REVIEW - Owon SDS7102 - A look at the SDS series from Owon

[pascal_sweden]

I agree that the first step is to use the existing hardware, and see how far you can get with that approach.

Keep us posted about the progress of your project!

[wingel]

By the way.  I'm quite convinced that I saw a open source scope project where they used a FPGA and had a fairly nice looking plastic housing for it.  I think it was a german project and I think they had some major issues with it.  But I can't find that project any more.  Do any of you remember it?

And since I basically know how everything about how the scope fits together now I need to start writing some software for it.  Is there any open source software out there that I can reuse?  Should I just use fbdev for the graphics?  SDL, X, Wayland?  GTK, QT?  Any existing controls that can be reused?  Or any existing verilog/VHDL code I can look at?

[nctnico]

There used to be a scope on the market from a German firm which went belly up. I think they released the firmware but it was such a huge mess people started from scratch. But as I already wrote: when it comes to oscilloscope firmware you probably end up writing at least 3 versions from scratch before you understand the difficulties OR you have to start with a very detailed plan. A good starting point is to assume each trace consists of sample & timestamp pairs which then need to be plotted onto a screen from time X to time Y.

[wingel]

Welec 2000A, that's the German scope.  I just needed something to jog my memory.  I was wondering if there was any source code for that scope that could be useful.

[nctnico]

Yes, the Welec Wittig scope! My memory was rusty as well. It seems they are still working on it (much to my own surprise!) because the last commit on the project was 6 days ago:
https://sourceforge.net/projects/welecw2000a/

[pascal_sweden]

Never heard about Welec or Wittig before. Must have missed that part!
But then again, at that time, I did not have much focus on test and measurement equipment.

[nctnico]

It would be interesting to see how easy/difficult it is to port the opensource Welec firmware to the Owon scope.

[pascal_sweden]

Or maybe it is better to start from scratch, and use High-Level Synthesis and an "All-in-C" based design.

[wingel]

It's too warm to sleep here in Stockholm so here's another blog post:

http://blog.weinigel.se/2016/07/25/sds7102-fast-buses.html

[Hydrawerk]

Yes, the Welec Wittig scope! My memory was rusty as well. It seems they are still working on it (much to my own surprise!) because the last commit on the project was 6 days ago:
https://sourceforge.net/projects/welecw2000a/

Already known: https://www.eevblog.com/forum/testgear/welec-w2000a-another-weird-oscilloscope/msg361147/#msg361147
Welec was probably shit.
This happened in 2007 when Rigol and Siglent were quite unknown in Europe.

[masterx81]

Little update on my 7102, that has always suffered of heavy backgroud noise (as much that i've had 200mvpp noise with the 20mv/div setting on 10x). Simply washing deeply the boards (first water+soap, then some degreasers, at the end demineralized water and isopropyl alchool) have solved all my problems, now almost no backgroud noise. Had also to remove the shield aroud the analog section for ash deeply under it.
A little thing to notice is that the ground of the "compensating output" have some nasty noise but don't bother me as not affect the measurements...

[wingel]

Little update on my 7102, that has always suffered of heavy backgroud noise (as much that i've had 200mvpp noise with the 20mv/div setting on 10x). Simply washing deeply the boards (first water+soap, then some degreasers, at the end demineralized water and isopropyl alchool) have solved all my problems, now almost no backgroud noise. Had also to remove the shield aroud the analog section for ash deeply under it.
A little thing to notice is that the ground of the "compensating output" have some nasty noise but don't bother me as not affect the measurements...

Interesting.  I've also noticed some ashy residues on the main PCB in my scope too.  I didn't think that would affect the measurements though.  But now when you mention it, if that ash is conductive it could carry noise into the weak signals at the AFE.

I'll have to make sure to clean my scope the time I pull it apart.

[wingel]

I'm on vacation and the weather isn't too nice so I've spent waay to much time hacking on my FPGA image.  I've now managed to make the emulated DDR2 memory on the SoC bus read/write and can capture data from the ADC directly into the DDR2 memory attached to the scope.

http://blog.weinigel.se/2016/08/15/sds7102-ddr-capture.html

[JoanTheSpark]

I used the search on this thread and the forum, but couldn't see any results for my problem.
I got a SDS7102V and have no luck recording screenshots/waveforms onto USB sticks/flash drives.

Firmware version is 5.4

What happens:
I use the menu to save an image, insert the USB stick, hit [Save] the keyboard comes up on screen.. either scroll and go back to [Enter] or change the filename and go then to [Enter] or stay on [Enter].. once I press the multipurpose knob for starting the transfer action nothing happens.
If there was a light on the USB device it now goes dark, the screen keyboard stays on the display and nothing further happens (10+ minutes pass by). If I pull the USB stick out and check it for images, there is an empty (or several, depending how often I do this) bmp file in it with the name I set/chose via the on-screen-keyboard, but size is 0 kB).
Oh, and after I pull the USB stick the scope works again (after showing for some moment a little bar instead of the on-screen-keyboard).

Anyone seen such behavior on those units?
Any fix?

I checked their website with the scopes particulars and it tells me I got the latest fw for it..

[Orange]

I used the search on this thread and the forum, but couldn't see any results for my problem.
I got a SDS7102V and have no luck recording screenshots/waveforms onto USB sticks/flash drives.

Firmware version is 5.4

What happens:
I use the menu to save an image, insert the USB stick, hit [Save] the keyboard comes up on screen.. either scroll and go back to [Enter] or change the filename and go then to [Enter] or stay on [Enter].. once I press the multipurpose knob for starting the transfer action nothing happens.
If there was a light on the USB device it now goes dark, the screen keyboard stays on the display and nothing further happens (10+ minutes pass by). If I pull the USB stick out and check it for images, there is an empty (or several, depending how often I do this) bmp file in it with the name I set/chose via the on-screen-keyboard, but size is 0 kB).
Oh, and after I pull the USB stick the scope works again (after showing for some moment a little bar instead of the on-screen-keyboard).

Anyone seen such behavior on those units?
Any fix?

I checked their website with the scopes particulars and it tells me I got the latest fw for it..

Have you tried an other USB stick ?

[JoanTheSpark]

I tried 2 different USB sticks and 1 SD card reader with 2 different cards.
I get empty bmp files written on them, but I essentially have to yank them out of the unit as there is no progress bar nor does it seem to finish (expected 35 seconds according to experience of others with this unit).

Just checked the manual from their site and see that they explicitly mention FAT32 with 4kB clusters.. well, seems I missed that in another online manual I checked earlier that wasn't directly from them were that wasn't mentioned at all.. 

It's working and feels like 5 seconds only to me from hitting [save] via the multiknob till it's done writing with the progress bar.
The [copy] button under the multiknob still doesn't seem to do anything (as I have read on the interwebs), so no fault there I guess.

Thanks for your help. Much appreciated.

[wingel]

I used the search on this thread and the forum, but couldn't see any results for my problem.
I got a SDS7102V and have no luck recording screenshots/waveforms onto USB sticks/flash drives.
ed their website with the scopes particulars and it tells me I got the latest fw for it..

Strange, my comment disappered, I might have had too many links in there.

Anyway, I think this is  a known problem.  The cluster size of the FAT file system on the USB drive must be 4 kBytes or smaller, large USB memories can have a cluster size of up to 32 kBytes.   Google for "owon sds fat32 cluster size".

Try formatting the USB memory with "FORMAT X: /FS:FAT32 /A:4K" or "mkfs.vfat /dev/sdx1 -s 8 -F 32" on Linux.

[AndrejaKo]

A question for other users of this scope:

I'm slowly getting back into electronics after a 3-year break. My scope was stored (in "normal" room conditions) unused for a while and now it developed a series of issues which are probably inter-related .

First thing I've noticed is that the MC34063A is getting quite hot.
The stock Chinese part had issues back when the scope was new. Basically, it would start switching, then stop at around 50% of transition, take a short break, drink a bit of coffee and then complete the switch. Changing the chip to an Onsemi part fixed the issue. For some reason I don't remember now, I also replaced the R33 current-limiting resistor with a short. Back when I was using the scope, things seemed to be working fine for at least 1 year if not more.
Now, on the other hand, the chip is getting really hot and is untouchable when the scope is connected to the PSU. After a while, it will stop working and the voltages on screen drift up to infinity. If the scope is disconnected from the PSU, the chip remains cool. A while later, the Onsemi regulator died completely, so I replaced it with another one. At the same time, I removed the short and put pins for a multimeter in its place. One thing I did notice is that when I have multimeter in series, it stabilizes the MC and the rest of the scope is using a varying amount of current that seems to drift between 3,5 mA and 5 mA. Of course, with such current, the regulator stays cool. If I replace the multimeter with a regular short,  regulator gets hot again. The power supply is PCB-T115-J rev4. Unfortunately, I don't have another handy source of negative voltage, so I can't measure how much current the scope would like to take if there's no limiting from the MC.

The other issue I have is auto-calibration. Basically, the scope won't properly auto-calibrate and on some voltage ranges, the zero volts line has offset. I'm not really sure how to go around troubleshooting this thing. Also the rechargeable battery on the ADC board is at around 0.8 V. This is with the multimeter connected to the MC34063A's current limit pins. The ADC board is SDS1209 version.

The third issue is that vertical position adjustment is a bit jumpy. Basically, I'd adjust it in one direction and the line would go up and down and sometimes multiple steps for one step of encoder rotation. My gut feeling is that it's most certainly a bad encoder, but I thought I'd just mention it here for completeness's sake.


So does anyone have any ideas what could be causing any of the issues?

[wingel]

The third issue is that vertical position adjustment is a bit jumpy. Basically, I'd adjust it in one direction and the line would go up and down and sometimes multiple steps for one step of encoder rotation. My gut feeling is that it's most certainly a bad encoder, but I thought I'd just mention it here for completeness's sake.

No idea about the other questions, but the  jumpiness is just crappy firmware in the scope.  All mechanical switches such as the ones in the rotary encoders are noisy and need some kind of debouncing.  The firmware in the scope does a really bad job with debouncing and the position will jump.  I have that problem with all rotary encory encoders on my scope even if some of them are worse than others.

[AndrejaKo]

The firmware in the scope does a really bad job with debouncing and the position will jump.  I have that problem with all rotary encory encoders on my scope even if some of them are worse than others.

Really?? Then I must have been lucky. I myself didn't have any jumpiness issues previously.

By the way, thanks a lot for the great work on firmware hacking!!

[wingel]

By the way, thanks a lot for the great work on firmware hacking!!

Thanks.  But I need to get back to it.  A lot of things have happened (playing Satellite Reign and my job getting bought by Nokia) and I haven't had time to do any hacking on the scope for months.

I know how all about how the hardware works now, now I "just" need to build a usable firmware from it. 

[lemon]

Wingel, we have subscribed on your project with interesting, many of us!

Your job with 7102 is marvelous and you had many congrats for that!

We will wait for the fw update from you...

@AndrejaKo = The "jumping" on vertical position adjustment is as fw issue as rotary encoder fault, specially if the device had no working for long time (it is know issue).
2-3 weeks on daily operation in normal temperature room, will recover it enough but no 100%.

[l3VGV]

Hello owon owners! I have a question to everyone who looked inside the scopes. I was replacing encoders, and decided to take a look to input stage of adc.

My unit use MXT2001, and channel 1 is a little bit different from channel 2. 

Look at the marked part, ther is a some part marked 30A and a capacitor. and second channel do not have a cap.
First of all, why are they even there? And what ar they doing, cap maybe to limit higher frequency, but if 30A is a resistor, what the point to attenuate signal?
Not every 7102 based scope have them, if u look at diferent posts sometime ther ar resistor and cap at the adc input, sometimes not.

In my case i measure 70Ohms at both adc inputs instead of 100. So is that 30A a resistor?

[lemon]

Yes, the smd 30A is the R230 resistor. It is marking with 30A, ie 200ohm.
I don't know why the above capacitor C83 is missing in your Owon.
Is that something to the final measurement (I doubt), but as I see at my Owon these are identical and there is no missing any capacitor !

[toxicshadow]

Good to see the amount of reverse engineering going on with this scope. For those who don't know the official OWON scope software has the ability to send commands via usb to control a large proportion of the scopes options. Unfortunately this is windows only and tbh the installer for the software seems a little questionable.....

Either way after a few hours with wireshark I have extracted most of the commands and have written a very basic proof of concept program to allow Linux (or any os that supports libusb1.0) to send and control the oscilloscope using a (very) basic command prompt interface.

I hope to one day create a QT interface to mimic the official owon app ( but better ) but progress may be slow based on work priorities etc...

For anyone interested in incorporating usb functionality into their apps, take a look at my usb_interface.c file to see the commands and important byte adjustments.

https://github.com/7oxicshadow/owoncontrol