OWON VDS1022I Quick Teardown (versus the Hantek 6022be)

[Gyro]

Yes, it threw me for a while too. You need to tick X-Y mode in the 'Display' menu, and then click on the single/split screen icon (the one to the left of the home button) on the main screen.

I'm not sure if Florent intended it this way - it could be more intuitive. 

[CheckOne Two]

It was me that requested the full screen XY mode,

I use a curve tracer (octopus) for troubleshooting difficult faults where nothing else is showing up anything significant and pcb comparison sometimes using the VDS on a 15" laptop. So much better on the VDS than using it on a standard analog scope ... e.g. probe first connection, hit spacebar (this freezes the image) .... move probe to the connection to be compared, hit spacebar ... and you get an instant AB comparison.

With getting old now the original tiny XY screen on the 3x screen display was difficult to see from where I sit the laptop ... after a few beers one night I requested the full screen version with no idea if Florent would implement it...

I've just installed it ... won't be able to try for a while yet so not seen it work. Yeah it was meant to be switched by the Single/To Three view button which IMO is intuitive.

I think this has also given a new full screen width FFT + normal scope screen too instead of 3 screen display in FFT mode ... switched by same button ... although I can't remember now if it was like that originally...

Here's my original request, this will explain why the "To Three view" button was used the way it was:

Is it possible to get XY mode in full screen like the FFT screen? Maybe I just haven't found how to do it?

When FFT is enabled it can be toggled to full screen with the "to three view/to one view" button above "trigger" (bottom right).

I use XY mode a lot sometimes when troubleshooting & it would help me if I could full screen view the XY screen to see it better from where I sit, is that possible?

[Gyro]

It was a good request, I suspect that several of us wanted full screen X-Y but didn't think that Florent would want to / be able to tweak the screen to that extent. Yes, it looks as if we got full screen FFT for free too!

Having to go to menu items to select either X-Y or FFT doesn't seem too much of a burden considering that you are going to be selecting those modes for a specific purpose anyway... and by default, if you don't select either, you still get the split screen scope + X-Y if you hit the 'Single to Three view' button.

The only improvement I can think of is to make the 'Single to Three view' button operate three ways (in keeping with its title), Full screen scope -> Full screen X-Y -> Split screen scope + X-Y, with similar behaviour if FFT is selected instead.

It would still require Menu action to enable FFT, but that is probably the less used function on an bit scope, albeit much more useful now it has a decent display size!


Edit: Full screen FFT confirmed working. 


P.S. I've posted an issue on Florent's github page, if you have an alternative view to what I suggested, please add a comment.

[CheckOne Two]

The only improvement I can think of is to make the 'Single to Three view' button operate three ways (in keeping with its title), Full screen scope -> Full screen X-Y -> Split screen scope + X-Y, with similar behaviour if FFT is selected instead.

Yeah that's why I initially asked if I maybe hadn't found out how to get the 3rd vlew LOL ... yes the software is getting good now I've sent him some euro's in appreciation ...

P.S. I've posted an issue on Florent's github page, if you have an alternative view to what I suggested, please add a comment.

I see he's already "on the case" ... agree with your ideas so I've added my interest in further view development too. It's funny if I hadn't had a few beers initially I would never have asked in the first place LOL ...

[Gyro]

Just a note that Florent has updated his S/W version (OWON-VDS1022-1.1.3-cf18) to implement the Full screen scope / Split screen X-Y + Scope / Full screen 3 way button as discussed.

Thanks Florent!

[indman]

Hi!
Is it possible to implement Bode Plot for this oscilloscope? That would be really cool.

[Anthocyanina]

Hi!
Is it possible to implement Bode Plot for this oscilloscope? That would be really cool.

This oscilloscope doesn't have a function generator, so very very unlikely (someone would have to write quite a lot, to read the input frequency and amplitude in both channels from the measurements then automate timebase and voltage range adjustments, from reading those measurements) I think possible but not worth trying to do it

[indman]

This oscilloscope doesn't have a function generator

You can use an external generator - it's easy. DSO5102 - has a Bode Plot option, albeit in a simplified form.

[Gyro]

Maybe the trigger (multi) BNC could be used to trigger/synchronise the external sweep generator.

[indman]

Gyro,yes it can be used as in this example with DSO5102

But how correctly and logically answered me florent:
"The VDS1022 has the same fake Bode Plot capabilities as the one from your link.
It's called roll mode or slow mode. I wouldn't call it a Bode plot given that it doesn't have a log scale.
It's also limited to the current time-range and buffer size."
I agree with his opinion.

[Pihkal82]

Hi guys (and possibly gals),

I've been reading through this thread and I'm blown away by all the great work that has been done here, many kudos to all of you!
I'm thinking about purchasing the Owon VDS1022I, it looks perfect for an occasional tinkerer like myself.
To not risk blowing up an expensive laptop, like someone else in this thread did, I would like to pair it with a spare Raspberry Pi 3b+ and an ancient 15" 1024x768 VGA monitor.
So far I've set everything up on the Pi side, including Florentbr's custom software, it all seems to be running rather well but since I don't own the Owon yet I can't really test how this software performs on the Pi.
I know people in this thread have been talking about running it on a Pi but did any of you test it and were there trade offs compared to pairing the Owon to a decently specced laptop?
Many thanks in advance!

[Gyro]

Hi, welcome to the forum.

The VDS1022(I) is very modest in it's USB requirements, it transfers data at USB full speed (12Mbps) rather than USB high speed (480Mbps) so it isn't going to put any particular strain on the Pi, either data speed or processing. I haven't run it on a Pi myself, but if you already have Florent's S/W working then I can't see a problem.

With the galvanically isolated VDS1022I, it is really very difficult to blow up a laptop via its USB port (that's its purpose). The non-isolated plain VDS1022 is a completely different matter, you can inflict damage with this as easily as you can with most other PC hosted instruments (logic analysers, Hantek, Picoscope, Analog Discovery, MicroVNA etc).


P.S. I think many of us may be interested in hearing about your results, so do please report back. 

[Pihkal82]

Hi Gyro,

thanks for the welcome and the reply.
My main concern with the Pi would be the actual computer side processing, but to my understanding most of it happens in the scope, unlike the Hantek 6022be (which was my first choice until I stumbled upon this thread) which does most of its stuff in software.
The software does run but  it can still fail on the usb connection, but we'll find out....
As for the "blowing stuff up', I've ordered the isolated version, that should give me some peace of mind if I need to hook it up to a laptop.
In the mean time I've also installed PulseView/sigrok on the Pi as I'm also getting a cheap logic analyzer, but that's beyond the scope (no pun intended ) of this thread. 
 
Anyway, I've just placed my order on Amazon, expected delivery is the 17th of August.
I'll report my findings on how it works with the Pi here.
Thanks!

[Pihkal82]

Hi all,

I've received my OWON VDS1022I today, a day earlier than expected and did some testing on the Raspberry Pi 3.
For starters let me first correct an error I made in my previous post, it's a regular Raspberry Pi 3b and NOT a 3b+ .
Let's move along to the good stuff now...

Setup:

 Platform: Raspberry Pi 3b 1.2Ghz / 1GB (no overclock)
 PSU: LiteOn 5.2V / 2A -> replaced by some Chinesium PSU 5V / 2A to combat the undervoltage warnings
 Powered USB hub: Dlink DUB-H4 (5V / 2.5A)
 OS: DietPi 8.7 (Debian BullsEye 32bit) + LXDE desktop environment
 Scope: OWON VDS1022I
 Software: Florentbr OWON VDS1022 1.1.3

Findings:

 It's kind of plug&play, connect the scope to the Pi's onboard USB ports, start the software and it works BUT....
 The above strongly depends on the PSU that's powering the Pi, the first PSU I used caused undervoltage issues, I needed to use a powered USB hub with it.
 However, the second PSU I used resolved the undervoltage issue as well as omitting the need for the powered hub.

 Next let's talk about performance...
 With the Pi at stock clocks it looks like it's performing ok, however since I haven't tested it on a decent laptop yet, I haven't gotten any point of reference on how it should perform. (number of trace updates per second)
 It does look like it's overloading a single core, htop shows the OWON thread as being at 123%, this will most likely have a negative impact on the update rate.
 To counter this I've overclocked my Pi's CPU to 1.4 Ghz, as you can imagine it generates quite some more heat now, to handle this extra heat I've added a heatsink to the SoC, it still gets rather toasty (60°C) while using the scope software, but since the final version will be cooled with a 40mm fan this won't be an
 issue.
 
 As for the scope itself, I've only tested it with the build in 1 Khz square wave signal and this is working really well.
 The noise on the trace is acceptable and I'm happy with my purchase.
 I paid 117€ for the scope (isolated version) on Amazon with free shipping, a pretty good deal if you ask me.
 The rest of the setup is all stuff that was lying around here collecting dust, I finally was able to upcycle it and put it to good use.   
 
TLDR: You can pair this scope with a Raspberry Pi, it appears to work pretty well but will most likely not perform as good as it would on a decent specced laptop.

If you have any questions regarding the subject please fire away!

Kind regards
 
 

[Gyro]

A good writeup Pihkal. Thanks very much.

[jasonRF]

Hi!
Is it possible to implement Bode Plot for this oscilloscope? That would be really cool.

This oscilloscope doesn't have a function generator, so very very unlikely (someone would have to write quite a lot, to read the input frequency and amplitude in both channels from the measurements then automate timebase and voltage range adjustments, from reading those measurements) I think possible but not worth trying to do it

I don't have one of these scopes, but am thinking about getting an isolated one for 'riskier' measurements to make sure I don't damage my PC or my more expensive Picoscope.  The simplest way I can think of to automate Bode plots with one is to write a Python script to control things.  This came to mind as I am currently playing with a very simple Arduino-controlled AD9833-based function generator project, where I send the desired generator frequency to the Arduino over the COM port using a Python script.   This is not an original idea of course. 

For the measurement part of the Bode plot, it looks like the Python API that florentbr built has an autoset function.  If that autoset works well enough, then florentbr did all of the hard work for us!  It actually only needs to get the amplitude settings okay, since we would know the frequency already so could do a final time-base adjustment if needed (eg, if the signal is aliased to ensure the signal is not aliased).  If the Owon works like my Picoscopes, having the wrong time-base that creates an aliased signal actually doesn't impact the amplitude measurement much, if at all, so it should work out fine. 

Of course, the physical setup for Bode plots would require the function generator output and one scope channel on the input of the DUT, and the other scope channel on the output of the DUT.  Automating the Bode plot would then amount to looping over the frequencies of interest.  For each frequency
1. set the function generator frequency
2. run the autoset function on the scope. Perhaps adjust the time-base based on the known frequency.
3. measure the amplitudes on each scope channel and the phase difference (or time-delay) between them.

After looping over the frequencies, the plots could then be made in Python pretty easily.  There are multiple ways to do step 3, of course.  Without having one of these scopes I wouldn't pretend to know how exactly, but surely the API, in conjunction with other Python functions, could be used. 

Anyway, I just thought I would put that idea out there in case someone cares enough to try and make it work. 

I suspect many folks around here could do the coding better than I can, but I would be happy to share my super-simple Arduino and Python code for the signal generator if folks are interested in trying it out.  This is my first ever project using either a microcontroller or Python, and since I just started yesterday it is a rough work-in-progress not worthy of posting as an actual 'project'.  The analog output I cobbled together with parts in my bin is okay, but still on a breadboard and powered using a bench supply. 

cheers!

jason

[negative_feedback]

I'm trying to read sine wave of AVR generator, mains and DC-AC inverter by using barebone 10x probe (9Mohm+1Mohm internal afaik). Can I make this graph "fit" or I have to use step-down transformer? 

Attached image is for single phase 230VAC/50Hz readout on grounded desktop PC.

PS Please don't suggest buying hardware (DP10013 in this case) that costs more then an oscilloscope at hand...

[indman]

don't suggest buying hardware (DP10013 in this case) that costs more then an oscilloscope at hand...

DP10013 is not necessary to buy, but a 100:1 probe will be just what you need.

[negative_feedback]

Is there half-decent software way of doing this, change the proble in settings to 100x and such ?

[indman]

https://how-to.fandom.com/wiki/How_to_make_a_100X_oscilloscope_probe

Building a 1000x probe will give you a lot of useful information that is relevant for a 100x probe as well 

[adam4521]

No you are saturating the hardware — the analogue front end and/or the ADC, and you need to scale it down. No reasonable software fix for this. But placing your own potential divider across this particular source to scale down for your existing probe is easy.

On the safety point, if you are relying on the USB isolation remember that the exposed part of the BNC connector will be connected to your source and even if is is a ‘neutral’ connection is considered to be a live conductor, and you should treat that and this whole scenario with some caution, due to the potential fault currents and shock risks. If you have a low voltage transformer kicking around, this is likely a safer answer: you can ‘calibrate’ the scale factor by comparing the input and output voltages using a multimeter.

[balnazzar]

Hi!
Is it possible to implement Bode Plot for this oscilloscope? That would be really cool.

This oscilloscope doesn't have a function generator, so very very unlikely (someone would have to write quite a lot, to read the input frequency and amplitude in both channels from the measurements then automate timebase and voltage range adjustments, from reading those measurements) I think possible but not worth trying to do it

I don't have one of these scopes, but am thinking about getting an isolated one for 'riskier' measurements to make sure I don't damage my PC or my more expensive Picoscope.  The simplest way I can think of to automate Bode plots with one is to write a Python script to control things.  This came to mind as I am currently playing with a very simple Arduino-controlled AD9833-based function generator project, where I send the desired generator frequency to the Arduino over the COM port using a Python script.   This is not an original idea of course. 

For the measurement part of the Bode plot, it looks like the Python API that florentbr built has an autoset function.  If that autoset works well enough, then florentbr did all of the hard work for us!  It actually only needs to get the amplitude settings okay, since we would know the frequency already so could do a final time-base adjustment if needed (eg, if the signal is aliased to ensure the signal is not aliased).  If the Owon works like my Picoscopes, having the wrong time-base that creates an aliased signal actually doesn't impact the amplitude measurement much, if at all, so it should work out fine. 

Of course, the physical setup for Bode plots would require the function generator output and one scope channel on the input of the DUT, and the other scope channel on the output of the DUT.  Automating the Bode plot would then amount to looping over the frequencies of interest.  For each frequency
1. set the function generator frequency
2. run the autoset function on the scope. Perhaps adjust the time-base based on the known frequency.
3. measure the amplitudes on each scope channel and the phase difference (or time-delay) between them.

After looping over the frequencies, the plots could then be made in Python pretty easily.  There are multiple ways to do step 3, of course.  Without having one of these scopes I wouldn't pretend to know how exactly, but surely the API, in conjunction with other Python functions, could be used. 

Anyway, I just thought I would put that idea out there in case someone cares enough to try and make it work. 

I suspect many folks around here could do the coding better than I can, but I would be happy to share my super-simple Arduino and Python code for the signal generator if folks are interested in trying it out.  This is my first ever project using either a microcontroller or Python, and since I just started yesterday it is a rough work-in-progress not worthy of posting as an actual 'project'.  The analog output I cobbled together with parts in my bin is okay, but still on a breadboard and powered using a bench supply. 

cheers!

jason

Some FRA code seems to be already there, in the example notebook Florentbr provided with the software distribution so to illustrate the API: https://github.com/florentbr/OWON-VDS1022/blob/master/api/python/vds1022.ipynb

Scroll down to "Frequency Response Analysis".

I haven't managed to understand the mechanics, though. He seems to increase the frequency (on the scope side) just by increasing the sampling rate. The external generator is not taken into account.

Maybe you can shed some light about these aspects..

[balnazzar]

More specifically, here is the code, leaving out plotting or importing helper modules:

Code: [Select]

dev = VDS1022(debug=0)
dev.set_sampling('10k')
dev.set_channel(CH1, range='10v', probe='x10')
dev.set_channel(CH2, range='10v', probe='x10')

points, freq_prev = { }, -1
try:
    for ch1, ch2 in dev.fetch_iter(freq=4):
        freq, phase = ch1.freq()  # compute frequency and phase
        freq = freq and round(freq)
        if freq and freq == freq_prev:
            points[freq] = 20 * np.log10(ch2.rms() / ch1.rms())
            dev.set_sampling(freq * 50)  # increase frequency
            print('%d: %dHz  %.fdBv   ' % (len(points), freq, points[freq]), end='\r')
        freq_prev = freq

except KeyboardInterrupt:
    x, y = zip(*sorted(points.items()))  # sort, unzip

Note that this just acquires the pairs (freq,phase), unless I'm missing something obvious...

[jasonRF]

Hi Balnazzar,

I do not have one of these scopes so have not used his code at all.  What I was referring to is the fact that the api appears to have a lot of nice features, including an autoset function.  So in principle it should be straightforward to implement Bode plotting as long as you have a function generator you can control from python.   Having to write your own code to automatically adjust the vertical scale can be a hassle. 

Jason

[Gyro]

This is sounding like it's gaining a little traction. Indman raised it on Florent's github page a couple of months ago but I just noticed that Florent closed it as being a bit too broad a definition. Maybe you could add some additional ideas or detail...   https://github.com/florentbr/OWON-VDS1022/issues/57