OWON VDS1022I Quick Teardown (versus the Hantek 6022be)

[indman]

How can I monitor battery charge/discharge over 24 hours?

And why do you need such a resolution and high accuracy when measuring battery charge /discharge?
For example, I use cheap ImaxB6 and LogView for these purposes 

[DAIRVINE]

How can I monitor battery charge/discharge over 24 hours?

And why do you need such a resolution and high accuracy when measuring battery charge /discharge?
For example, I use cheap ImaxB6 and LogView for these purposes 

I don't actually need to log battery voltage at the moment.  Though I would like the capability for off grid use (someone I know).

For VDS1022I it is more a case of what can we get out of it. And I may have a go at it if motivated.

[DX1]

I did some tests to determine what sort of results we might get with averaging to improve resolution
Using an undercompensated scope probe I captured the peak of the 1kHz output at 10us per division.
I uploaded 4 images:
Sample is the CSV saved sample fed into libre office and charted.
Rolling average  is the above sample with 16 elements averaged in libre office.
Rolling 64 is the above sample with 64 elements averaged together in libre office.
Scope average is the scope average using 16 elements of the same signal.

Clearly we are trading bandwidth and resolution.

The scope average function does not improve resolution, but keeps it at 8 bits. This could be improved.

Is anyone interested?

I assume Owon are not going to let us near the FPGA and firmware?
BTW I changed my account because I messed up changing to new email.

[DX1]

I made another test.
Scope average is a repeat use of an earlier capture using scope average function with 16 elements.
Average 16samples was made by capturing 16 csvs in sample mode as csv and averaging them in libre calc.

It does, I think, show that the average function in software could be improved by using more bits.
There are a number of ifs and buts. Still it could still be useful.

Additional improvements are possible, for example detecting and displaying glitches outside of normal noise range.
If using rolling average, detecting and displaying changes outside the noise range.

[DX1]

Of course, to go really mad we can do both.
16 samples averaged, then 64 elements averaged as rolling samples.

[DX1]

Can anyone please give me some guidance on downloading the latest code into eclipse and debugging it.
I am familiar with eclipse, but not with java.

[jasonRF]

Adrian's digital basement has a review of this scope


The video triggering is clearly quite useful for the work he does, and really seemed to work.  He hated the interface, and would have benefitted from spending a couple of extra minutes learning how to use it (especially keyboard shortcuts for horizontal/vertical scale adjustment) before diving in.  Am interested in seeing the follow-up review he promised that will use florentbr's software. 

jason

[themadhippy]

and would have benefitted from spending a couple of extra minutes learning how to use it (especially keyboard shortcuts for horizontal/vertical scale adjustment) before diving in

I class that as a real world test,how many of us get something new and put it to one side whilst we read the manual from cover to cover ,compared to  plugging  it in and start stabbing random buttons?

[Gyro]

and would have benefitted from spending a couple of extra minutes learning how to use it (especially keyboard shortcuts for horizontal/vertical scale adjustment) before diving in

I class that as a real world test,how many of us get something new and put it to one side whilst we read the manual from cover to cover ,compared to  plugging  it in and start stabbing random buttons?

I have to say I have mixed feeling on this. Out of the box reviews are good, and this one particularly as he was actually using it on live circutry to do something useful, but it was kind of embarassing to watch him blundering around getting the basics of the UI (particularly the timebase). After 15 minutes he was getting there with just random stabbing, he would have got there in another few minutes (trigger edge, channel and measurement shortcuts etc). He hadn't even got as far as the keyboard shortcuts on even the stock s/w.

He did get the capabilities pretty well tied down though.

[jasonRF]

and would have benefitted from spending a couple of extra minutes learning how to use it (especially keyboard shortcuts for horizontal/vertical scale adjustment) before diving in

I class that as a real world test,how many of us get something new and put it to one side whilst we read the manual from cover to cover ,compared to  plugging  it in and start stabbing random buttons?

Fair enough.   I guess I have used picoscopes for so long that one of the first things I did when getting my rebadged vds1022 was learn the basic keyboard shortcuts.   It literally took one minute, and there is a tab on the “tips” screen that pops up when you start the software so they are not exactly hard to find. Changing timebase or vertical scales with a menu/mouse is just too annoying to me.   

[Gyro]

One thing that I forgot to mention last night - The isolated version choice early in the video, where he said that it would be useful for live chassis working! Granted, there is going to be a lot of exposed live metal in the DUT, but that's not a reason to add more to it by making the case and other exposed connections of the VDS live too!   Granted, the USB isolation would still protect the PC but that's not the point.

Again, this comes from the 'out of the box' style review, but I would have expected him to do some basic research into safe use before suggesting it in a video review. Floating mains use is something that has come up a few times in the thread and each time it has to be explained why it wouldn't be a safe practice. I don't know if anyone has put a link to this thread in the comments yet (I'm not registered).


On the subject of the scroll wheel use. I'm not sure if it would be useful to have some default action for what happens if you scroll it while the pointer just generally hovering over the screen. It sounds as if some other USB scopes (and Saleae) change the timebase in this case, however it might be equally appropriate for it to change the vertical sensitivity, trace vertical position, trugger level, buffer position? etc. Hence the requirement to hover over the parameter in question when scrolling. I'm not sure if a general default 'general screen area' action would make it more intuitive? One for Florent I guess.

[rsjsouza]

and would have benefitted from spending a couple of extra minutes learning how to use it (especially keyboard shortcuts for horizontal/vertical scale adjustment) before diving in

I class that as a real world test,how many of us get something new and put it to one side whilst we read the manual from cover to cover ,compared to  plugging  it in and start stabbing random buttons?

I have to say I have mixed feeling on this. Out of the box reviews are good, and this one particularly as he was actually using it on live circutry to do something useful, but it was kind of embarassing to watch him blundering around getting the basics of the UI (particularly the timebase). After 15 minutes he was getting there with just random stabbing, he would have got there in another few minutes (trigger edge, channel and measurement shortcuts etc). He hadn't even got as far as the keyboard shortcuts on even the stock s/w.

He did get the capabilities pretty well tied down though.

The out of box experience is important to show people what to expect when purchasing the product and eventually catch a few pitfalls. However, when it enters the realm of "first poking at the equipment", editing and corrections (by means of captions, for example) is always warranted especially if you dislike or can't understand something as basic as moving the waveform or getting the timebase correct. What I think it happened is that he was initially miffed by the interface and this colored all the further explorations into it - especially the expectation that the entire interface should have worked with the mouse only.

One of the aspects that I disagree is the on screen waveform dragging, as it can be hard to figure out the pick point on the screen that will drag just one channel instead of the other - especially when waveforms overlap. I have seen UIs be extremely confusing on this as the number of displayed channels increase and it is not uncommon to find yourself dragging the wrong thing on the screen. 

On the subject of the scroll wheel use. I'm not sure if it would be useful to have some default action for what happens if you scroll it while the pointer just generally hovering over the screen. It sounds as if some other USB scopes (and Saleae) change the timebase in this case, however it might be equally appropriate for it to change the vertical sensitivity, trace vertical position, trugger level, buffer position? etc. Hence the requirement to hover over the parameter in question when scrolling. I'm not sure if a general default 'general screen area' action would make it more intuitive? One for Florent I guess.

Florent's software uses the scroll wheel to change the timebase while the pointer is on the screen. However, I would love to see it take a different action if Ctrl or Alt were depressed. The problem of changing other settings with the scroll usually boils down to which channel will this be applied. You can make the channel active by clicking on its box, but by then you are already closer to the V/div settings anyways (which also works with the scroll button).

Anyways, the video shows the first impressions but I feel that corrections on post editing would have improved it.

[jasonRF]

One feature of florentbr's software that I haven't seen other posts about are the lowpass filters he added to the channel setup menus.  This week I had a practical situation where they were quite useful.  I was trying to measure the output of a class-D audio amplifier; my particular model has differential outputs and the signals look nothing like the pretty traces I am used to seeing from class-A and class-AB amps.   EDIT: for those that ae interested, the amp is an SMSL AO200, which is based on the Infineon MA12070 chips that run without an output filter. 

I have attached a few images of measurements using the scope, where the two channels are making single-ended measurements of the two outputs for the left channel while playing a single 1 kHz sine-wave.  The raw signals are messy, as is the output of the mathematical difference between the raw channels.  You can see the sine-wave, but it has a ton of high-frequency content:








However, with the lowpass filters set at 20.5 kHz, both the signals and their difference are quite clean:







I think it is interesting that florentbr setup the filters as a percentage, so the cutoff frequency will change when a timebase change modifies the sample rate.  For comparison, the filtering options for my Picoscope are set by cutoff frequency.   Either way is fine, and while I do prefer the way the Picoscope software works, it is mostly a matter of taste.  I am just happy that florentbr added these, since it can save the hassle of cobbling together hardware filters on a breadboard just to make some quick-and-dirty measurements. 

This is yet another way that florentbr made these scopes more valuable. 

jason
 

[Gyro]

One feature of florentbr's software that I haven't seen other posts about are the lowpass filters he added to the channel setup menus... 

Damn, That confused me, I was wondering why I couldn't find it! It turns out that I while I have CF19 installed, I was using an old desktop shortcut, that's what happens when you have multiple versions installed in different locations - I must have deleted the wrong one.  I must have been running it at some point because, according to changes.txt, it was introduced in CF15 and I tested the XY / FFT mode cycle button in CF18.

Thanks for pointing it out.


EDIT: That's also why I wasn't seeing the scroll wheel timebase that rsjsouza pointed out the other day.

[trungdt90]

Hi everyone, I was looking for an oscilloscope for my DIY audio, after researching and thanks to this thread, I bought a non (i) vds1022
everything works fine, however when I did the noise test of the oscilloscope itself, I see it to be not very stable, despite using an external power source (lt1764 low noise LDO)
Will adding a power supply isolation chip affect this?

[Gyro]

Hi,

If I'm reading the trace correctly, at 10ms/div, that looks like mains pickup. It could either be from a ground loop between your PC and audio source (which USB isolation in the 'I' version is there to break), or noise pick up from your source. It's impossible to say without knowing rather more about the source and your setup. If the mains noise disappears when you unplug the probes (which I'm sure it will unless the scope is faulty) then it is an issue with your test setup.

I'm not exactly sure what you mean by "power supply isolation chip", are you talking about turning the scope into the 'I' variant or your external power source. To cure a ground loop issue, you need to break the loop which is passing between the grounds of your source, the PC ground, and possibly your PSU ground.


EDIT: On second reading, I am now not sure what you have your probes connected to. If they are just lying on the bench, they will be picking up low level mains noise from the local environment, in this case, shorting their tips to their ground clips should make it go away.

[trungdt90]

Hi,

If I'm reading the trace correctly, at 10ms/div, that looks like mains pickup. It could either be from a ground loop between your PC and audio source (which USB isolation in the 'I' version is there to break), or noise pick up from your source. It's impossible to say without knowing rather more about the source and your setup. If the mains noise disappears when you unplug the probes (which I'm sure it will unless the scope is faulty) then it is an issue with your test setup.

I'm not exactly sure what you mean by "power supply isolation chip", are you talking about turning the scope into the 'I' variant or your external power source. To cure a ground loop issue, you need to break the loop which is passing between the grounds of your source, the PC ground, and possibly your PSU ground.


EDIT: On second reading, I am now not sure what you have your probes connected to. If they are just lying on the bench, they will be picking up low level mains noise from the local environment, in this case, shorting their tips to their ground clips should make it go away.

Thank for reply, well, I didn't plug in the probes when do this test. And yes, the "power supply isolation chip" I'm talking about is the DCP020505U at "I" version .

[Gyro]

A couple of things that come to mind. You say that you have a low noise external power supply feeding the scope, If you are feeding this in via the USB 'Y' cable that comes with the scope then your supply is effectively in parallel with the 5V USB supply comming from the PC USB port (It takes power from both connectors).

No, just fitting the internal isolated DC-DC converter won't break any ground loop. To be isolated like the 'I' version you would also need to fit the ADuM3160 USB data isolation chip and remove the zero ohm resistors bridging the ground planes (not that you can't remove these without fitting the ADuM3160).

Try attaching the scope probes and grounding the tips with their ground clips. It's possible that the high impedance inputs are stil picking up a small amount of mains hum if your environment is particularly noisy.

P.S. Before ordering any parts, check the footprint of the DC-DC converter. From a recent post, it looks as if they have changed from the DCP020505U to a more common Chinese DC-DC module. Before opening also consider whether it is feasible to return it in exchange for an isolated model before modifying it yourself. I'm also not sure which would be the cheaper option in your local market conditions.

[trungdt90]

A couple of things that come to mind. You say that you have a low noise external power supply feeding the scope, If you are feeding this in via the USB 'Y' cable that comes with the scope then your supply is effectively in parallel with the 5V USB supply comming from the PC USB port (It takes power from both connectors).

No, just fitting the internal isolated DC-DC converter won't break any ground loop. To be isolated like the 'I' version you would also need to fit the ADuM3160 USB data isolation chip and remove the zero ohm resistors bridging the ground planes (not that you can't remove these without fitting the ADuM3160).

Try attaching the scope probes and grounding the tips with their ground clips. It's possible that the high impedance inputs are stil picking up a small amount of mains hum if your environment is particularly noisy.

P.S. Before ordering any parts, check the footprint of the DC-DC converter. From a recent post, it looks as if they have changed from the DCP020505U to a more common Chinese DC-DC module. Before opening also consider whether it is feasible to return it in exchange for an isolated model before modifying it yourself. I'm also not sure which would be the cheaper option in your local market conditions.

I break 5V+ from usb, connect it to low noise external power, not parallel 2 power source.
Follow datasheet of DCP020505U, i see it breaks both V+ and GND, also checked the VDS board, 4 big 0ohm resistors (2 on top, 2 on bottom) are correct input and output with DCP020505U. I had order 2 isolator, replace it soon and check again

[Gyro]

Ah, I see, that makes sense. All of the internal analogue and digital supply rails are separately regulated (apart from the probe cal output) so it shouldn't be very sensitive to input supply noise, certainly not at mains frequency.

Yes, the large 0R resistors that isolate the ground planes and +5V are fairly obvious. A quick check with a meter will show that you have achieved isolation. The resistors that bridge the USB data signal isolation are on top of the ADuM3160 footprint, so you will automatically have to remove them to fit the IC (again don't attempt ground isolation before you have fitted the ADuM3160 as this would risk injecting damaging voltages into the USB signals).

My unit is now 7+ years old, so many revisions before yours, but you can see the noise levels on mine (PC powered) in replies #2 and #3 for comparison. Nobody else seems to have reported high levels of mains frequency noise pickup, maybe somebody else with a recent revision unit can post some screenshots for you to compare. Again I would check with shorted inputs in case you have a particularly noisy environment where mains hum is getting into the open BNC connectors - and maybe also try powered by the normal USB 5V.

[jasonRF]

My unit is now 7+ years old, so many revisions before yours, but you can see the noise levels on mine (PC powered) in replies #2 and #3 for comparison. Nobody else seems to have reported high levels of mains frequency noise pickup, maybe somebody else with a recent revision unit can post some screenshots for you to compare. Again I would check with shorted inputs in case you have a particularly noisy environment where mains hum is getting into the open BNC connectors - and maybe also try powered by the normal USB 5V.

Here is what the noise looks like on my MP720016 (rebranded vds1022) with no probes on; it looks the same with shorted probes attached.   It also does not matter whether my laptop is running from mains or battery.  I purchased the scope from Newark within the past two months, and the PCB is labeled VDS1022_V2.1 with a date code 2021.07.12. 

jason

[trungdt90]

I have soldered both DCP020505U and ADUM3160 isolator ic and put it back in the aluminum case but the main noise still doesn't go away.
Then I dissemble it and found that when touching the gnd part on the motherboard the main noise disappears, but leaving it alone the main noise reappears.
Can anyone give a hint on this issue?

[Gyro]

That's a curious one. In a noisy envronment, maybe touching the ground plane injects some common mode noise into it which compensate the signal noise pickup (?).

Some things to check...

1. Continuity checks that USB ground and +5V and VDS are now fully isolated.

2. That the internal screening cans are properly grounded to the ground plane.

3. I suspect most likely - That the Aluminium case is making good ground connection with the ground plane. I think on later PCB revs there are bare tinned areas running down each side of the PCB to make contact with the case grooves. It could be that you have a case which is extra-heavily anodised and isn't making contact. Check against the BNC sleeve or the gnd tab next to the probe cal output. I would expect touching the case (or indeed the probe ground clips) to have the same effect as touching the PCB.

[Bigwizman]

New member here and a huge amount of appreciation to all contributors for so many interesting comments over the current 31 pages, it must be quite a surprise to find so many following for so long Gyro?  The level of expertise shown by so many, is quite reassuring and having worked in electronics for my whole career, although now retired, it shows how much dedication so many have for the subject.  I've recently been playing with micro-controllers to improve certain features of my old property and add some smart systems.  My old oscilloscopes have seen better days, both sporting Advance logo's but one after Gould took over, although both names were added to the front panel. One is analogue and the other the 1st digital scope Gould made, being the OS4000, sampling at 1.8MHz.  They claim it manages to display a bandwidth of just over 400kHz, although has a 10MHz front end for the analogue display.  They all used to be mixed signal 'scopes during that phase.  They have both served me well but I was struggling to verify some of the new hardware I'm using, as well as having several faults that appear every time I actually need them, so decided to get myself a cheap small unit as I'm trying to stop filling the house even more than it already is!  I visited many sites but this was the one that gave me the best insight, although I'd seen it on the digital basement review 1st, so both areas had my respect.
I only received it yesterday but it is indeed impressive for the price and does do what the makers claim, not a normal experience for most of the far eastern imports, who shall remain nameless.
I started with florent's SW, which I am also appreciative of, it is an impressive feat, although I know that many have helped contribute to the project.  I spent an afternoon running around the various features and got excited as each had such capability beyond something at this price point.  Today, I coupled up one of my experiments and decided an external trigger would be useful.  I noted the multi channel automatically switched to "external in" and being TTL, only edge trigger is available, both what I needed.  It was the 1st item that I'd tried that didn't go as planned.  No triggering at all and to cut a long story to the chase, I found that the multi channel appears to load the circuit, as though it is still switched as an output (I'd used it during my familiarisation to see the trigger output delay), as when attached to the probe calibrator, the signal no longer swings to zero, being around a 2.5v to 5v square wave instead.  I tried pre-switching it to an input, prior to selecting it as a trigger source, as well as several other combinations.  It isn't a show stopper and didn't curb my admiration for all that has been done but perhaps somebody might confirm my findings and make sure it isn't something peculiar to my unit or circumstances.  My unit is the VDS1022I and I have checked all the grounding arrangements.  Removing the ext connector returns the probe output back to normal.
I will end there, as it is already long for a 1st comment but I'm glad to be part of this community and a rather pleased owner of the device too!

[Bigwizman]

While I'm still thinking about it, I didn't seem able to enable the channel filters, although I used averaging instead to achieve a better image on repetitive waveforms.  Some of the single shot traces would have benefitted from some bandwidth limiting (despite the lower apparent sampling rate), as I was at very slow sweep rates and in roll mode.  However, what a blessing having triggering during roll, I'd have been lost without it, as it is one of the best features of my old Gould OS4000, along with 4 trigger points to serve either pre or post trigger data.  The features on this SW (cf19) are making me smile every time I capture a new trace.

To add a little background, I put my controllers to sleep for efficiency, so set a single trigger, waiting for the wake-up. Although most of the signals are quite fast, bursts of data appear at various points afterwards, so a slow'ish sweep eases the visualisation of all the relationships.  The peak detect allows the odd single pulses to be detected and yet another smile elicited.  I assume that is possible because the ADCs are always flat out, which I'm sure I read here too.

Thanks again.