibly with custom cases and probes, but that's a huge investment that I have no way of making.

One option to fund the production run would be to set up a kickstarter to ensure the customers are at hand before starting the factory.

Another concern is that I suspect there aren't enough cheapskates with MSO5k scopes in the world that would be willing to plunk down 150 EUR for a knock-off LA probe.

All of this applies to my current design which is more or less a straight clone with a slightly cheaper, but equal performance, comparator chip and the split-pod physical layout.

The connectors and comparators make up 85% of the BOM cost out of that about half is the comparators part.

To make a truly low-cost alternative someone would have to put in the legwork to attack the BOM cost at the expensive end:

1.a: Connector cost could be slashed by finding the parts at the source in .cn in stead of Mouser, I suspect waltzing down to any market and browsing stalls in the pearl river delta would yield massive savings:)
1.b: Another option could be using off-the-shelf consumer cables, but I don't know of any modern standard with 8 LVDS pairs, ground and 4 extra power wires, so that's probably not going to happen.

2: Lower performance comparators must exist, but likely with single-ended output, so one would have to find a cheap LVDS transmitter to pair it with.

[bmx]

That's exactly what I had in mind since the beginning of this thread: Is the rigol price really high? No sure, considering it comes with ``quality`` hooks.

It's pretty hard to find good hooks that don't cost a finger each.
After initial disassembly of the pod, everyone noticed quality components inside, compact design, etc. I'm not quite sure at all rigol makes profit on those probes. All in all, fpc, board, case, cables, hooks, labels, lead time, packing and shipping, it's not worth it.

Unless someone reports on quality inspection of the rest of the rigol probe (plastic, leads, heat resistance, hooks, ..) as a $5 piece of crap.

[joeyjoejoe]

I think the biggest opportunity missed here is the ability to design a simple low cost 4-8 channel version.

[dren.dk]

I have just assembled the first set of pods and breakout board: https://photos.app.goo.gl/qMiUNHk9Tk6krjD9A

Holy hell, those 250 parts by hand at 0603 was hard without going insane.

I ended up sticking 48 0402 10 nF caps on the back, so each of the 12 power pins per IC has a super-local decoupling cap.

I still want to play around with a cheaper and lower performance frontend.

[MegaVolt]

Input band at the probe 200 MHz. Perhaps changing the input chains can raise this band to 500? Sampling frequency allows.

[felix ch]

I have just assembled the first set of pods and breakout board: https://photos.app.goo.gl/qMiUNHk9Tk6krjD9A

Does it work as you expected?
If so, I'll going to make it.

Thanks for you a lot!

[dren.dk]

Not quite.

Good news first:
* I have verified all the output voltages on the breakout board without the pod, so that's nice.
* I have also verified the connections from the breakout to the pod and that there are no shorts.

Then the bad news:
* I fucked up the charlieplexing for two of the LEDs, so that will require either a new board or some bluewires.
* I'm seeing -2.5V on Vref when the pod is plugged into the scope, which should not happen as nothing but high-impedance inputs are connected to Vref and no shorts to any power rail.

[dren.dk]

Close inspection of the board file revealed that the problem was that I fucked up and overlapped the -2.5V and +4V zones so they ended up shorted.

The good news is that the MSO5k has a nicely robust set of power supplies for the LA probe, they survived the abuse just fine.

I've created a new board and I'm just waiting for another layout for the same order before sending it out.

The changes are:
* Fixed the Charlie plexing fuckup.
* Fixed the power rail short.
* Moved the parts away from the edge, so there's less risk of damage when depanelizing.
* Wired the ISP tracks out to the panel frame, so it's easy to program without fiddling with the annoying SOIC clip.

[johnmx]

Close inspection of the board file revealed that the problem was that I fucked up and overlapped the -2.5V and +4V zones so they ended up shorted.

Always run the Design Rule Check before sending the design for production. Most people don't care about this tool, but I think it is essential.
I am only happy when it returns 0 errors.

[thm_w]

Always run the Design Rule Check before sending the design for production. Most people don't care about this tool, but I think it is essential.
I am only happy when it returns 0 errors.

Wow, that is wild that people would not care, it should be run as default function when generating gerbers. Of course the best is live DRC, but that is probably much harder to implement software wise.
Good to know it didn't blow the Rigol, but I'm thinking to add some 1A fuses on the 4V/2.5V rails, just in case.

[dren.dk]

I certainly always run DRC before tapeout, but in this case it generated a single warning that didn't sound dangerous and when I went over the board I could not see any problems at all, so I figured that it was a false positive.

Good thing 4 layer boards are so cheap now, so it's only a 50 USD lesson:)

BTW: Someone warned against having decoupling caps on the bottom of the board because it would be too hard to assemble, but that was really not a problem at all.

[voltsandjolts]

Maybe they meant it is more hassle for automated assembly and reflow - all components on one side would have to be glued in place?

[dren.dk]

Well, caps on the bottom get held on with the surface tension of the solder during reflow of the top side, so there's really no reason for glue.

... unless you have very heavy components on the bottom that is.

[felix ch]

I was noticed you ordering PCB from szjlc.com and soldering it by yourself.
Szjlc provide SMT service for one side parts in chinese. Not for every single part of course, but resistors and capacitors are supported.
I'm not sure if the service is available for you.

[dren.dk]

Yeah, I assumed that they would have that service, though I admit I didn't look into it as I assumed that the startup cost would be too high for a single prototype.

I just looked around for it, but I could not find any information on how to submit such an order from the international site, so I think the assembly service is still Chinese-only.

[johnmx]

I was noticed you ordering PCB from szjlc.com and soldering it by yourself.
Szjlc provide SMT service for one side parts in chinese. Not for every single part of course, but resistors and capacitors are supported.

Those SMT services in China are cheap, but there is a catch.
You have to choose if they use their parts or you send the parts to them.
If you send the parts, the shipping cost plus the increase of customs taxes when importing the boards will make the final price too high. So it will be cheaper to assembly by my self.
If they use their parts, the circuit quality will be extremely affected. For sure they will use the cheapest and crappest ceramic capacitors they can find. Probably they will even use counterfeit ICs.

[ebclr]

What evidence do you have of this assumptions ?

[SilverSolder]

I would love an assembly service, even if it was just for the passive components.

[dren.dk]

I finally received the updated PCBs monday and yesterday we assembled two sets.

There's a photo of the frame just before going into the the oven.

We took a flir image of the board after letting it run a little and it seems the solid ground plane does it's job and is very much needed as the comparators still got up to over 60 degrees even with the copper plane spreading the heat.

Two things of note about the boards and stencil from https://jlcpcb.com/
* One via on one of the pods was disconnected, so we had to run a wire, I checked out a couple of the other boards, but did not find the same problem, so it's certainly not a design problem. Perhaps I'm pushing capabilities by using 0.3 mm via holes and would be better served with 0.4 mm drills and larger pads in future designs.
* I wanted to have alignment holes in the stencil, but for some reason jlcpcb ignored my paste gerbers and produced a stencil from the soldermask, thus handily ignoring my alignment holes that I designed into the stencil, drilling those holes was super fiddly.

I remember reading about the stencil problem, but it slipped my mind when I placed the order; aparently the way to getting the stencil you want from JLCPCB is to order the stencil separately from the PCB with only the paste gerbers in the zip file.

[dren.dk]

Note also that the scope itself gets quite toasty, I wonder what this will do for the life time.

One of my collaborators is in the process of designing a 3d printed case which will be printed with slightly conductive filament and get lightpipes for the LEDs and a bottom covered by an 1mm silicone sheet.

[TK]

In my opinion the LEDs should be removed from your design, they don't add any information to the end user and can be distracting.  They  might be useful only after assembly if you just want a visual confirmation of the rails being there, but it is adding complexity and power consumption to your design.

[dren.dk]

I see, good thing we didn't make it for you then.

Beside your personal preference, everything you said is just plain wrong:

0: If someone wanted to build one of these pods, then it would be very easy to leave out the offending parts and the space they take up was added to be able to have proper mounting holes, so there's no disadvantage to removing the option.
1: The LEDs show the current threshold voltage, so it's easy to notice if it has been set incorrectly, without digging down to the appropriate menu on the scope.
2: IOW: the LEDs do not show the power rails are there, but rather make critical information more readily available exactly where it's needed.
3: The added complexity is 2 extra BOM items (LEDs and micro controller, the resistors and the decoupling cap are the same as other items already used) in total it's 10 extra parts, the total number of parts on a pod is around 50.
4: Seeing as how the comparators dissipate enough power to reach 60 degrees, nobody is going to notice the  extra 10 mW taken by the microcontroller and a led.

Now, if I was building a mass-market product, then bean counters might force me to drop the reference-voltmeter option or start optimizing away decoupling caps, but as there is probably never going to be built in more than 5 of these sets, then cost optimization is not at all a priority.

The most important priority is that I build what I want, especially when the cost is basically zero.

[TK]

The most important priority is that I build what I want, especially when the cost is basically zero.

I was just trying to comment on your design, as you made it public on this forum.  The idea of sharing something in eevblog is to get other's opinions, right?

[TK]

I see, good thing we didn't make it for you then.

Beside your personal preference, everything you said is just plain wrong:

0: If someone wanted to build one of these pods, then it would be very easy to leave out the offending parts and the space they take up was added to be able to have proper mounting holes, so there's no disadvantage to removing the option.
1: The LEDs show the current threshold voltage, so it's easy to notice if it has been set incorrectly, without digging down to the appropriate menu on the scope.
2: IOW: the LEDs do not show the power rails are there, but rather make critical information more readily available exactly where it's needed.
3: The added complexity is 2 extra BOM items (LEDs and micro controller, the resistors and the decoupling cap are the same as other items already used) in total it's 10 extra parts, the total number of parts on a pod is around 50.
4: Seeing as how the comparators dissipate enough power to reach 60 degrees, nobody is going to notice the  extra 10 mW taken by the microcontroller and a led.

Now, if I was building a mass-market product, then bean counters might force me to drop the reference-voltmeter option or start optimizing away decoupling caps, but as there is probably never going to be built in more than 5 of these sets, then cost optimization is not at all a priority.

The most important priority is that I build what I want, especially when the cost is basically zero.

Your eyes will be 90% looking at the scope anyways and not trying to look at some LEDs blinking on a sea of wires.  At least that is my experience debugging logic circuits.

[dren.dk]

That's fair, I probably ended up sounding much more argumentative than I wanted, sorry if that was the case.

The point of the leds is not to blink out information, but to make it clear what the threshold is set at when I'm hooking up the probe, which is especially important and easy to miss if both pods are in use as the two groups of inputs have separate threshold voltages.

The state of the leds only change when changing the threshold voltage, so there should not be any need to stare at them or to get distracted.

[TK]

Good points about the LEDs and the threshold settings

[oliv3r]

just make sure the leds are dimm enough to not be a distraction; using large resistors to reduce the current to 5mA may be just the solution Personally, I like the idea of the LEDS

[dren.dk]

Don't worry about the LED brightness I'm very annoyed by overly bright LEDs myself.

The LEDs used are these ones: https://datasheet.lcsc.com/szlcsc/Lite-On-LTST-C193TGKT-5A_C12065.pdf

From the datasheet we get a forward voltage of about 2.5-3.1 V, so let's call that 2.8 V and a max current of 20 mA.
The supply to the MCU is the -2.3/2.5 V rails, so that's 4.8 V in total.

I designed with 2x 220 Ohm resistors for a peak current of (4.8V-2.8V) / 440 Ohm = 4.5 mA.

I'm running the LEDs with a fixed rate charlieplex where each LED gets one out of ten PWM periods, so that means that the average current at 100% PWM is 0.45 mA per LED.

As I'm not a fan of bright LEDs I've turned down the PWM to 1/255, which yields an average current of 2 uA

The initial result was too bright, so I upped the resistors to 2x500 Ohm for a peak current of 2 mA and that seems about the right brightness.

I must say I'm very surprised about how much light is yielded by the charlieplexed LEDs even at a crazy low peak current.

We assembled serial number 3 last night, here's the recipient of #3 stuffing 0402 10 nF decoupling caps and swearing a lot.

#3 ended up working perfectly.

The recipient of #2 is in the process of printing cases for us.

The firmware for the voltmeter is also working perfectly, so after calibration the exact threshold voltage is reflected by the LEDs.

[Sprite_tm]

As a warning for other people that try to DIY this using an LMH7324: there are fakes of this chip on the Chinese market, as I can unfortunately tell you from experience... I built my pod, and 8 of the 16 channels refused to do anything. Luckily, I own an advanced device for looking inside of chip packages to debug this issue... Seems they forgot to put the actual silicon into half my chips, grmbl. FYI, these come from Taobao; that was the easiest source I could get them from seeing as I live in mainland China.

Images (click to embiggen):

[pmnxis]

I ordered PCB and parts with following this thread.

Thanks. I will try to assembly and test on my new mso5000.

in my case I spent 150usd more for parts and pcb. but it's interesting and cheaper than pla2216.

BTW I changed some connector and cable assembly for rapid works (there's no stock for connector, so changed little bit)

And I don't have experience about making IDC 1.27 cable by hand. so ordered already made one.

[thmjpr]

Started on the USB C probe idea from Muza giving 4 signals per probe pod, 4 pods total. Schematic is basically straight copy of dren.dk then the connectors were changed out.

Main connector: https://www.digikey.ca/product-detail/en/sullins-connector-solutions/SFH11-PBPC-D25-ST-BK/S9201-ND/1990094 $3
Type C connectors: https://lcsc.com/product-detail/USB-Connectors_Jing-Extension-of-the-Electronic-Co-LCSC-USB-3-1-C-TYPE-DIP-SMTFemale-10-0L-High-temperature_C168698.html $0.50

Board is 1.6mm, ideally would be 2mm to match the header connector pin spacing but JLC just has their cheap impedance controlled board in 1.6 (JLC7628).

In retrospect, HDMI is a much better solution in almost all aspects: you don't have the polarity issue, and proper spec cables are easier to find. Probably why we've seen HDMI on siglent scopes and on CC payment terminals, etc. But, once you realize the quirks of USB-C its not too bad, its just that there are sooo many of them. Most cables wouldn't work, flipping the cable will of course screw up your channels (if you want to be lazy you can simply mark the "up" end on your cables), and active cables won't pass through our signal or could blow from the voltages we are using.

For finding the right USB cable, what you want is a passive "super high speed" or "USB 3.1 gen 2" or "10Gbps". These, if correct, indicate that all four high speed pairs are wired inside the cable. I bought two cables and it looks like they both are suitable (very surprising given how cheap they were, about 19 wires! need to be run and cables are hand assembled for the most part). They have the labeling: "VW-1 type 2725", unsure what it refers to (related).

Cable 1: https://www.aliexpress.com/item/30cm-USB-C-USB-3-1-Type-C-Male-Connector-to-Male-Data-Cable-for-Tablet/32825241462.html $3
Cable 2: https://www.aliexpress.com/item/USB-Type-C-Cable-USB-3-1-Gen2-USB-C-Male-to-Male-Data-Sync-and/32893896190.html $6

First cable is way too short (30cm), I bought it thinking it would be fake/mislabeled, but its not, has all the right pins wired up.
Second cable appears to be very high quality, again same internal wiring. Of course I can't tell the power cable diameter inside or bandwidth of the HS pairs, so quality is purely based on physical feel.

Neither of these cables seem to have the internal 1k pull-down resistor on CC2 line, which I was expecting. Also, all multi-pin connections VBUS/Gnd/etc. are connected internally on the cables. Knowing these two things would have made layout a bit simpler. Shield is internally connected to Gnd as well, no messing around with having that connected in some strange way.
For the CC2 line, basically there are two polarity detection lines, CC1 and CC2. One is connected through the cable and one is not. So based on where that pin ends up, you can determine which one of the four orientations the cable is in.

[gizmo5418]

Hi,

I want to thank you for this great LA project. I successfully made LA to MSO5000. All channels work, but channels 1, 3, 5, 7, 9, 13, 14, 15 have a negated level, as in the screenshot of the MSO5000. Voltages -2.5V, + 2.5V, + 4V, Vref1, Vref2 are correct. I noticed the same negation of channels in the user's picture dren.dk.
It seems to me that the signals of negated channels in the LA project are reversed - one should introduce a correction on the PCB Breakout..

I am sorry for my English, I use a Google translator.

Regards,
Pawel

[pmnxis]

also i had same issue.

[dren.dk]

Sorry, I fell off the project and just came back to the thread, cool to see that people have actually built the boards!

I just did a basic smoke test and put the project on the shelf, so I never noticed that some channels were inverted, what an embarrassing mistake to make.

I wonder where the error was made, I guess I'll have to do some measuring to figure out where the signal is swapped, but the only way it could be someone elses fault than mine is if it's the pinout I used for the scope.

The best solution would be to redo the breakout board, but it ought to be possible to swap the affected pairs in the cable and put the connector back on.

[pmnxis]

Thanks for comeback and your participation for made them!.

There's another issue that some channel cannot change state or mixed remarkable noise.
I cannot sure mine was broken while connect cable with wrong angle with short time.
or cable or assemble is strange.
I will try to figure out what was problem
I hope there's no problem on my scope

BTW are there any people have same noise issue like me?

[gizmo5418]

Hi,

I have noise if I have a "PCB Breakout" attached without connected logic probes.

pmnxis, connecting the ISP programmer at the same time that LA was connected to the oscilloscope was probably not the right solution.
Preservation of my oscilloscope with "PCB Breakout" plugged in without logic probes:

[pmnxis]

thanks for you pictures.
Mine got some "too much" noisy issue on 6th channel when didn't connect cable 0to7 channel.
BTW I solved issue that channel 1ch(start from zero) didn't work.
I was soldering each only one point to 2 bypass capacitor near the MCU.
and re-soldered array capacitor. and then it works.

for now, a left issue is a known issue. some channel is inverted.
hope it would be nice fix later. but I think that's not immediately to me and enough for see extra point for change or triggering something.

[thmjpr]

Can confirm the noise, even with the Logic analyzer completely disconnected, some channels seem to have injected noise and go wild. Wonder if a high value (100k) pull-down on the connector board would help to get rid of that. Seems odd that Rigol would leave it high-z.

Also confirm the original pinout of the connector has many signals flipped (D1, D3, D5, D7, D9, D13, D14, D15), see attached for what it should be. On mine, dumb luck meant I flipped one of the pinouts of the LM7324, so I only have these flipped: D3, D7, D14, D15. I'll think if I want to mod the main board or the pods to fix this.. Once I've done some high speed testing, can send out USBC boards if anyone wants the extras (knowing that you have to do some mods).

Noticed that 2V5 seems to get pulled down a fair bit, with just one pod connected I was reading 2.2V. Not sure what the wire gauge in my USB cables is.
Also if it wasn't clear (because I didn't realize this), the Vref is: -1.5 to +1.5 to represent a -15 to +15V trigger level. How I have it now, the STM32 can only read positive trigger voltages (0 to 1.5V).

[Rerouter]

100% certain on D11 and D14? they break the pattern of all the others

[TK]

Can confirm the noise, even with the Logic analyzer completely disconnected

By completely disconnected you mean the minimum required for the Scope to detected and activate the logic analyzer, correct?  The microcontroller that was added for the LEDs, is it active when you detect the noise?

[thmjpr]

See attached for 40MHz SPI signal on analog + logic input, seems to work quite well other than a bit of jitter.
One issue is there is a ~150ns offset between the analog channels at the top, and the same digital channel on the bottom. I'm thinking this is worth reporting as a bug.
One possible workaround for this would be adjusting the analog channel offsets, so all are -150ns. But that is a lot so it may not be possible.. will check.

Temperature is 55C on the chip and 45C on the bottom of the board (in open air at 23C). It will get a bit worse inside the case, but would likely be fine without a heatsink.

By completely disconnected you mean the minimum required for the Scope to detected and activate the logic analyzer, correct?  The microcontroller that was added for the LEDs, is it active when you detect the noise?

Yeah thats right, nothing connected to the Rigols 50-pin port at all. Sometimes the LA will run for a bit, even after unplugging the pod (some bug or timing detection). But you can manually short pin 1 to ground too, to enable the LA and test this way as well. I suppose in Rigols design case it doesn't matter, as you either have all channels connected, driving the inputs hard, or not connected and LA disabled on the screen.
Good point about the micro though, I could see it having some effect.

100% certain on D11 and D14? they break the pattern of all the others

It is an odd pattern, and it would make more sense if it was N, P, N, P repeating consistently. Actually now that you mention it, the bit order does have a "logical" pattern: if you skip two pins, it does go from D0, D1 .. to D7, and the same for D8 to D15. At first I thought it was completely nonsensical.

I will go over them again tomorrow now that I've modified my pod to swap D4 and verify the pinout (having the resistor pack made it easier, see photo below).

[thmjpr]

With the pod now having D1 and D3 flipped, the signals D11 and D14 are still backwards. So they don't match the pattern seen on the other signals. To eliminate all possible sources of error it could be checked manually at the input pins.

Digital to analog delay cant be adjusted with ch-ch skew adjustment. This only delays the analog channel further forward on the x-axis (+/-150ns max). Maybe someone with the official logic pod can verify they see this digital-> analog delay as well. Adding a fixed offset in firmware would be nice, an adjustable offset in the menu even better still.

Channel to channel jitter is OK, I will get a shift back and forth of 1ns on D3, D2, D1 when triggering on D0 for example. Rigol spec is 2ns typical, 5ns maximum. Sample rate is 1GS/s at 200MHz, which all adds up.

Going to print the pod case again, font was too small to show up. Clear color would be cool too, to see the LED inside.
Adapter case coming soon.

[Rerouter]

Does the Channel - Channel skew line up with your trace lengths, if not we may be able to forward correct for it on a new PCB version.

[thmjpr]

Does the Channel - Channel skew line up with your trace lengths, if not we may be able to forward correct for it on a new PCB version.

The trace lengths were matched to under 1cm or so, speed of signal in FR4 is ~15cm/ns, so 70ps. I can't see it making too much difference, its well under their specifications. The skews sort of just jump around, if I averaged a bunch of captures then I might see the relationship you are referring to.

If we wanted to extend the traces to make up for the 150ns LA->analog delay in the scope, we are talking about 22m of trace! So I'm sure the official Rigol pod does not line up with the analog channels either.

I originally had the case try to duck under the cutout in the rigol scope, but it didn't quite fit. So pulled it forward to be in front instead. The OLED module sits inside that top bezel part.
If you don't need a screen (its not as necessary as I thought because the idle signals will be clearly inverted if the cable is upside down), then removing that top makes it a simpler design.

[Rerouter]

Ah sorry, misread, though it was digital - Digital skew, I would assume the rigol would have a setting somewhere to adjust digital skew.

[TK]

Today I noticed the LA can be activated without the digital probe connected (I don't have one).  What is the purpose of the probe detect on pin 1?

I don't have any noise in the digital channels when the LA is active and there is no probe connected to the scope

[thm_w]

Today I noticed the LA can be activated without the digital probe connected (I don't have one).  What is the purpose of the probe detect on pin 1?

I don't have any noise in the digital channels when the LA is active and there is no probe connected to the scope

You get to adjust LA settings and see what is in the menus with nothing connected which is OK. But yeah the software should probably auto-disable the channels from showing up on screen.

The noise will appear as soon as you short pin 1 and 2 together. So the probe detect pin is enabling buffers/inputs, might save some power and reduce noise, assuming its a hardware disable. The calibration menu option is also disabled.

[John_Cavanaugh]

Any thought on using mini displayport for cables?   They have orientation and same lanes as hdmi.

Cables are pretty cheap.   https://smile.amazon.com/Cable-Matters-Mini-DisplayPort-Black/dp/B0777RKTJB/ref=sr_1_8

But I didnt check on the connector prices.    I suspect the usb-c are much cheaper...

[thmjpr]

Added code for stm32 here: https://gitlab.com/thmjpr/stm32f03_la_monitor
Will show cable status ("A: -" means A cable not connected, etc.) and the two threshold voltages (0 to +15V).

I have one partial built usbc main board and un-built PCBs if anyone is interested. 3D printing is possible but would cost more for shipping.

BTW there is a SCPI command for LA time offset, :LA:TCALibrate 0.0000002 +/- 200ns but its not working (read back is always 0), suspect in another thread that its not actually been implemented yet.

Any thought on using mini displayport for cables?   They have orientation and same lanes as hdmi.
Cables are pretty cheap.   https://smile.amazon.com/Cable-Matters-Mini-DisplayPort-Black/dp/B0777RKTJB/ref=sr_1_8
But I didnt check on the connector prices.    I suspect the usb-c are much cheaper...

That would be a good choice. Its also possible to mix it up with mini->normal size cables as those are more common, but $9 is great already and seems robust. Connectors, I see two on digikey for $2-3 each:

https://www.digikey.ca/product-detail/en/te-connectivity-amp-connectors/2129320-3/A120537CT-ND/5021748
https://www.digikey.ca/product-detail/en/pulse-electronics-network/E9320-001-01/553-3756-1-ND/4169580

[jMachina]

Wow, I am so glad that I just made this account! Would you be willing to share design files and firmware so I can make a run of PCBs?

Thank you kindly,
jMachina

[thmjpr]

Wow, I am so glad that I just made this account! Would you be willing to share design files and firmware so I can make a run of PCBs?

Thank you kindly,
jMachina

Thats great. I've added files for revision 1 board here: https://gitlab.com/thmjpr/stm32f03_la_monitor/tree/master/PCB_Rev1

As mentioned in PM, I would recommend going with that and modding the pod PCBs, or getting me to do it. That way you'd have 14 usable channels out of 16, then two inverted. Shipping the extra boards is not expensive. But if you want to wait for Rev2 files that is OK.

[thmjpr]

Revision 2 PCB is here: https://gitlab.com/thmjpr/stm32f03_la_monitor/tree/master/PCB_Rev2

- See readme.md
- A 50 pin connector available from LCSC but has no polarized part in the middle, so could accidentally be plugged in upside down. One polarized version is: SFH11-PBPC-D25-ST-BK (digikey)
- CC1/2 resistor values might need to be played with or voltage measured, can't recall if I used what was on the schematic.

Attached some parts I bought from LCSC but its not the complete BOM, just use for sourcing ideas.

[AngusBeef]

I'm a little uncertain reading through this thread - does the logic analyzer you guys designed improve anything off the Rigol design, or is it simply a matter of reducing the $400 price tag off the Rigol product to BOM plus hobby time?

[spongle]

Revision 2 PCB is here: https://gitlab.com/thmjpr/stm32f03_la_monitor/tree/master/PCB_Rev2

- See readme.md
- A 50 pin connector available from LCSC but has no polarized part in the middle, so could accidentally be plugged in upside down. One polarized version is: SFH11-PBPC-D25-ST-BK (digikey)
- CC1/2 resistor values might need to be played with or voltage measured, can't recall if I used what was on the schematic.

Attached some parts I bought from LCSC but its not the complete BOM, just use for sourcing ideas.

This is fantastic, if you have any spare boards / parts you'd like to sell I'm interested.

[thmjpr]

I'm a little uncertain reading through this thread - does the logic analyzer you guys designed improve anything off the Rigol design, or is it simply a matter of reducing the $400 price tag off the Rigol product to BOM plus hobby time?

Its mostly reducing the price tag. There are some small improvements, like if you want a longer cable you could buy a different length USB-C cable.

This is fantastic, if you have any spare boards / parts you'd like to sell I'm interested.

I have 3x rev 1 bare boards, I don't plan on buying rev2. $4 for just PCBs in a bubble envelope, modding required. Will send a PM for the pre-made board price.

[whowe]

Hello thmjpr,
Is Rev2 like a "perfect" version so far?
Really love your USB-C idea. I'm thinking to make one and guess it can bring some fun time with my new MSO5072.

[tv84]

BTW there is a SCPI command for LA time offset, :LA:TCALibrate 0.0000002 +/- 200ns but its not working (read back is always 0), suspect in another thread that its not actually been implemented yet.

First of all, congratulations for your work! 

Now, my small contribution:

Those SCPI commands are implemented in the FW. I've checked them in the latest FW.  (and there are plenty of other functions that are not currently exposed to users, but equally implemented)

Don't know if you are aware of all the MSO5000 SCPI commands that are currently externally exposed? but maybe you have to manually enable "calibration mode". Try these commands:

:CALibration:ENTer
:CALibration:EXIT

I think Rigol has been using this type of protection in the latest equipments so that one doesn't execute unwanted calibrations. My guess...

Edit: These commands are only for the analog channels! 

Do you do a :LA:ACTive command before the calib to choose the specific digital channel before setting the delay?

[thmjpr]

Hello thmjpr,
Is Rev2 like a "perfect" version so far?
Really love your USB-C idea. I'm thinking to make one and guess it can bring some fun time with my new MSO5072.

Thanks, the ideas are mostly combinations from others in this thread.

I haven't built V2 so I'm not sure, it will still measure only positive trigger voltages, and cable orientation resistors may need tweaking. AngusBeef has ordered boards so he will be the one to try it. Although sounds he will be busy with moving shortly, perhaps after that he will be able to sell the extra PCBs.

BTW there is a SCPI command for LA time offset, :LA:TCALibrate 0.0000002 +/- 200ns but its not working (read back is always 0), suspect in another thread that its not actually been implemented yet.

First of all, congratulations for your work! 
Now, my small contribution:
Those SCPI commands are implemented in the FW. I've checked them in the latest FW.  (and there are plenty of other functions that are not currently exposed to users, but equally implemented)

Don't know if you are aware of all the MSO5000 SCPI commands that are currently externally exposed? but maybe you have to manually enable "calibration mode". Try these commands:

:CALibration:ENTer
:CALibration:EXIT

I think Rigol has been using this type of protection in the latest equipments so that one doesn't execute unwanted calibrations. My guess...

Edit: These commands are only for the analog channels! 

Do you do a :LA:ACTive command before the calib to choose the specific digital channel before setting the delay?

Thank you, there is a LA threshold calibration as well but I'm not sure why that one doesn't work, it should be reading the voltage internally. Can't recall if we've heard from someone with the official probe if they can confirm. Anyway, it will just cal the trigger threshold voltages, so its not too critical, as they can be read out or manually adjusted if ever needed.

For ":LA:ACTive" command, that is the same as the LA -> Select option on the menu screen. So you can select none, D0, D1, etc. But I tried with either of those and still the LA:TCAL command does not save any offset data. What we want is an entire shift in time from all digital channels to all analog channels.
I could see individual channel adjustment being useful in some niche cases, but IMO its something left to very high end logic analyzers.

[AngusBeef]

I got one of the Rev2 logic pods done, I'm short a few caps that'll come on Wednesday so I can finish the other 3.

I will probably finish the rev2 main board tonight or tomorrow and so I *should* be able to test it on Wednesday assuming I track down any shorts or non-connected wires. I did the comparators with a cast iron skillet and my stove  and everything else with a soldering iron so I have run across several components I had to fix up already. I messed up one comparator which hurt my soul a lot.

I already have some boards I screwed up, but assuming all goes well in completing this, I should have 2 main boards and 7 logic pod boards left over. I'll be packing up my place in about 4 days but I'll have them available when I put stuff into storage in another state mid-October and then whatever is left is coming with me to Europe.

I really only started learning electronics about 3 months ago and focusing on it about a month ago, so this is my first 'real' circuit besides breadboarding with an Arduino so I'm just trying to keep my head above water here.

Edit: There could be something worthwhile with the LCSC assembly service seen here - I'm already committed to doing this project but someone else may want to check it out. - https://www.eevblog.com/forum/manufacture/jlcpcb-smt-assembly-service-208764/?topicseen

[AngusBeef]

Got two logic pods and the main board done. With just the main board in the scope, I have a lot of noise as mentioned above. But when I plugged my pods in, they light up for a second and then turn back off. I think either I messed up all of my USB connections since I tried the majority of them, or the USB-C cable I have lying around is not a Gen-2 so it may not work correctly. I may get a chance between now and next week to take a peek, otherwise I'll probably get it running by mid-October.

One change I had to make was using an 1uF electrolytic cap for C11 because I must have already packed up my 1uF ceramic caps. I've got it in polarity correct ( ) so my assumption was that it will work the same in this situation. Would love to hear some feedback whether or not it's true. From the schematic it looks like the STM32F030F4P6 and all the associated passives are for the OLED screen. Is it possible that they're causing issues if I don't have the screen hooked up yet?

I'll clean up the flux and upload a photo later when I get a chance.

[whowe]

I also have made a copy of your Rev2.
I planned to use JLC, but the verification girl insisted it's a "5-board" one and the price was almost 100 USD. I failed to correct her that it was "2 designs only" board. Finally I had to change to another vendor who does not understand impedance part.
Anyway, the board is being shipped to me now. And I also purchased 2 LMH7324 directly from TI (FedEx from Texas) which is expensive but should be genuine.

Hoping for the best result.

[thmjpr]

Got two logic pods and the main board done. With just the main board in the scope, I have a lot of noise as mentioned above. But when I plugged my pods in, they light up for a second and then turn back off. I think either I messed up all of my USB connections since I tried the majority of them, or the USB-C cable I have lying around is not a Gen-2 so it may not work correctly. I may get a chance between now and next week to take a peek, otherwise I'll probably get it running by mid-October.

One change I had to make was using an 1uF electrolytic cap for C11 because I must have already packed up my 1uF ceramic caps. I've got it in polarity correct ( ) so my assumption was that it will work the same in this situation. Would love to hear some feedback whether or not it's true. From the schematic it looks like the STM32F030F4P6 and all the associated passives are for the OLED screen. Is it possible that they're causing issues if I don't have the screen hooked up yet?

I'll clean up the flux and upload a photo later when I get a chance.

A random USB-C cable is unlikely to work, as most will not have all wires connected internally (passively). But you can check by plugging it in and beep out the connections if you want. I linked two specific alix vendors earlier in the thread.

Caps won't make a huge difference so should be fine for the moment.

STM32 is purely for the OLED display, if you don't use the display, don't need to populate it or supporting circuitry. It isn't connected to any of the differential signals, and the only connection to Vref0/1 is to monitor those voltages. So just 100 ohm resistors + fuses on the power supply rails would be needed if you want to operate the board passively.

[whowe]

I also have made a copy of your Rev2.
I planned to use JLC, but the verification girl insisted it's a "5-board" one and the price was almost 100 USD. I failed to correct her that it was "2 designs only" board. Finally I had to change to another vendor who does not understand impedance part.
Anyway, the board is being shipped to me now. And I also purchased 2 LMH7324 directly from TI (FedEx from Texas) which is expensive but should be genuine.

Hoping for the best result.

Got them. Although the PCB manufacturing is really not good.

[ebclr]

Do you have any spare set of boards to sell/share

[whowe]

I would like to share them, but I'm in China. Oversea shipping is costy.
And what's worse, is that I might have made a mess. It's not working, yet.
Could be bad PCB manuafacturing, could be bad soldering, or maybe partially wired USB cable. I don't know yet.
Will update later, mmm, when I figure out what's wrong there.

[luma]

  the various silkscreen jokes on that PCB are great!

[whowe]

Check this problem.
Without connecting any satellite adapters. Some channels are high, some are low. Especially the D6, it has a 32MHz (not accurate) pulse. Don't know where it picks up from.
As a reference, a put a 10MHz sine ware on Analog Channel 1. See the screenshot.

[thmjpr]

Check this problem.
Without connecting any satellite adapters. Some channels are high, some are low. Especially the D6, it has a 32MHz (not accurate) pulse. Don't know where it picks up from.
As a reference, a put a 10MHz sine ware on Analog Channel 1. See the screenshot.

This is normal. Once you plug in a working adapter all channels on that adapter should drop low (if its high cable is backwards). I was considering terminating the diff pairs, specifically D6 to try to get rid of this trace "noise" but I got down to 10kR and it was still there so I stopped. Also it would take a ton of resistors.

The issue is some internal noise in the oscilloscope and the fact that they seem to have not pulled/terminated the signal inputs in any way. Which is probably an OK decision if you want to save some cents.

Turn off any channels you are not using to avoid the distraction.

[whowe]

Check this problem.
Without connecting any satellite adapters. Some channels are high, some are low. Especially the D6, it has a 32MHz (not accurate) pulse. Don't know where it picks up from.
As a reference, a put a 10MHz sine ware on Analog Channel 1. See the screenshot.

This is normal. Once you plug in a working adapter all channels on that adapter should drop low (if its high cable is backwards). I was considering terminating the diff pairs, specifically D6 to try to get rid of this trace "noise" but I got down to 10kR and it was still there so I stopped. Also it would take a ton of resistors.

The issue is some internal noise in the oscilloscope and the fact that they seem to have not pulled/terminated the signal inputs in any way. Which is probably an OK decision if you want to save some cents.

Turn off any channels you are not using to avoid the distraction.

Thank you.
When probe 1 (D0-D3) is connected, D0 has opposite polarity than D1-D3. Same thing happens with probe 3 (D8-D11), D11 is opposite to D8-D10. (Maybe the pin definition is incorrect?)
And as probe 2 and probe 4, it's a mess, too much noise, I even can't tell the polarity. Maybe the voltage reference pin is floating because of bad soldering.

The good thing is that I've finally found a USB type-c cable that works.

My bad, I forgot to solder F1-F3 and R5/R6, I thought they're only for the optional MCU. lol

Everything works as expected now!

My next step would be 3D-printing a proper case/handle for the main board.



Again, thank you thmjpr and all other friends who devoted time and knowledge in this thread/project.

[whowe]

A small tip for choosing USB type-c cable: buy those 3.1 Gen2 ones (with or without e-Markers does not matter).
Those cheap ones claimed as "JEM" or "Google" 3.1gen2 cables are known working.
Mine in the attached photo is called Google cable, 0.5M length.


Here's the link for these cables, visit and scroll down you can see detailed pictures (even teardown) of them.
https://item.taobao.com/item.htm?spm=a230r.1.14.16.41cc4735N9EF3q&id=574580414281&ns=1&abbucket=3#detail

[thmjpr]

A small tip for choosing USB type-c cable: buy those 3.1 Gen2 ones (with or without e-Markers does not matter).
Those cheap ones claimed as "JEM" or "Google" 3.1gen2 cables are known working.
Mine in the attached photo is called Google cable, 0.5M length.
(Attachment Link)

Here's the link for these cables, visit and scroll down you can see detailed pictures (even teardown) of them.
https://item.taobao.com/item.htm?spm=a230r.1.14.16.41cc4735N9EF3q&id=574580414281&ns=1&abbucket=3#detail

Thanks, those look like decent cables, incredible for $3-4. Also interesting about the emark chip, it appears to just connect to CC1/CC2, so the only possible issue it would cause would be there, not a big deal.

I've uploaded solidworks files here for a beta case: https://grabcad.com/library/beta-la-case-parts-1 (I think login is needed, can use bugmenot.com). Nothing really holds it together, my plan was to glue it once the board was fully modified.


edit: I may still have D11 flipped on Rev 2, not sure, user WhoAmI has measured it as being OK, thank you.

[hedehede81]

guys, is there an estimate how much it costs right now to make this cable?(counting time as hobby time) is there any possibility we can see a production of these(Kickstarter etc)? or waiting for Rigol cable to go on sale makes more sense?

[joeyjoejoe]

LM7324 is going to be your biggest BOM cost, and then the 4-layer boards themselves (22USD shipped).

I think it's challenging to have someone on Tindie/kickstarter produce this. The demographic is people who are rather experienced/technical, so there's not much value in selling these since most likely the people buying it could make it themselves as well if they wanted to save money. (If they didn't want to save money they'd buy the Rigol unit).

[hedehede81]

For my Fluke 289 there was also a proprietary IR serial cable, which after sometime appeared on Taobao as a bootleg. Maybe this can also make its way some day. As you said, finding LM7324 at a reasonable price point is the bottleneck here, Digikey wants to sell 1000 units to you where as for Mouser international shipping is very expensive. I never ordered anything from TI directly, however I doubt they are cheaper for only 4 chips. If there was an alternative board with LM7322 things can be easier since it is available on Ebay. There is an alternative PCB on gitlab for LM7322 but as an unfinished experiment I think?

[Noy]

Any news here?
Bugs with the wrong signal (high/low) fixed?

In a few weeks i will build a pod but with the "Classic" design. 1 Pod , 2.54mm IPC cable, 1.27mm socket for using my coax cables from the Ds Logic +. And i want to know If there are still some issues with the schematic? The LMs are expensive so hopefully only one shot.

[TopLoser]

Just to try and save people time and effort (and maybe money!)... I have a number of boxed, unopened, Rigol MSO1000Z Upgrade Kits (RPL1116, connecting cable and set of grabbers) for sale.

£95 plus shipping (free UK, £10 anywhere else).

[thmjpr]

Any news here?
Bugs with the wrong signal (high/low) fixed?

In a few weeks i will build a pod but with the "Classic" design. 1 Pod , 2.54mm IPC cable, 1.27mm socket for using my coax cables from the Ds Logic +. And i want to know If there are still some issues with the schematic? The LMs are expensive so hopefully only one shot.

If you are using the Kicad design you will have to fix the swapped +/- of some of the differential signals.

Micro-coax probes is a good idea, although the dslogic has embedded resistance in the connector. So you could consider what effect that has on the signal levels and signal integrity.

[Noy]

Yes i want to use the Kicad Design.
Which differential signals have to be swapped? Is there a list with the wrong ones?

I know about the embedded resistance. I bought a second set of cables and will replace them  with zero ohm resistors.

[memeruiz]

Great thread!

Awesome contributions from dren.dk and later from thmjpr and whowe, thanks guys!

I have seen the gitlab repo from dren.dk, but not the one with the USB-C modifications. Is it somewhere? Would be nice. Maybe I can make some contributions to that version.

Thank you!

[thmjpr]

Yes i want to use the Kicad Design.
Which differential signals have to be swapped? Is there a list with the wrong ones?

I know about the embedded resistance. I bought a second set of cables and will replace them  with zero ohm resistors.

You'd have to compare the connector pinout to what is shown in Rev 2 schematic. I was told the pinout is correct by users above, but I have not personally tried that board.

Great thread!

Awesome contributions from dren.dk and later from thmjpr and whowe, thanks guys!

I have seen the gitlab repo from dren.dk, but not the one with the USB-C modifications. Is it somewhere? Would be nice. Maybe I can make some contributions to that version.

Thank you!

thank you, https://gitlab.com/thmjpr/stm32f03_la_monitor

If you need any specific info/files let me know via PM.

[Rerouter]

If it was my version, looking way back at my last upload, I did have pins flipped, and looks like I never uploaded after fixing it, so here they are minus the last few connections I got stuck on.

If the other, If you can point me to what is wrong, I should be able to correct them,

[Gandalf_Sr]

I've been looking at this thread with interest; I'm close to ordering an MSO5074 but I'm not keen on paying an extra $360 for the PLA2216 cable (36% of the cost of the scope).  I looked on Digikey and there are other LVDS comparators available although many seem nearly as expensive as the TI part suggested.  The difference in price between the MSO5074 without/with the PLA2216 is $360 and you can buy the probe on its own for $400 so I may well just buy without for now.

So I'm considering a poor man's logic probe that would be good for up to 100 MHz ish, I realize low propagation delay may be a necessity for digital timing alongside the analog signals but some logic families such as TI's AUC go down to 1.5 nS propagation delay although I think that the MSO5000 needs +/- LVDS signals.

Then I looked on TIs website and came up with the DS90LV047A 4-channel LVDS driver IC which are $2.62 each so about $11 to get 16 channels, and have a 1.7 nS propagation delay, that's more like it!

Anyone think this could work?  Tell me I'm insane but, if this is a reasonable idea, let's get the circuit designed! I will create the schematic do the PCB layout in Diptrace.

[MarkL]

...
So I'm considering a poor man's logic probe that would be good for up to 100 MHz ish, I realize low propagation delay may be a necessity for digital timing alongside the analog signals but some logic families such as TI's AUC go down to 1.5 nS propagation delay although I think that the MSO5000 needs +/- LVDS signals.
...

Just a comment on the propagation delay...

If you're concerned about matching the analog channel delay to a home-brew digital probe delay, the Rigol 5000 has a channel delay adjustment.  It's not clear from the user manual if you can adjust the digital channels, but you can certainly move the analog channels relative to the digital.  And although it doesn't say, I'm a sure a few nanoseconds would be within its range.

Most scopes have this, even old analog scopes.  Also known as channel-to-channel skew adjustment.

[Gandalf_Sr]

OK, here is the simplest form I can imagine for my design idea:

1 There could be 4 x DS90LV047A converting single-ended 3.3V CMOS logic levels into LVDS differential signals.  Cost about $11
2. A single 74AUC16244 acting as a 16 bit buffer and voltage level translator following the advice in the TI little logic guide starting on page 19.

The AUC device at 2 above could be multiple devices e.g. 4 x 4bit buffer.

The different voltage levels being sensed could be handled by different Vcc values used for the AUC buffer.  How did the original design do this and was there some sort of setting in the MSO5000 to define what was being sensed?

If the DS90LV047A devices convert single ended into LVDS signals then I think it makes more sense to have them on the smaller PCBs close to the point of sensing and then send LVDS signals over the ribbon cable to the board that plugs into the MSO5000.

What about the input protection circuitry?

Comments please.

[Gandalf_Sr]

OK, to self-comment.  I've been looking for comparators that would be a front end for the DS90LV047A but they are either too slow or too expensive so the LMH7324 seems like a good option and, at 700 pS propagation delay, it's the fastest that TI has listed.  However, the minimum Vcc is supposed to be 5V, was that catered for in the designs you guys have implemented?

[TopLoser]

There's only a maximum of 4V supply available on the scope connector.

Start reading the thread from the beginning, it's got answers to lots of your questions.

[Gandalf_Sr]

There's only a maximum of 4V supply available on the scope connector.

Start reading the thread from the beginning, it's got answers to lots of your questions.

I've read the whole thread a couple of times.  I saw the comment that the comparator only had 4V and needed 5.  Someone responded that there was a -2.1V available on the scope connector which could be added to the +4V, so in the designs, does the comparator run on 4V or more?

Maybe it runs OK at 4V?

[Gandalf_Sr]

I've ordered my MSO5074 - without the PLA2216 because I refuse to pay 36% of the cost of the whole scope for a couple of ICs, a bit of cable, and the clip thingies.  I looked at the good work done here to reverse engineer the PLA2216 but the end result seems too complicated and expensive - mainly because it seems to emulate a probe that might be used on very high frequencies; much higher than the 200 MHz that the LA inputs are rated for on the MSO5000.

So I plan to design and build a simpler, lower frequency probe that uses cheaper DS90LV047A LVDS drivers and a simple 74AUCxxx front end buffer with open drain outputs that will allow it to perform voltage level translation using a single resistor per channel.  The 74AUC family is newish and has very low propagation delays of around 2 nS. I will post a new thread for this which I will come back and link from here but I do have a couple of questions that I hope you guys can answer that will speed me on my mission:

1. Regarding the 1/10 reference voltages that the MSO5000 puts out:
    a. Is this as straightforward as it sounds? For 3.3V logic the scope outputs 0.33V etc?
    b. How many different voltages are selectable from the scope's menu?
2. The LVDS buffer IC I'm considering shows that the output is done by sending current out of the +/- output pair and requires a 90 - 130\$\Omega\$ termination resistor across the output to generate the +/- voltage signal.  My driver IC will run at 3.3 volts and my question is - is the termination resistor required or are resistors already there internal to the MSO5000 LA input?
3. I plan to use a controllable Vcc to the 74AUCxxx buffer to adapt to the input voltage selected - this is going to be a microcontroller which I'm well able to do but I would prefer a simpler solution - any thoughts or ideas?

Thanks in advance.

[ebclr]

Interesting Idea, Keep us informed on the development

[thmjpr]

1. Regarding the 1/10 reference voltages that the MSO5000 puts out:
    a. Is this as straightforward as it sounds? For 3.3V logic the scope outputs 0.33V etc?

Yeah, its commented in my code somewhere. Roughly -1.5 to +1.5V

    b. How many different voltages are selectable from the scope's menu?

-15V to 15V in 10mV steps if I remember correctly.

2. The LVDS buffer IC I'm considering shows that the output is done by sending current out of the +/- output pair and requires a 90 - 130\$\Omega\$ termination resistor across the output to generate the +/- voltage signal.  My driver IC will run at 3.3 volts and my question is - is the termination resistor required or are resistors already there internal to the MSO5000 LA input?

It looks like there are 33 ohm series resistors. I don't think there is any internal parallel termination, but I can measure it later.
https://gitlab.com/riglol/rigolee/-/wikis/MSO5000-teardown

[Gandalf_Sr]

Thanks so much for the information

2. The LVDS buffer IC I'm considering shows that the output is done by sending current out of the +/- output pair and requires a 90 - 130\$\Omega\$ termination resistor across the output to generate the +/- voltage signal.  My driver IC will run at 3.3 volts and my question is - is the termination resistor required or are resistors already there internal to the MSO5000 LA input?

It looks like there are 33 ohm series resistors. I don't think there is any internal parallel termination, but I can measure it later.
https://gitlab.com/riglol/rigolee/-/wikis/MSO5000-teardown

Hmmm, this is worrying.  If the current driven out of the LVDS doesn't generate a big enough voltage across the 33 resistor then my whole approach isn't going to work without changing the internal resistors which I'm not going to do.  I think the only thing to do is to get an LVDS IC (I'm planning on the DS90LV047A) and try it from a breadboard to make sure it triggers the level changes accurately.  It puts out a nominal 20 mA which, if driven across a 33 resistor, would give 0.66 Volts across whatever IC is receiving the signal.  I looked at the GIThub tear down but couldn't see what that might be?

[EDIT] Oh, you said SERIES resistor so maybe if I include 100 parallel resistors just before the connector, they will feed through to the MSO5000 and work properly? Will need to test.

I have created the 50 pin connector component in Diptrace and the screen capture of it is below.  Can anyone confirm what pin 10 is and whether the other pins are correct, particularly the Ps and Ns?

[thmjpr]

Hmmm, this is worrying.  If the current driven out of the LVDS doesn't generate a big enough voltage across the 33 resistor then my whole approach isn't going to work without changing the internal resistors which I'm not going to do.  I think the only thing to do is to get an LVDS IC (I'm planning on the DS90LV047A) and try it from a breadboard to make sure it triggers the level changes accurately.  It puts out a nominal 20 mA which, if driven across a 33 resistor, would give 0.66 Volts across whatever IC is receiving the signal.  I looked at the GIThub tear down but couldn't see what that might be?

[EDIT] Oh, you said SERIES resistor so maybe if I include 100 parallel resistors just before the connector, they will feed through to the MSO5000 and work properly? Will need to test.

I have created the 50 pin connector component in Diptrace and the screen capture of it is below.  Can anyone confirm what pin 10 is and whether the other pins are correct, particularly the Ps and Ns?

Series resistors, yeah. In the teardown photos they go through the resistor pack, a via, then disappear to an inner layer. I can't measure any relevant sort of parallel resistance on the connector itself, its ~1M. So probably goes straight into some ICs input pins.
Connector pinout looks correct.

I don't even know if parallel resistors would be needed on your board? Unless you had them for signal integrity purposes. No harm in adding a footprint for them though.

The 74AUC gates look cool https://hackaday.io/project/28833-microhacks/log/157535-just-how-fast-are-74auc-gates

[Gandalf_Sr]

thmjpr
Thanks for the commentary. The AUC gates are cool but when I asked on TI's e2e forum about implementing voltage level translation using open drain outputs and pull up resistors, the expert advice was that, although this works, it's very power inefficient and the value of the pullup resistor needs to get smaller as the signals go faster.  They recommended the SN74AXC8T245 which is an 8-bit voltage level translator that uses uAmps and I've got one of these per 8-bit pod.

I will start a new thread soon and link to it from here.

[EDIT] First vn is up in the thread here

[joeyjoejoe]

I think a lower cost pod is a great idea. As a hobbyist the 5000 series scope is at a nice price/feature point, but a LA at 200MHz just isn't needed. My cheap chinese one has been doing fine, so a cheaper pod would also be fine.

[Gandalf_Sr]

My boards are due to arrive today but now I'm starting to worry that the LVDS output of the DS90LV047A will not work properly with the LVPECL input (assuming that is what it is) for the MSO5000 LA interface connector.  The LMH7322 outputs a differential voltage that swings between 0.35 and -0.35 volts  (see first picture) but I've made my DS90LV047A LVDS output circuit drive into a 100 Ohm resistor so my output is not going to centered around zero.  Have I fucked up?

I think I may need to drive into 2 x 50 Ohm with a center offset voltage but it's not clear to me from the DS90LV047A datasheet that this is needed (see second picture).

If this is a mistake I've made, I may be able to make these boards work by putting 2 x 50 Ohm resistors in place of the 100 Ohm one and then feeding a 1.2V supply into the center point of the resistors.

Any comments are welcome.

[Vtech]

It's been a while since I've worked on the probe.

The LMH7322 outputs a differential voltage that swings between 0.35 and -0.35 volts  (see first picture) but I've made my DS90LV047A LVDS output circuit drive into a 100 Ohm resistor so my output is not going to centered around zero.  Have I fucked up?

Why do you think it's zero centered? It's RSPECL. At this VCCO voltage (2.4V) it should output levels H=1.3V, L=0.9V. These should be exactly LVDS levels (common mode voltage around 1.2V).

I can't measure it currently becouse I'm on a leave. I'll be back at work in two weeks.

[Gandalf_Sr]

It's been a while since I've worked on the probe.

The LMH7322 outputs a differential voltage that swings between 0.35 and -0.35 volts  (see first picture) but I've made my DS90LV047A LVDS output circuit drive into a 100 Ohm resistor so my output is not going to centered around zero.  Have I fucked up?

Why do you think it's zero centered? It's RSPECL. At this VCCO voltage (2.4V) it should output levels H=1.3V, L=0.9V. These should be exactly LVDS levels (common mode voltage around 1.2V).

I can't measure it currently becouse I'm on a leave. I'll be back at work in two weeks.

Thanks for the fast response. 

I was looking at the schematic in this post and the driver has +2.4 and -2.4 volts hooked up to it so I was assuming that the output was centered on zero.  Can anyone else confirm whether the voltage on the differential pair is centered on zero or 1.2V?

[EDIT] It does appear that, the way that the real LA probe circuit is connected, the LMH7322 Vcco is set at 2.5V which makes the H/L outputs run at +1.5V/+1.0V which is LVDS-compatible, so I think I'm OK (see picture).

[Gandalf_Sr]

I built one and it works - at 10 MHz in my tests so far.

See these new pictures in my thread.

Thanks for all the support and ideas from this thread.

[EDIT] I have now run my budget probe set at 50 MHz (square wave 3.3V logic levels) and it works as expected but there is an occasional 1nS jitter on the signal.  How does this compare to the real LA probes and the build specified in this thread?

[bigboss59]

So far there is no RPL1116 design ?
(I know it can be easily adapt; I don't want to invent the wheel again)

If not is there anybody have one set of the MSO to sell ? (preferably in Europe)

BR

[koljsch]

Hello!
Someone can tell - after assembling the circuit on the pins, where the voltage should be -2.5 V, the voltage rises to 0.9 V. What is the problem?

[oliv3r]

Here are some photos of RPL1116 Active Logic Probe pod used in MSO1000Z series scope.

Scope has 16 channel logic probe. The probe pod uses 8x LMH7322 (Dual 700ps High Speed Comparator with RSPECL Outputs). 4 comparators on each side. The pod is connected to the scope using 68 pin connector and flat cable. Reference voltage is generated inside the scope.
Cable connector has:
16x2 pins for PECL signals,
22x GND between diff pairs + some extra,
6 pins for VCC
2x2 pins for +/- supply voltage
Some 4 more signals (reference voltage??)


Most likely, the same design is used in new pod for MSO5000 series - PLA2216 Active logic probe. Here are some x-rays of the new pod: https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2055715/#msg2055715

Vtech, I know we kind of hi-jacked your thread with the MSO5k probe.

I too did a design https://www.eevblog.com/forum/testgear/another-low-cost-la-probe-for-rigol-mso5000-by-oliv3r/ but am really interested if we can port it also to the DS1000Z series. Do you still have one? Would you experiment/help. I'd really need the pinout of course, and while looking at those two pictures, I could get like 33% of the pinout, the other 66% probably needs to be measured. E.g. which pin is what channel, which pins are gnd, which pins are power, and what voltages at that. I know it'll be quite similar to the MSO5k, but still.

[seppeltronics]

Hello,

for the MSO500/PLA2216:

1) Are the schematics and layouts which are a 100% Clone ?
2) Is the circuitry in the genuine big connector and the pod's ?
3) Is there a good clone that is commercially available ?

Thanks a lot.

Best Regards, Seppeltronics.

[Mechatrommer]

Hello,
for the MSO500/PLA2216:
1) Are the schematics and layouts which are a 100% Clone ?
2) Is the circuitry in the genuine big connector and the pod's ?
3) Is there a good clone that is commercially available ?

Thanks a lot.
Best Regards, Seppeltronics.

to answer your PM as well here...

the original circuit and IC teardown and reverse engineering for RPL1116 was made by Vtech...
https://www.eevblog.com/forum/testgear/rpl1116-active-logic-probe-pod-for-1000z-series-teardown/
https://www.eevblog.com/forum/testgear/rpl1116-active-logic-probe-pod-for-1000z-series-teardown/msg2088667/#msg2088667

the pin connector for PLA2216 (please note PLA 2216 and RPL1116 have different pin out into the scope) was described by TopLoser (and maybe others)...
https://www.eevblog.com/forum/testgear/rpl1116-active-logic-probe-pod-for-1000z-series-teardown/msg2086255/#msg2086255

there is no voltage regulator in the pod, you can see from photos here.. Q3: lets hope there will be soon. currently there are few cheap clones based on some driver/logic IC for less than $100, perhaps similar to this...
https://www.eevblog.com/forum/testgear/low-cost-logic-analyzer-probe-for-rigol-mso5000-easyeda-project
https://www.eevblog.com/forum/testgear/another-low-cost-la-probe-for-rigol-mso5000-by-oliv3r

you can follow those useful threads or maybe this thread in case if you want to build your own.  cheers.
https://www.eevblog.com/forum/testgear/low-cost-compatible-rigol-pla2216-logic-probe-for-dho900-(and-hacked-dho800)/msg5352677/#msg5352677
https://www.eevblog.com/forum/testgear/low-cost-compatible-rigol-pla2216-logic-probe-for-dho900-(and-hacked-dho800)/msg5412416/#msg5412416

[seppeltronics]

Hello,

the issue that I have is that these EEVBlog-Forum-Projects contain hundreds of posts and if you do not read them all you might miss information. So I wonder if these are in a git-repo somewhere? Got it: https://gitlab.com/riglol/lapod

I did the calculation, about 60..120€ for the comparators(),... + enclosures, PCB's,... if that is no production-run that is commercially available, it's really hard to do it for less than the original probes?

Best Regards, Seppeltronics