Rigol's New DHO800 Oscilloscope unbox & teardown

[Fungus]

Will the DHO800 run on a plain 12V source like a cig socket in the car?

I don't see why not.

Watch out though, in old cars the cig lighter might be connected directly to the alternator. Rev the engine and... 

(I finally got to use that emoji!)

[ptluis]

Will the DHO800 run on a plain 12V source like a cig socket in the car?

Yes but for your own safety its better to not make a cig adapter -> USB C cable that outputs 12V without negotiation.
As mentioned previously in this thread you can blow up devices. Not to mention possible voltage spikes.
A suitable USBC PD car adapter that is safe to use can be had for <$30.

The first power supply shipped with the scope only have 2 wires positive and negative, this means it's possible to build a powerbank outputing 12.6V (3S pack 18650 or 21700 cells with BMS off course), and feeding the scope without the need of any kind of negotiation.
I wonder what's the minimum voltage the Scope runs. Anyone tried?

With a car, the regulator inside the alternator stabilize between 13.8V up to 14.8V depending on car make and model. The only risk is if the alternator regulator fails and then yes magic smoke.


First power supply shipped with DHO800:

[EEVblog]

Will the DHO800 run on a plain 12V source like a cig socket in the car?

It negotiates 15V @ 2.5A from the USB-C, so likely, no. You could use could use a 12V to USB-C adapter

[PELL]

The RK3399 CPU is quite struggle.
So that may explain why things are really slowing down when using the protocol decoder.
I installed a Launcher and a Guesture software btw 
Then you can use split screen.
The RIGOL.SCOPE software is monitored by a daemon, If the SCOPE software crashes, it will automatically restart.

[Fungus]

I installed a Launcher and a Guesture software btw 

With ADB?

[Dacian]

I played a bit with the DHO804 and frequency generator and zooming on vertical is super buggy with cut out waveform and completely wrong voltage. You need to always change the horizontal time base in order for the signal to be displayed correctly.
Seems inexcusable to ship a product with this sort of bugs as this will be the first things you will test.
It is like they have a single software guy working on this and no test engines. Even in that case I do not see how such basic errors where not detected as zooming a signal both horizontal and vertical should be the first thing you test after any change.
Hardware seems excellent so I hope they will fix this soon. I seems to have version 1.00 maybe it is some sort of early release software.

I see comments about the RK3399 struggling. This is a very powerful SOC for an oscilloscope so what it is struggling is the software the is improperly written.
I find it decently responsive but have not tested the protocol decoders up to now. I prefer to use a computer based logic analyzer for that sort of work.

[PELL]

I installed a Launcher and a Guesture software btw 

With ADB?

Of course.
But the ADB and SSH leave wide open also makes me wonder if someone can just hack in my scope and "rm -rf /*" all my scope files 

[ebastler]

Will the DHO800 run on a plain 12V source like a cig socket in the car?

It negotiates 15V @ 2.5A from the USB-C, so likely, no. You could use could use a 12V to USB-C adapter

You forgot about the early samples which were shipped with the fixed 12V supply. The scope clearly can run from a 12V supply, but if the power supply offers USB PD negotiation, the scope negotiates 15V. 12V was available in the USB PD 1.0 standard, but is no longer supported in versions 2.0 and 3.x in my understanding.

[Malmen]

So maybe not ideal to run this from the car battery directly then. And I suspect those cheap 65w PD cig plugs adds lots of noise. But I also have a 4300mAh Lifepo4 S4 pack that gives me about 13,3 volts. That would be safe to use then. Would not expect more then an hour or so of use from it but that's ok.

[Fungus]

I played a bit with the DHO804 and frequency generator and zooming on vertical is super buggy with cut out waveform and completely wrong voltage. You need to always change the horizontal time base in order for the signal to be displayed correctly.

Yeah, that was shown in Daves video at 36:18, here: https://youtu.be/S8jrpCoZyx8?t=2155

Seems inexcusable to ship a product with this sort of bugs

You should send yours back immediately. 

[dreamcat4]

Yes but the general question still remains:

to test the scope actual performance characteristics across a variety of common and different types of power sources. To make sure that a poor quality of power does not adversely affect or detriment scope measurements.

So having some standardized set of tests (that uses a generally affordable and widely commonly available other test equipment). Makes sense. That user A can test with power source X and user B can test with power source Y, yet to be able to meaningfully compare the scope performance results. And across a variety of metrics or modes.

And we don't necessarily have to throw our hands in the air and give up just because theres a lot of random chinese junk flooded in the market.

For example, the 2 known manufacturer provided power bricks might be a starting point. However we could then also consider other **quality** and commonly available (reputable) power banks, or otherwise step-down voltage converters (as some high quality and known electronic module, to then buffer / plug into the car, or otherwise step-down from higher battery voltages).

The point being if you can step down voltage. Then the input power source can just be some cell pack which is smoothly and reliably not introducing noise itself... rather you are only having to worry about the step-down regulator for contributing noise. And to have it efficient enough (so switching probably). To then be efficient not to unduly drain the cell pack too quickly above the already astronomical idle power draw of this scope.

Of course we can also consider popular and readily available brands like Anker or Apple, or such other ones. For their USB-c power banks too. But those doesn't converge on a common homebrew solution (the internals). So is less widely applicable in that respect to only be of value to those other users who happened to pick the same exact retail product.

[Fungus]

to test the scope actual performance characteristics across a variety of common and different types of power sources. To make sure that a poor quality of power does not adversely affect or detriment scope measurements.

It's 36W from a switching power supply so I suspect it's already dealing with a lot of ripple, etc. It must have some fancy filtering and power conditioning on the analog area of the PCB or it wouldn't work at all.

So... I suspect it will work well with just about anything.

[Fungus]

So maybe not ideal to run this from the car battery directly then.

You could try looking at the output of your socket with the engine switched off (I assume you wouldn't try to use it with the engine running)

(Use an oscilloscope, not a multimeter...  )

[Veteran68]

You forgot about the early samples which were shipped with the fixed 12V supply. The scope clearly can run from a 12V supply, but if the power supply offers USB PD negotiation, the scope negotiates 15V. 12V was available in the USB PD 1.0 standard, but is no longer supported in versions 2.0 and 3.x in my understanding.

My understanding is the PD spec still supports the 12V profile as optional. So charger manufacturers can choose to implement it or not. The minimum that must be supported are 5V/9V/15V/20V to be considered PD compliant, with PD 3.1 adding support for 24V/36V/48V.

I just got a new 65W PD charger last week that still supports 12V/3A. So they're still plentiful I think, you just need to check specs before you buy if you need 12V support.

[Neutrion]

Just a few questions:
Did I understand right that the waveform update rate was down to 1000/sec with only one channel enabled at 50 ns/s with full mem depth ? And the screen actually even showed a lower than 1M mem depth?   And it went up only when Dave lowered the mem. depth?
How is the update rate with two channels enabled?
Because the 1104x-e has many times of that with full mem depth and two channels enabled in the same group.

Also interesting weteher it is possible to have a zoom in view on the vertical scale without changing the vertical scale, so getting the getting the full height of the signal?

So to ansver my question after watching Daves video again.
With a 10 mHz signal the scope only gets the advertised 27000/sec display waveform update rate if the mem depth is set at 250 !! points, which seems to be even lower than the 1Megpoints the scope has with all channels active.
When it goes to 650 points mem. depth, the update rate slows down to 1000/s (50ns/d)

As Dave edited in the 1104x-E, with 17000/s I have to add that it is wrong, as the max of it with single chanel active with 14Megpoints is around 107000/20000 dots/vector mode and  with even two channels active it never goes below 3000 but mostly it stays above 6000/sec.
With a single channel active it mostly stays above 10000.
So the 1104x-e has at least 10X the displaywaveform update rate it seems, but if we go down to a few hundred points mem depth it possibly goes much higher, the official max number is 100000 with whatever setting.

So either the Rigol keeps those numbers with more mem. depth, and more channels active, but even than it is still much slower than the advertised value, or if it goes still lower, than that is a bit atrocious.

Edit: OK I see I forgot about Dave turning on everything on the Rigol which bogs it down to about 2500/sec with 1000 points mem. depth. Not tested on max. mem depth it seems.

[Fungus]

My understanding is the PD spec still supports the 12V profile as optional. So charger manufacturers can choose to implement it or not. The minimum that must be supported are 5V/9V/15V/20V to be considered PD compliant, with PD 3.1 adding support for 24V/36V/48V.

Yes, it's obvious why it's 15V nominal.

I don't think there's anything on the PCB that needs anywhere near that voltage. They must be regulating it down as a first step.

We know 12V works, so what's the limit? Somebody needs to hook up an oscilloscope to the internal rails and gradually bring down the input voltage to see where it fails.

nb. The current will probably go up as well so we need to look at component temperatures in the power area as the voltage drops. It's not an easy test to do - maybe one for Dave?

It might even be that it's happier with 16V than 15V, but 15V is the closest widely supported voltage in the PD spec.

PS: How many amps will that input connector take?

[dreamcat4]

What i'm actually interested in would be some step-down switching regulator module. That took an input up to 22v and then brought it down to a reliable and stable +15v (or like whatever you say Fungus, maybe 16v). However the thing of it is that Makita LXT 18v battery packs start at about 22v fully charged state. And then the voltage degrades to about 14-15v when they are almost empty. Or at 10% charge state whatever.

So this fits well with the nominal input voltage of 12-15v for this scope. However the other thing that this module, or strap around circuitry needs to do is A/B power switching failover. So that when battery pack A dies, and falls below threshold voltage. Then the circuit will seamlessly switch over to battery pack B. (but only if present, and detected as power good input rail).

Now switching over function might not be as important for others. Or (for example) powering from a vehicle +12v. However it would be quite nice feature I believe, and mostly dues to the high idle power consumption of 35w. Which is otherwise probably not needed.

[metrologist]

There are plenty of USB-C 15V power packs and automotive supplies designed for 15V laptops.

https://www.ebay.com/itm/166056322833

[ebastler]

to test the scope actual performance characteristics across a variety of common and different types of power sources. To make sure that a poor quality of power does not adversely affect or detriment scope measurements.

I think this is a relevant point. I am wondering why Rigol switched from the fixed 12V supply to the proper, PD-enabled Lenovo supply -- which must cost them extra money.

The 12V supply was apparently not planned as a stop-gap solution for early shipments only. It was meant to be the regular power supply shipped with the scope: It is described in the manual, and the scope's backpanel has a special hole next to the USB-C connector which accepts a little coding pin molded onto the cheapo 12V supply's "USB-C" plug. (To prevent users from killing other USB-C devices with that pseudo-USB-C power supply.)

So why did they swap it for the Lenovo supply? I'd guess that either the 12V thing did not meet international electrical safety standards, hence was not suitable for export. Or its output voltage has such nasty spikes that the scope's S/N suffers -- in which case some other power supplies might create problems too. I'd be curious to learn the reason for the late power supply swap!

[2N3055]

to test the scope actual performance characteristics across a variety of common and different types of power sources. To make sure that a poor quality of power does not adversely affect or detriment scope measurements.

I think this is a relevant point. I am wondering why Rigol switched from the fixed 12V supply to the proper, PD-enabled Lenovo supply -- which must cost them extra money.

The 12V supply was apparently not planned as a stop-gap solution for early shipments only. It was meant to be the regular power supply shipped with the scope: It is described in the manual, and the scope's backpanel has a special hole next to the USB-C connector which accepts a little coding pin molded onto the cheapo 12V supply's "USB-C" plug. (To prevent users from killing other USB-C devices with that pseudo-USB-C power supply.)

So why did they swap it for the Lenovo supply? I'd guess that either the 12V thing did not meet international electrical safety standards, hence was not suitable for export. Or its output voltage has such nasty spikes that the scope's S/N suffers -- in which case some other power supplies might create problems too. I'd be curious to learn the reason for the late power supply swap!

There were some reports that scope had visible noise in traces and PSU was simply spraying enough EMC (radiated and conducted) that it affected measurements. There was few videos about that. EU EMC compliance could have been the factor too but that PSU was too cheap and degraded otherwise good scope performance.

So Rigol, being honest company, had to do something. And they did. They might use Lenovo as intermediate measure until they source/make cheaper alternative that is of adequate quality.. Anyway, classy move on their side..

[dreamcat4]

and can i just say: that this was not the 1st time i have seen a product coming from china with a +12v fixed (2-wire) over usb-c without any usb-compliant power negotiation. There was also some raspberry-pi type of SBC equivalent (or other such boxes, like android tv box) that supplied such solution. Perhaps 1 year ago. And it's a thoroughly abhorrent practice (in terms of damaging people's other devices), i would say even with the safety pin.

OTOH i do value and absolutely appreciate the general capability of supplying a simple fixed voltage input around 12-15v level. And whether via a usb-c connector or not. But what else would you prefer to use? I looked about, and it's like... well theres things like XT-60 which seems a bit large and bulky. So I got some XT-30 which looks about right size and quite nice. However you try getting an XT-30 with a locking clip and proper panel mount male and female versions (that have screw holes etc...) And you will be looking for a very long time. Or they will cost too much. Or they will be difficult to order economically etc.

So then you look for other dc power connectors that are locking. With some sorts of locking clip. And this is where GX-* aviation connectors are just... less convenient (and more physically bulky). Because those retaining collars screw on rather than click. So that's not as nice. And while mini XLR does click and lock nicely, OTOH it's also similarly bulky like the aviation connector. And also not specc'd properly for DC power like the XT-30 is.

So I am still scratching my head a bit here, if not usb-c, and not XT-30, then for a simple DC power connector that is actually nice to use. And you know rugged / durable / cannot be damaged so easily as some barrel jack. And does not have polarity ambiguity issue.

[duckduck]

Just got an email from Rigol NA that they are shipping my DHO804. I wasn't expecting it so soon. I'm looking forward to it because I sold my old scope already and I've got (hobby) work stacking up on the bench.

I previously owned a Rigol DS1202Z-E (200Mhz, 2-channel). The DS1202Z-E had a sluggish UI and the screen update rate in XY mode was slower than a holiday picture slide show. I'm hoping these are not issues in the DHO8xx series. I plan on backing up the flash and upgrading the unit first thing. I would not have bought this unit if it were not for the hack.

[Fungus]

The DS1202Z-E screen update rate in XY mode was slower than a holiday picture slide show.

That's usually the operator's fault for picking the wrong memory size.

[Warhawk]

The DS1202Z-E screen update rate in XY mode was slower than a holiday picture slide show.

That's usually the operator's fault for picking the wrong memory size.

You could work for Apple. Any problem is always customer's fault

[ebastler]

For use with the DHO900 (and hopefully the DHO800 after a few mods?) I've created a new v3.1 of the 16-channel LA clone board that is cheaper to make and easier to hand-solder:
https://climbers.net/sbc/clone-pla2216-logic-probe-analyzer/

Many thanks for that! I have a loosely related question, but relevant to a DIY logic probe:

While looking at the PCB and at a picture of the DHO900 front, it seems that the connector on the scope's front has a different shroud. The notch which keys the plug direction seems to be off-center on the DHO900? Is there an off-the-shelf IDC connector which is styled that way, or do people use a regular IDC connector and file off the centered tab? (Or did I simply look crosseyed?  )

EDIT: Hmm, people clearly seem to use regular IDC connectors with the MSO5000. But the notch in this photo does look off-center, doesn't it? (Aligned with pins 13 and 14 in a row of 25 total.) https://www.batronix.com/shop/oscilloscopes/Rigol-DHO914S.html

EDIT²: Ok, it seems to be a regular connector with the notch in the center. Must be just a weird distortion in the photo; so this is a non-issue.