Products > Test Equipment
LaPOD: Low cost Logic Analyzer probe for MSO5k, DHO900 and more!
oliv3r:
--- Quote from: ebastler on November 02, 2023, 08:43:23 am ---
--- Quote from: oliv3r on November 01, 2023, 10:22:31 pm ---
--- Quote from: ebastler on October 28, 2023, 02:14:02 pm ---@oliv3r, is there enough room for the HDMI plugs' plastic bodies? Looks like the jacks are spaced very closely together.
--- End quote ---
crap! I Didn't even think about that one ... i'll measure my cables and adjust accordingly. Only bit that's frustrating, is that'll probably mean I have to do my differential lines, again, and the longest one will probably be ending up even longer :S
--- End quote ---
Making the PCB wider might also run into collisions with the connector on its left -- even more so on the DHO900, which has a USB jack quite close to the logic analyzer port. Maybe look into mini-HDMI or mini-DVI again?
--- End quote ---
Hmm, I'll check the sizes again, my el-cheapo cables are at 21mm, so 22.5 mm would be the minimum with for each connector. That would yield 900mm total width, which seems to fit the MSO5000 it seems. Using a ribbon extension cord is an option, but defeating the purpose of course :D
But if you find a mini-hdmi port on lcsc we can talk :)
What is spacing like on the DSO900? That USB connector is awfully close. I was thinking of doing double sided connectors (J1 J2 top, J3 J4 bottom, but those solder lugs get in my way. Offsetting things so both holes overlap, e.g. 2 connectors in 1 hole is an option, just not solderable one. Getting connectors without the lugs is an option, but makes it much to easy to 'rip off'...
I suppose a nother option is an ugly one, current design for the MSO5000, an asymmetrical for the DSO9000, and have it stick out only to the right side. Since the boards carry no components, they are cheap as butt of course, just the design effort.
--- Quote from: ebastler on November 02, 2023, 08:43:23 am ---
--- Quote ---8 cables is still too much for most of my tasks and still a big-ish cable. Also, those flat cables aren't differential cables, which is why I dislike them. 4 was the sweetspot for me, not just because hdmi has exactly 4 differential channels, but also because 4 is what you'd use for i2c + irq, or SPI + irq etc.
--- End quote ---
Agree that 8 signals plus 8 GND wires is a lot already. I am not sure why all the individual GNDs would be needed though. My preference for an 8-signal pod would be a 2*5 pin connector (8 signals, 2 GND). That also works well with an IDC connector and colored ribbon cable, with the colors nicely matching the channel numbers.
--- End quote ---
I think this also comes from the fact, that you may want to have a (coax) wire per signal, and connect each individually to power and signal? Use nice coaxial mini cables? idk, i'm not EE enough (at all) to know what's the better design idea here. For me, using a single row or double row header doesn't make a difference, and I think a 5 pin (or 6) single row might actually be wider then the HDMI plug :)
--- Quote from: ebastler on November 02, 2023, 08:43:23 am ---
--- Quote ---I think [...] nikki's v3 are then just what you are after.
--- End quote ---
Nikki Smith did an updated version of the low-cost (TTL/CMOS) probe only, right -- or did I overlook something? I definitely want a probe with the large voltage range and configurable threshold provided by the LMH7324, since one of my use cases is work on vintage (tube) computers with 20V logic levels.
--- End quote ---
Very true, it was the first design I looked at actually, because it was in kicad, and I had forgotten about dren.dk's, which I re-found and used to do the LM variants :)
--- Quote from: ebastler on November 02, 2023, 08:43:23 am ---The design by @dren.dk is the configuration I personally like best, https://www.eevblog.com/forum/testgear/rpl1116-active-logic-probe-pod-for-1000z-series-teardown/msg2316423/#msg2316423. It needs some work though: Some signal polarities are flipped due to early mis-information about the 50 pin connector's pinout, and he did not bother with length matching of the traces.
But I am in no rush -- will wait how your design pans out, and probably borrow (and use in duplicate) your very nice LMH7324 pod layout. :-+
--- End quote ---
Keeping the HDMI connector on the pod isn't unreasonable, because you can't make the pod much smaller anyway, as most space is used by those 2012's that are needed. You could use smaller components there, but then traces could become an issue ...
--- Quote from: UK on November 02, 2023, 09:30:35 am ---
--- Quote from: ebastler on November 02, 2023, 08:43:23 am ---Making the PCB wider might also run into collisions with the connector on its left -- even more so on the DHO900, which has a USB jack quite close to the logic analyzer port. Maybe look into mini-HDMI or mini-DVI again?
--- End quote ---
Why not just use vertical ones?!
Previously, I also suggested mini-HDMI and mini-DP, but after looking at the range of cables available I agreed that the ordinary HDMI is the best solution... since slim connectors, flat or soft versions of cables are not even available for mini-HDMI / DP / DVI ...unless you're a fan of cables that look like a garden hose on your table ;)
--- Quote from: oliv3r on November 01, 2023, 10:22:31 pm ---... that'll probably mean I have to do my differential lines, again, and the longest one will probably be ending up even longer :S
--- End quote ---
If you going to redo all differential lines again )) pls look closely at the upright HDMI sockets I've already mentioned before. Then you won't have to make a board that looks like a palm tree. And all four HDMI sockets can fit on a pcb in the width of a 50-pin header.
--- End quote ---
I actually had considered those as well, (I need HDMI connectors for another project actually, and I really wanted a normal and a reverse (e.g. upsidedown) HDMI connector) but these things are almost impossible to source ...
So the vertical ones, are ugly :p and then stick out either at the top, or the bottom (and I can't place the LED in the center :D) But yes on all other points. But the breakout board can be done in various versions :p
Not that cost is the major driving force (well the topic suggests otherwise), the vertical ones do cost 10x more then the normal ones (65c vs 7c)
Also, I'm not convinced it would make layout any easier, the board might end up being even longer due to the size of the connector, and on top of that the connector sticks out quite a bit, and we still need some jigglies to route the pins ...
So while I certainly feel the challenge, I'm worried about its use/success :p Unless I hear a lot of shouts to do it anyway :p But feel free to try :D I'd share my PCB file, but the forum won't let me :( so i've renamed it to .txt :D
mwb1100:
--- Quote from: UK on November 01, 2023, 07:22:29 pm ---I've dived a bit deeper into the MSO5000 look and made several touches... now it definitely looks in its authentic style.
--- End quote ---
Those do look nice!
oliv3r:
Them new dimensions look something like this, 85mm it looks like atm, dunno if i can route it though :)
ebastler:
Thanks for sharing the draft. For the DHO900 series, 85 mm would already be too wide, I'm afraid. As mentioned earlier, I don't have the scope (yet). But I took a photo from the net and used the connector width for scale -- attached.
The circuit board itself would probably already get in the way of inserting a USB stick or plug. If you add an enclosure for the breakout board or an HDMI plug in the leftmost position, it would definitely block the USB port.
LCSC do offer mini- or micro-HDMI connectors, by the way -- they just don't put that in the part name and have no way to filter for them... ::) Part numbers C2962409, C2962410, C720620 for example are all micro-HDMI connectors which are in stock and cost around $0.50 each. C2962409 and C720620 are also available for JLCPCB's assembly service. Not the friendliest parts for PCB layout though: Those pads are fine-pitched, and I don't see a way to get to the inner row without dropping down to another layer.
Edit: Switching to mini-HDMI would add a bit of parts cost, both for the jacks and the cables. But not significant compared to the LMH732x cost, I'd say. If you don't want to bet on mini-HDMI entirely, there could always be two versions of the breakout board. Probes should keep the full-size HDMI connector in any case, since cables with mini-HDMI on both ends seem very uncommon.
Edit²: Removed the picture, since the scale was not correct. See UK's post below for a corrected picture, which shows an even bigger interference.
Edit³: Oops, the HDMI part numbers above are actually micro HDMI (type D), it seems. They would definitely fit, but may be unnecessarily finicky (both from a layout perspective, and regarding handling/robustness of the plugs). Proper mini-HDMI (type C) connectors are available too, e.g. C2682170. More layout-friendly since they bring all pins out in a single row. Also they are mechanically more sturdy than the micro variant, and should be small enough to fit the four connectors on a breakout board which is not wider than the 2*25 pin port.
ebastler:
One more comment, if I may, regarding the LMH7324-based pod from reply #155:
The blocking capacitors for the various supply voltages are place a bit too far away from the LMH7324, for my taste. Especially when considering that we are already making compromises with the inductance by using the larger 0603 parts and 100 nF caps only (as used by Rigol too), vs. the (additional?) 10 nF suggested '7324 datasheet.
If you drop the differential output signals down to another layer right away, you would have room for the capacitors directly to the left and right of the '7324. You could bring the differential signals back to the top layer in a convenient place, just below the RN1 and RN2 arrays. That would also allow you to use arrays for the 220 Ohm termination resistors, placed just below RN1 and RN2 respectively.
Oh, and one more thought: When length-matching the four channels on the pod, it should be sufficient to make the total of "trace length on the input side + differential trace length on the output side" equal for all channels. It looks to me like, at the moment, you have separately balanced everything on the input side, and also on the output side. Maybe you can save a few wiggles and some PCB space?
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version