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

[lemon]

172mVp-p is an average noise level with Acquire Mode to Peak Detect, but it is the best that you have measured.
If you set Acquire Mode to Sample, you are going to measure a range of 140-160mV.

[lemon]

Look at these red arrows.
All these are decoupling smd capacitors that Owon adds to psu board for decreasing the noise level.

At the yellow one, they cutting the connection of power switch metal cage with the gnd (it was an antenna of radiation).
At the orange one, you see the rerouting of ground plane of board.

I think all these modifications was suggested to Owon from rf-loop member.

[TomC]

Please can you take pictures from new adapter board (whole board) bottom and top side.
Same for new PSU board.

This picture what you now show about adapter board..... 
I really think now that Owon want totally shoot down themselves.
I can not see on the adapter board any signs of Owon made needed modifications.
5.5V and TFT backlight SMPS circuits need redesign! There need only follow simplest rules for design low noise SMPS circuits. I have done lot of free work for Owon and send lot of documents how to do it. 
(there need redesign PSU 8.4V  and -7.6V,  Adapter board TFT backlight, 5.5V and also 3.3V.
Also there is some other small things but they are not so acute - Example TFT databus)

And I can see they have not understood anything. (in this partial picture what TomC show)

I'm so angry now that it is better I do not write more...

rf-loop, the Adapter board pictures I posted are from my original adapter board with the modification Owon had me do as part of the upgrade. So apparently, the only difference in the adapter boards of the scopes now being produced is that additional 1000pF capacitor bypassing the 10µF electrolytic on top. I'm surprised about that too since you had quite a few more modification in version 2b of your DIY that Owon approved.

The pictures of the new PSU board and the little package containing power button caps and SMD capacitors are on post#1370. I can take pictures of the original PSU if you want to compare the two.

I'm also disappointed at Owon's decisions, they redesigned the PSU board and added a capacitor to the Adapter board, but their solution don't give as good results as what you did for free for them with your version 2b DIY. I wish they had paid more attention and at least compare the results and pick the best solution.

[TomC]

I suggest to him, add a ferrite to battery ribon cable, also)

lemon, I don't have the battery option, that's why I didn't put a ferrite on the battery cable. However, I agree, that anybody that has the battery option would benefit from a ferrite on that cable. I didn't order the battery option because I don't need portability, I just use the scope around the house and there are AC outlets everywhere. If I need to do a floating ground test, I have an isolation transformer. Warning: Never use a scope for a floating test that is not specifically designed for this, the SDS7102 is not designed to do floating tests even if battery operated! Failure to heed this Warning can result in physical injury or death! However, I don't always do as I preach

[J4e8a16n]

What is a floting test?

JP

[marmad]

If I need to do a floating ground test, I have an isolation transformer. Warning: Never use a scope for a floating test that is not specifically designed for this, the SDS7102 is not designed to do floating tests even if battery operated! Failure to heed this Warning can result in physical injury or death! However, I don't always do as I preach

For the SDS Series, it's safe to “float” the “signal common” for making measurements provided you don't connect a signal greater than 30 VRMS (>42 Vpk) from earth ground to either the probe tip or common lead. For measurements where higher voltages {>30 VRMS, >42 Vpk) are present and running the DSO off the battery, the instrument’s chassis must be connected to earth ground using a grounding wire from the screw at the back of the DSO to prevent electrical shock to the operator.

If there's any doubt whether more than 30 VRMS is present or not, a grounding wire should always be used when running off the battery - and floating measurements SHOULD NOT BE ATTEMPTED! Be aware that hazardous voltages may exist in unexpected places due to faulty circuitry in the device-under-test.

[TomC]

What is a floting test?

JP

For example, you want to view the signal across a resistor, but neither end of the resistor is attached to ground. If you attach the scope's ground lead to either end of the resistor you effectively create a short to ground. One way around this, but strongly not recommended because of the danger involved, is to float the scope. For example, a battery operated scope can be used without plugging it into an AC outlet. Since in this case the scope's ground lead is not connected to earth ground, connecting it to the aforementioned resistor wouldn't cause a short. However, the scope's exposed metal parts are now at the same potential as the resistor's end where you attached the ground lead. If this happens to be a high enough voltage and you inadvertently touch a metal part of the scope, or if there is an insulation breakdown, you'll be lit up like a light bulb.

[digsys]

I've run isolated CROs / DSOs for 20+ years. I have custom made mag shielded / filtered / line conditioner units with an available hard earth
connection if required. I have them set up in production areas as well. YES, you HAVE to be aware of possible dangers, but they are no
different to the days we had to work on 3-phase (415V) LIVE, or half the other dangerous stuff people have to do. It does NOT harm the CRO/DSO.
We often need to see AC current waveforms into capacitor banks and sometimes I connect across a 3 Phase 200VAC+ motor to check switching
noise and current waveforms. YES, you have to be AWARE and careful, but if you NEED to do it - DO IT.
Once it's isolated, touching ONE end doesn't give you a shock. Worst you feel, is a tiny "capacitance" or "rf" tingle - which reminds you it's there :-)

[marmad]

I've run isolated CROs / DSOs for 20+ years. I have custom made mag shielded / filtered / line conditioner units with an available hard earth
connection if required. I have them set up in production areas as well. YES, you HAVE to be aware of possible dangers, but they are no
different to the days we had to work on 3-phase (415V) LIVE, or half the other dangerous stuff people have to do. It does NOT harm the CRO/DSO.
We often need to see AC current waveforms into capacitor banks and sometimes I connect across a 3 Phase 200VAC+ motor to check switching
noise and current waveforms. YES, you have to be AWARE and careful, but if you NEED to do it - DO IT.
Once it's isolated, touching ONE end doesn't give you a shock. Worst you feel, is a tiny "capacitance" or "rf" tingle - which reminds you it's there :-)

From Tektronix document on floating:

"This is dangerous, not only from the standpoint of elevated voltages present on the oscilloscope (a shock hazard to the operator), but also due to cumulative stresses on the oscilloscope’s power transformer insulation. This stress may not cause immediate failure, but may lead to future dangerous failures (a shock and fire hazard), even after returning the oscilloscope to properly grounded operation.

Not only is floating a ground-referenced oscilloscope dangerous, but the measurements are often inaccurate. This results from the total capacitance of the oscilloscope chassis being directly connected to the circuit under test at the point where the common lead is connected."

[TomC]

I can't really disagree with anything that has been said here. In the past I also often worked with floating CROs, never had one fail as the Tektronix article states, even though every one of them was built by Tektronix. However, that doesn't mean it can't happen, after all, the manufacturer warns against doing it instead of claiming that it is a unique feature or capability. One may say that this is just a way to avoid liability, but in all probability they experienced failures while repeatedly testing their instruments under similar conditions. To me the main point is that floating scopes that are not specifically designed for this purpose can be very dangerous, even life threatening. So I wouldn't want to be the one that encourages someone that doesn't understand the risks or what they are doing to use this method. Yes, I have done it myself, but I was fully conscious of my actions and fearful enough of the risks to double check everything.

[tinhead]

Thanks to rf-loop's post I was able to convert to .csv. Here are the files renamed to .txt.

thanks TomC,

yeah, with GND was of course nonsense, i forgot that Owon is really switching to GND so lot of zeros inside csv.

I did FFT on the csv to check the if there is something visible from the sampling clock, it seems that the 250MHz image frequency is there.
Something is at 21.4MHz, and few third source signals between 75 and 86MHz as well.

[lemon]

Thanks tinhead for the fft results.
Your results is almost identical, with ones that we had to an unmodified SDS7102.
As you can you see at the 2:30 there is something at 80-90 MHz and 250MHz.

https://www.youtube.com/watch?feature=player_embedded&v=R5c1QNiBnQo#at=162

There was an older post of rf-loop, very interesting about two units of SDS7102 one without "ground noise" and other with it (post #1102).

https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/?action=dlattach;attach=47346;image
https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/?action=dlattach;attach=47348;image

At these two captures seems that the unit without noise there is no any 250MHz peak and of course has a low peak at the zone of 80-100MHz.

I think that will be very interesting if this fft test repeated with a new modified unit, as TomC's unit. TomC repeat this test and with low frequencies like 21.4MHz, please.

[TomC]

I think that will be very interesting if this fft test repeated with a new modified unit, as TomC's unit. TomC repeat this test and with low frequencies like 21.4MHz, please.

I captured some FFT images. The first two are with the scope setup to view GND noise. The next three are with nothing connected to the BNC. In Run mode the signals are random and the capture don't really represent what you are looking for, like spikes at the 250MHz and vicinity of 25MHz.

So what I did is set the trigger to Normal for the GND noise images, and bring the level up until it no longer triggered, then back until it triggered once in a while. This way I could capture when the highest noise spikes were present.

For the images with nothing connected to the BNC there is not enough signal to trigger, so I had to use Single trigger mode and Force trigger until I got some images that showed the area of interest. But with the scope set to 5MHz/Div if you are in Auto trigger all you see is random, so I don't see anything significant in the 25MHz vicinity.

Now that I have a spare PSU, I've been examining it and trying to draw a better schematic than what they have on Version 1 of the Service Manual. The one thing that struck me as odd, so far, is that there isn't a bypass capacitor anywhere between either PSU ground plane and Frame ground. However, there is one between the two ground planes. I haven't checked to see if that's something that rf-loop added on his DIY version 2b. Let me know what you think.

[lemon]

I see at the captures 1 & 3 (gnd noise & nothing to bnc) that the spike of 250MHz is smaller than before but is remaining. Interesting!
As I had told, I am in summer holidays and I can't try with my mod Owon.
I think that the 250MHz spike has a relation with the adapter board and perhaps explains why at rf-loop's captures is missing.

The pattern of spikes at range 80-100MHz, as you told, it isn't stable and there is no conclusions by static image.

At the old psu board, there is no any decoupling capacitors. These are added by rf-loop and this solution suggested to Owon.

As you can see at post #1383 there are a lot of them at yours board. The improvement of psu board has finished to my opinion.

If you have open the unit still, you can investigate the noise on specific area of adapter board.
This investigation method has analysed before from rf-loop (You must used two probes, the first one on Z-plate both tip and ground together, the second without ground clip and with protected tip - with the second probe approach all the specific areas) - sorry but I can't to explain better with this level of my english.

[rf-loop]

250MHz peak come from ADC system. (this is proofed and it is exactly synch with ADC clock)
This peak can see also with Owon FFT.
(there is other peaks then what are related to several SMPS circuits but it os other story and these do not pollute signal pathway so much internally. This individual 250MHz peak (+ 500MHz) is there internally. But ... front panel read, this is 100MHz oscilloscope....

Interesting is that (many cases) if scope is cold and start up, then run FFT and look this peak. Note level. Then do fast power break, just off and immediately on. Many times this peak level drops. But if it is very low, then it may also in some cases rise). But this experience is with scopes what have used RuiFeng ADC when I have done more inspections. It is possible this is somehow related to ADC chip internal selfcal routines?)

(Run scope to warm, reboot.. sometimes it give small improvement for signal quality.)

Typically this peak have been well under -40...-50dB from 0dB (@1MHz) reference level (full scale) using 100mV range and FFT  (10 dB  scale) If I remember right. (Im out from work/workshop, can not now test. After some kind of partial summer holiday can do more )

(ADC have 4x250MHz datastream or better say 2 x 2 x 250MHz. Perhaps this is related to this peak.  PCB topology is not best possible for isolate digital and analog sides.)

[lemon]

rf-loop very interesting information.

As you said the peak of 250MHz is from ADC clock and this can't effect the operation of scope becouse it is a 100MHz scope.
As we know at the -3dB, BW is almost 200MHz (for units with version 2.4.1 and after) but this is under from 250MHz peak.

[rf-loop]

Here in test image is visibble this self generated 250MHz "spurious" signal.

This image show how it is related to real signals in practical signal levels.

Scope Ch1  10mV/div
Signal adjusted so that using 1MHz this level give around full 10 divisions oscilloscope image.
Oscilloscope display infinite persistence.
Acquire mode normal.
FFT set for 1+dB/div and vertically adjusted bottom to bottom. Horizontally zoomed x2 and shifted 0 near left side.
First sweep from 1MHz to 360MHz
Then stopped and set for 107MHz.
(signal generator (HP8644B) level display -16.3dBm and this have measured that flatness is far better than specs limits) Around 34.24mV rms.

Botom part of image exatly same signal dropped 50dB to -66.3dBm (around 108uV)
(not in this image but 10dB less and still clearly detectable (34.24uV))

Impedance mathing is very poor,  only 50ohm terminator with T so freq response suffer some amount about this.

250MHz (ADC circuit produced) spur is marked (also 500M 2. harmonic (and ADC base clock)  is there but not in this image). Its level vary some amount due to  base noise.  (In this image it is more low than typical. This peak level also is different in different individual scopes. It may be even up to near  this low cursor level. Specially in these cases fresh calibration may help and also power cycling "trick" after scope is warmed some minutes. Also there may be small differencies between RuiFeng or other brand ADC and scope manufacturing versions - these are so small things... )

If reader is not familiar with dB scales, it is good to realize what these images show.
50dB difference is roughly 1:320 voltage and 60dB means 1:1000 voltage ratio.
Oh - but it have 8 bit ADC.

[rf-loop]

Here continue
Exactly same settings, signal dropped 10dB more.
Now 107MHz signal level is 34.2uV rms. (-76.3dBm) from HP8644B
Cable M17/84 and scope end  T with 50 ohm terminator.

This signal is clearly detectable over noise floor.

Voltage is now 1/1000 from max level. (-60dB from ref level (1))

250MHz spurious level have now changed (it also vary due to noise floor, and may also change as temp in ADC circuit change. There is perhaps 20 minutes between these pictures due to signal level check with other equipments)

If Owon front end signal pathway is noisy, it really can't do this.
What under 0.5k$ chinese new scope can do this same in these images? I do not know, I have not yet seen any.

(and this is normal factory stock unit)

(1) ref level mean in this case:

Use oscilloscope mode and adjust 1MHz sine signal level so that peak to peak all is visible in TFT area (Owon TFT vertical area is around same as systen full scale)
In this case -16.3dBm signal give this level. (nearly 100mVpp in 50ohm system)
More high level start clipping and produce lot of "false" signals)

Changed to FFT and set upper cursor to this peak.

[lemon]

Thanks rf-loop for all these information.
It is a pitty such hw to have so basic fw.
All we know that the fw is stable but doing only the basics.

At post #1285 we had including our observations about fw and what we want to changed. I have sent this list to Owon (info@...) but I have no any respond until now.
The Owon is the smalest company of all Chinesse companies that make oscilloscope, with good results on hw but they have a weakness to make a good fw.
At the last two year from SDS Series introduction, they've done very little extra work in the FW department (I know they fixed some bugs, of course).
Perharps the department of research and development of fw doing their best, perhaps they know that sold a cheap oscilloscope (under 0.5K$) with a well made of hw but basic fw and they wouldn;t to nothing at all.

I know rf-loop that you have some contact with them often. Do you have any information about how they think to development the fw?

[tinhead]

I did FFT on the csv to check the if there is something visible from the sampling clock, it seems that the 250MHz image frequency is there.
Something is at 21.4MHz, and few third source signals between 75 and 86MHz as well.

i should probably mention that the only reason i asked for csv was to see if there are any spurious when in lower timeabse (that works on most DSOs, here the firmware will not filter anything). Normally the firmware will anyway filter that 250MHz (and in 8/9xxx Owons that will be 500MHz signal) and on the other side it is above the DSO bandwidth, so not an issue at all.

[TomC]

This is a video that was given to me by Owon tech support when they shipped the new PSU for my SDS7102. It shows how to completely teardown an SDS scope in about 5 minutes. I asked them if they could upload to YouTube and they said there was no access to YouTube in China but that it was OK for me to upload it. So it's now in my channel (TomC3106), only video there so far.

The only thing you have to watch for that I don't see in this video is what to do about the power button. So be careful and make sure that you push in the power button while removing the cabinet. On older scopes it was possible to remove the power button cap with pliers beforehand, but in my scope this was not possible because the cap has two plastic protrusions that catch under the cabinet hole. However, on the new power button caps that they sent me these protrusions have been cut off. So it's possible that the latest scopes are shipped with these modified power button caps.



I also asked Owon if there were any plans to implement additional modifications on the Adapter board to further reduce the GND noise. The response is that for now there are no plans for further changes.

[AndrejaKo]

Just a note: The older power caps (mine is from early 2012) also seem to have the protrusions cut off. There are visible marks where they used to be.

Also I wouldn't recommend numerous dis assemblies of the scope, because the screw posts can wear out relatively quickly.

[lemon]

Thanks for sharing TomC,

It's really! The youtube is forbidden to the people of China.

Please look if you have add metadata (or keywords) to youtube metadata search (like: Owon, SDS, 7102, teardown, disassemble e.t.c).
From what I have seen this video isn't easy searchable to youtube.

[TomC]

Thanks for sharing TomC,

It's really! The youtube is forbidden to the people of China.

Please look if you have add metadata (or keywords) to youtube metadata search (like: Owon, SDS, 7102, teardown, disassemble e.t.c).
From what I have seen this video isn't easy searchable to youtube.

Thanks for the tip Lemon,

I added some more tags as you recommended, so hopefully the video is easier to find now.

[mlhstock]

I received the new PSU board. After I opened the case, I noticed my adapter board is different (see the picture below). What should I do?