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

[Carrington]

@Carrington How much spray did you use to cover the inside of the scope?

You need a full pot of EMV35. I also use adhesive paper tape in big holes, you need to fill all button holes with plasticine without dirtying the rest around.

[Carrington]

Nice paint job!

Thanks, is easy to do.

Did you fix your adapter board? Or did you get new boards from Aidatek?

Yes, I get the new boards from Aidetek, but this was not easy. Owon don't want selling directly to me the adapter board. Why? I don't know.

[Carrington]

I have pending shield the screen, the necessary material to do this is expensive and hard to find.
Is important find the best point to joint the metal Z-shaped plane with the housing shield.
Without the shield and using the short gnd probe cable the noise was around 50 mV.

[lemon]

Your final results after the changing of two boards and your works with the spay shielding is marvelous.
Your gnd-noise level of 28mVp-p isn't with a loop cycle of probe cable but with the straight. If you measure it again like as we have said, I am very sure that you'll have under 20mVp-p.

Your test with FFT and the small radio was "all the money"! 

The next one for all our forum company is to make a good fw for this oscilloscope and change the label as "AWLtB" (As We Like to Be)" 

[Carrington]

Your final results after the changing of two boards and your works with the spay shielding is marvelous.
Your gnd-noise level of 28mVp-p isn't with a loop cycle of probe cable but with the straight. If you measure it again like as we have said, I am very sure that you'll have under 20mVp-p.

Your test with FFT and the small radio was "all the money"! 

The next one for all our forum company is to make a good fw for this oscilloscope and change the label as "AWLtB" (As We Like to Be)" 

Well, now there is not much difference.

[TomC]

  - Note the peak over 106.5MHz, i.e. internal FM receiver Osc. to get 10.75MHz IF.

Great results! Congratulations!

The effect of the radio on the FFT is interesting. With the shielding on the cabinet I assume the radiation is coming through the probe cable. Did you try moving the radio further away to see when the effect decreased?

[TomC]

Well, now there is not much difference.

Looks like 5mVpp better!
When the GND noise is that low a few millivolts less are hard to achieve. With the ferrites on my probes my noise is still a little higher than that. However, the position of the probe cables in respect to the scope don't have much of an effect.

[Carrington]

The effect of the radio on the FFT is interesting. With the shielding on the cabinet I assume the radiation is coming through the probe cable. Did you try moving the radio further away to see when the effect decreased?

Obviously that's what happens.

  - "Radio Osc" -> 106.5MHz + 10.75MHz. Away from gnd loop ~1cm.
  - "Cadena Ser" -> 106.5MHz FM. Away from gnd loop ~1Km.

[AndrejaKo]

Great work Carrington! I agree with lemon, the demonstration with FFT and radio is really great.

Anyway, I did the backlight mod and it got rid of the peaks in its area. Results are same as lemon's.

I've also changed resistors for the comparator pin of the MC 34063A, but that gave no improvement to the duty cycle issue. It' still changing all the time. I took some time and probed around the PSU and I've noticed that the whole +8.4 V rail is contaminated with noise. I'm not sure from where it could be coming. It doesn't look to me as if the main source could be coming from the MC. I've probed the comparator pin on MC and results are attached. I also removed the capacitors that were going from the resistors to ground at the sense pin.  This made the noise a bit clearer to see.

Basically the noise is cyclic and here's how cycle looks like: First, there's the quiet period seen on the left side of the picture. Here we have the usual switching noise from the MC. Then near the trigger point, comes the first wave of the noise. Then there's a short quiet period (10 microseconds here). After it comes the second noise noise wave. After comes the long quiet period. I'm also not sure how to actually trigger on this event. I captured it by repeatedly pressing the single button.

The probing was done with the oscilloscope running on battery.

If I understand how this PSU works, when on battery, everything south of D8 should be off. This means that the only two ICs running are the MC34063A and the LM324. The LM shouldn't do anything smart when the PSU is running on the battery. My idea is to try replacing it and see if anything happens.  I'll also see if I can re-trace the board and make a schematic of the area near the LM. In my version of the PSU, all four op-amps are in use.


If this doesn't help, I'll see if I can run the PSU without both the MC and the LM.

[TomC]

Great work Carrington! I agree with lemon, the demonstration with FFT and radio is really great.

Anyway, I did the backlight mod and it got rid of the peaks in its area. Results are same as lemon's.

I've also changed resistors for the comparator pin of the MC 34063A, but that gave no improvement to the duty cycle issue. It' still changing all the time. I took some time and probed around the PSU and I've noticed that the whole +8.4 V rail is contaminated with noise. I'm not sure from where it could be coming. It doesn't look to me as if the main source could be coming from the MC. I've probed the comparator pin on MC and results are attached. I also removed the capacitors that were going from the resistors to ground at the sense pin.  This made the noise a bit clearer to see.

Basically the noise is cyclic and here's how cycle looks like: First, there's the quiet period seen on the left side of the picture. Here we have the usual switching noise from the MC. Then near the trigger point, comes the first wave of the noise. Then there's a short quiet period (10 microseconds here). After it comes the second noise noise wave. After comes the long quiet period. I'm also not sure how to actually trigger on this event. I captured it by repeatedly pressing the single button.

The probing was done with the oscilloscope running on battery.

If I understand how this PSU works, when on battery, everything south of D8 should be off. This means that the only two ICs running are the MC34063A and the LM324. The LM shouldn't do anything smart when the PSU is running on the battery. My idea is to try replacing it and see if anything happens.  I'll also see if I can re-trace the board and make a schematic of the area near the LM. In my version of the PSU, all four op-amps are in use.


If this doesn't help, I'll see if I can run the PSU without both the MC and the LM.

As far as I know, you are right in all the assumptions about how the circuit works. I assume you are doing all testing in the no-electronics room, so we can assume there is no external interference. I doubt that the LM can cause any trouble here, but you can probe the output pins to see if there are any oscillations. The noise in the 8.4V rail may be causing the MC to misbehave, there are not very many other suspects, since I think you have replaced just about every component in that area. There may be something wrong with the battery itself, or its protection circuits. Have you tried a ferrite on the battery cable to see if this reduces the noise on the 8.4V rail?

[TomC]

The effect of the radio on the FFT is interesting. With the shielding on the cabinet I assume the radiation is coming through the probe cable. Did you try moving the radio further away to see when the effect decreased?

Obviously that's what happens.

  - "Radio Osc" -> 106.5MHz + 10.75MHz. Away from gnd loop ~1cm.
  - "Cadena Ser" -> 106.5MHz FM. Away from gnd loop ~1Km.

Thanks for the info!
That FM station is really close to you!

[AndrejaKo]

As far as I know, you are right in all the assumptions about how the circuit works. I assume you are doing all testing in the no-electronics room, so we can assume there is no external interference. I doubt that the LM can cause any trouble here, but you can probe the output pins to see if there are any oscillations. The noise in the 8.4V rail may be causing the MC to misbehave, there are not very many other suspects, since I think you have replaced just about every component in that area. There may be something wrong with the battery itself, or its protection circuits. Have you tried a ferrite on the battery cable to see if this reduces the noise on the 8.4V rail?

I added a ferrite to the battery lead, but it did not help. I also removed the PSU from the scope and did some measurements with it connected only by cables. The Z-plate is still full of noise spikes and I think that the noise could be coming from the Z-plate to the PSU. The type of noise I previously described to me seems a bit stronger on the Z-plate than on the PSU ground. That should be expected, since I have ferrite on PSU to adapter board cable.

A few things to try out:
I have a linear PSU which should be capable of outputting up to 2 A. If that's enough I could try running the scope from that and another battery (for negative voltage) and see the amount of noise inside the scope. This should allow me to see how much noise is generated by the adapter board and the main board by themselves.
Next, I could run the PSU with a dummy load and see how it behaves. If the previous step works out fine, I could easily test that out, otherwise I'd have to wait for access to another scope.

[TomC]

I added a ferrite to the battery lead, but it did not help. I also removed the PSU from the scope and did some measurements with it connected only by cables. The Z-plate is still full of noise spikes and I think that the noise could be coming from the Z-plate to the PSU. The type of noise I previously described to me seems a bit stronger on the Z-plate than on the PSU ground. That should be expected, since I have ferrite on PSU to adapter board cable.

A few things to try out:
I have a linear PSU which should be capable of outputting up to 2 A. If that's enough I could try running the scope from that and another battery (for negative voltage) and see the amount of noise inside the scope. This should allow me to see how much noise is generated by the adapter board and the main board by themselves.
Next, I could run the PSU with a dummy load and see how it behaves. If the previous step works out fine, I could easily test that out, otherwise I'd have to wait for access to another scope.

I measured the current consumption for the 8.4V and -7.6 a while ago. The 8.4V used about 2.2A with no battery option. Here is a link to where I reported those findings:

https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg281838/#msg281838

If you have a 2A PSU I would try it anyway, it may be able to supply a little more, just watch for overheating. The info on the link should also be helpful to set up a dummy load.

I agree with your reasoning about the noise in the Z plate versus PSU GND and the ferrite, so I think that would be a worthwhile experiment.

[lemon]

Yes, I get the new boards from Aidetek, but this was not easy. Owon don't want selling directly to me the adapter board. Why? I don't know.

The Aidetek starting to apply waranty cover to gnd-noise issues to their customers.
I'll have more information at the end of this month, about this.

...
Without the shield and using the short gnd probe cable the noise was around 50 mV.

Well, now there is not much difference.

Looks like 5mVpp better!
When the GND noise is that low a few millivolts less are hard to achieve. With the ferrites on my probes my noise is still a little higher than that. However, the position of the probe cables in respect to the scope don't have much of an effect.

I agree with the TomC, 6mVp-p isn't a big improvement but it is hard to achieved from 28mV to 22mV.
Also, it seems that your shielding works very good.
If you see at the capture with 22mV all the periodicaly noise (spikes) has gone and the only noise that remaining is only the base. Very good result

...
Anyway, I did the backlight mod and it got rid of the peaks in its area. Results are same as lemon's.
...

It is fine, seems to work to all versions 3.0.

You thinking right about test with the external PSU to the adapter board, but have in mind that the circuit around of AOZ1094 is the biggest source of noise on this adapter.
My opinion is to modificate this area too.

[lemon]

How can I save to external usb stick or to pc via Oscilloscope software a record waveform?

For example go Save->Type-Record->Mode-Record and click to Operate-Play, when you want to stop press simple the Click.
If you want to see what you have record follow the Mode-Playback->Operate-Play. With the Playmode-Loop the capture waveform plays without stop.

Just fine until now, but I can't find how this captured waveform can to export.

[TomC]

How can I save to external usb stick or to pc via Oscilloscope software a record waveform?

For example go Save->Type-Record->Mode-Record and click to Operate-Play, when you want to stop press simple the Click.
If you want to see what you have record follow the Mode-Playback->Operate-Play. With the Playmode-Loop the capture waveform plays without stop.

Just fine until now, but I can't find how this captured waveform can to export.

As far as I can see, you can only save to internal memory. I don't see any way of exporting to USB stick or PC.

I also noticed that when I use this function it interferes with the AC trigger. Once I use it, AC trigger don't seem to work until I change to DC trigger. Then, when I change to AC trigger again it works OK.

[TomC]

Firmware Upgrade for the SDS scopes ? 

While checking the Owon site for firmware updates I happened to notice that the RAR file for the 3.2 & 3.3 patch was slightly longer than the one I had downloaded some time ago. In addition, the date of the upgrade had changed to a later date. Knowing that Owon is a little unorthodox in the way they name and issue new releases, I went ahead and downloaded the file (took more than 15 minutes) and installed the patch.

As far as my scope is concerned, this is a different and improved firmware release even though it still has the same patch number that was released quite a while ago.

One thing I noticed is that the functionality of the LAN interface is better, I can again transfer images, a functionality that hadn't work for me for quite sometime now. The high memory depth transfer works most of the time up to 1M acquire length. However, my scope still freezes if I attempt a 10M transfer.

I also noticed that the firmware version on the About screen displays 3.2. I installed the patch a second time to see if would change to 3.3 but it didn't.

The last thing I noticed is that the "SCPI Command Line" button on the Remote Control Panel appears to be functional. I didn't try any commands, but it no longer emits an error message when you click it.

While at the Owon site I also downloaded the latest SDS manual. This is also a newer manual (V1.6.3). A quick look revealed that they have added the SDS7072 to the manual. They have also added and rearranged to the information for the PC software and LAN interface.

[TomC]

I tried the SCPI commands, they all seem to work.

The one thing that puzzles me is the Load File button. I can create a text file with SCPI commands in it and load it via this button. The first command in the text file then appears on the command line and I can press SEND to execute it. But that's seems to be the end of it. I couldn't find a way to progress to the next command in the text file. So it seems that it only works for a single command per file, or perhaps is not yet completely implemented. If anyone finds a way to make this work more like a macro, please post it!

[lemon]

How can I save to external usb stick or to pc via Oscilloscope software a record waveform?

For example go Save->Type-Record->Mode-Record and click to Operate-Play, when you want to stop press simple the Click.
If you want to see what you have record follow the Mode-Playback->Operate-Play. With the Playmode-Loop the capture waveform plays without stop.

Just fine until now, but I can't find how this captured waveform can to export.

As far as I can see, you can only save to internal memory. I don't see any way of exporting to USB stick or PC.

I also noticed that when I use this function it interferes with the AC trigger. Once I use it, AC trigger don't seem to work until I change to DC trigger. Then, when I change to AC trigger again it works OK.

Yes, I noticed the same function to internal memory but nothing to export. The Oscilloscope pc software has the save from hi memory depth but saves only static images.
The menu Save is a little complex, there is save output (internal or external) to Wave Type or Image Type that the usb stick works but the same to Record Type isn't.
Also, I noticed that the Record Function is only to view the waveform to loop or at once, not to stop this and zoom with Time button. If you try this you'll laugh a lot of...

[TomC]

Also, I noticed that the Record Function is only to view the waveform to loop or at once, not to stop this and zoom with Time button. If you try this you'll laugh a lot of...

I had the same impression, it's a joke
Maybe there is a reasonable application, but for the life of me, I can't figure out what that may be!

By the way, there is a similar feature from the PC software that I tried last night. You got to set a data storage folder to store the BIN files, and then you can play them with the Auto Player. Uses gobs of memory. But here you can at least do a little bit more with it.

[lemon]

TomC, I''ll look the pc software another time because I have the scope full unfit for painting with Contact Chemie 35 now.

By the way, I can't remove the last board, this of buttons and there is no any information at the video or service manual.
Have any ideas, how to remove it? Carrington, where are you?

[Carrington]

Pulls the buttons strongly, you can use a rubber band and a pliers. Roll the rubber band arround the button and strip with the pliers, but be careful.

Notes: Do not paint this part (red) cover it with adhesive paper tape, like a cigar. Use also the adhesive paper tape to cover the outer edges of the plastic housing, i.e. all around. You can glue too a foil of paper around the housing using the adhesive tape. By this way you do not paint where you should not. Please use a protective mask too.

Edit: Remember you need a full pot of EMV35, nothing remains. Also you need fill all the button holes with plasticine without dirtying the rest around. To clean the plasticine rest you can use use alcohol and cotton tipped swabs. Do not forget cover the hole for the screen with adhesive paper tape.

I forgot, does not spray the paint from afar, and sprayed with the pot in vertical. Better read the instructions.

Good Luck. 

[Carrington]

The board in the photo above radiates a lot.

[AndrejaKo]

TomC, how exactly the AC trigger does not work? I think that I used the waveform record function once on AC trigger and it seemed to work fine to me back when I did it. I don't remember if it was with the newest firmware.

Lemon, the only way I see that the recorded waveform can be transferred is to play through the recording frame by frame and then copy the waveform using wave or image types. This is very time-consuming!

A small update from me: I added the SMD capacitors from lemon's adapter board mod for the middle voltage regulator (specifically one 220 nF for D127 and one 220 nF in parallel with C33), but it seems that the result was negative.
I can now see a nice waveform on scope screen even with no probes connected and the peak to peak voltage with nothing in the BNC is around 32 mV. Previously it was around 26 mV. I'll try undoing the mod and see if that returns previous noise level.

After removing the capacitor going from D127 to ground, the noise level with nothing connected went back to normal.

I also managed to trigger on those "noise bursts" I mentioned several times before. Required settings for me were slope trigger, when=30ns, Threshold&SlewRate High level= 120 mV, Low level=- 120 mV. Now I'm almost completely sure that the noise is in fact coming from somewhere around the MC. I set the channel one to trigger on the noise and probed around with channel two. I placed the probe on top of almost every IC I could access and only near the MC did I see waveform on channel two synchronized with channel one.


It's also interesting to see that Owon expanded the SDS-E line to whole range of scopes. High memory seems to be an optional feature now. The other differences appear to be a slight update on of the shell and removal of the battery. I also noticed that it can be self calibrated with no warm-up period. In fact, they mentioned that in the new SDS manual as well. This seems a bit strange to me. Also since I'm bashing their autocalibration instructions, I'll mention this gem from the manual: "If the change of the ambient temperature is up to or exceeds 5?, the self-calibration procedure should be executed to obtain the highest level of accuracy." What's the point of mentioning the temperature of the change is up to or exceeding that value? Perhaps they meant equal or greater than 5?.

The SDS7072 has adjustable output for the scope compensation port. Frequency can go from 1 kHz to 100 kHz and duty cycle from 5% to 95%. I wonder why that would be a useful feature. I know that high frequency probes need higher frequency for compensation in addition to low frequency, but that high frequency is usually around 1 MHz. Also I'm not sure how variable duty cycle would help. Any ideas?

All in all,  it seems they added four new models of scopes based on SDS7102. This seems a bit too much to me.

[Carrington]

Shit, I forget this  :

  Covers the same part that you see in yellow, otherwise the paint will cause a short circuit.