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

[rf-loop]

There is a new SDS-E series of Owon scopes. But it is not bettter than the original SDS series.

Also SDS-E is NOT more new version for SmartDS series scopes.

This SDS-E is more like upgraded very old PDS serie. And added some more new versions
First there was SDS5032E what was time ago made for replace ancient model PDS5022S. (STN display) Also there was middletime  version PDS5022T (TFT diplay) and later it was replaced with SDS5032E  and now they have made more models for this E-conomy "basic school classroom"  product serie.

SmartDS serie  continue living its life independent of this "E"-conomy" series..
They are different product series. SDS-E   and then SmartDS series.

[Hydrawerk]

 

[TomC]

Today's test involved installing a Belly Strap on TR1. This caused a decrease in the field surrounding TR1, specially around the area where D5 is installed. As far as the noise is concerned there is a visible improvement on the noise observed on the Pwr Sw cover, however, the noise observed on the pseudo Z-plate is around the same.

The Belly Strap was built with HVAC aluminum tape. For the strap to work it must be electrically connected at the joint, so as to form a continuous conductive ring. So I folded the tape at the joint of the top & bottom ends of the tape so that the aluminum side of the top end would contact the aluminum side of the bottom end. Then I placed a piece of tape over the joint to hold the assembly together.

The one negative about this mod is that the current flowing through this strap, which partially cancels TR1's near field, also produces heat. The strap's temperature eventually rises to 50C, the top of TR1 which normally runs around 40C rises to 45C. Naturally this wasted energy will have some impact on the efficiency of the PSU, something that I didn't quantify during this experiment.

At this point I'm not convinced that the moderate reduction in the field out weights the negative side effects of this mod. I need to see if there is significant improvement in the GND noise when the PSU is installed in the scope. On the side by side captures the images on the right are with the Belly Strap installed.

[lemon]

Step by step to the sucess, TomC.
It is very interesting all your findings.

Although there isn't noise improvement on the pseydo-z plate, however we don't know how will be on the real z-plate. Keep walking to your magic psu investigation.

At the end if I remember well, someone (AndrejaKo or you) had used fast diodes on psu and the results were no better than the initial slow diodes.

[AndrejaKo]

Yeah, I tried fast diodes and they didn't help at all. In fact, one specific diode had negative effect, since it prevented battery charging under certain conditions.

[Carrington]

Well done Tom, maybe that the effect is not noticeable in the GND noise, but for sure it reduced EMI.
Both, the use of the small toroidal core for L2 and the Belly Strap are cheap and easy to do.
Unfortunately I believe that little more can be done, because make all the necessary changes is almost impossible. For example how apply a improvement to reduce the EMI in the connector between the "mainboard" and the DC/DC board?

Well I think that still remain something that can be improved easily, replace the DC/DC (see atached image) by a good linear regulator, like LP38853 or LP38513-ADJ (see pag 11 at ADC0xD1520RB_R1_Schematic_07-27-2K12.pdf). Well may not be so easy.

http://www.ti.com/tool/adc08d1520rb

[AndrejaKo]

Can anyone try to confirm this bug:
I had channel one active and channel two disabled and I captured data from channel one in single shot mode. After capturing data, scope went into stop mode, as expected. Then I turned off channel one and turned on channel two. Channel two displayed signal from channel one! Firmware version on my scope is 2.8.1.6 from the newest version of 3.3 patch.

It does same thing if capture is made on CH2 and them we move to CH1.

[lemon]

I confirme the same.
At Run Stop if you activate/deactivate the channels the captured image transfer to other channel.

[bob808]

Hello.
Does anyone have the SDS6062 model? I bought one and it will arrive next week. I am interested in this "noise" issue as it may affect the 60mhz model as well. The one that I bought is 1301 series (first week of 2013?). The seller was kind enough to give me the serial number.

[TomC]

Hello.
Does anyone have the SDS6062 model? I bought one and it will arrive next week. I am interested in this "noise" issue as it may affect the 60mhz model as well. The one that I bought is 1301 series (first week of 2013?). The seller was kind enough to give me the serial number.

The information that we have is that serial 1319xxx is the breakpoint for the low noise mods. However, I seem to remember someone else that bought an SDS6062 prior to that serial and had good results. Perhaps someone else can confirm this?

Edit: This is the post I was thinking about: https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg212002/#msg212002
However, I'm not sure that he was performing the noise test correctly.

[TomC]

Today I tried my Old PSU in the scope. The only alterations at this point are the L2 inductor that I replaced with a toroid and the Belly Strap on TR1. There are no additional bypass capacitors or alterations to the printed circuit.

I started by calibrating the scope and capturing the noise for my New PSU. Note that I had a ferrite on the PSU to adapter cable during this first set of captures. Next, I installed my Old PSU but left the ferrite on the PSU/adapter cable off. For the last set of captures I installed the ferrite on the PSU/adapter cable.

For all three sets of captures the probe cable was looped and sitting on the same surface as the scope. Also note, that as I have indicated before, my probe cables are equipped with ferrites at both ends to limit external interference. While viewing the signals only the CH1 probe was connected to the scope, the USB cable and the Laptop were kept away from the scope and all nearby fluorescent lights were off. To capture each signal I stopped the scope and then connected the USB cable and the Laptop.

On the side by side images the image on the left is for my new PSU (with a ferrite on the PSU to adapter cable), the center is for my old PSU, and the right for my old PSU with the ferrite on the PSU to adapter cable.

[lemon]

Congratulations TomC.
From what I see, the old psu with only two mods (belly strap on TR1 & new L2) is at least identical to new psu.
I don't know what else you are thinking to try about old psu but I am thinking if there is a reason to put the decoupling capacitors.
From the average measurement seems that the ferrite absorbs the noise almost 4mV. From this aspect the decoupling capacitors perhaps to help more.

It is interesting at the sample measurement the capture of old psu with frt hasn't any periodical noise that the new has. I don't know if this is stable all the time like here but if it is true is more important from the improvement of value (26->22mV).

[bob808]

Hello.
Does anyone have the SDS6062 model? I bought one and it will arrive next week. I am interested in this "noise" issue as it may affect the 60mhz model as well. The one that I bought is 1301 series (first week of 2013?). The seller was kind enough to give me the serial number.

The information that we have is that serial 1319xxx is the breakpoint for the low noise mods. However, I seem to remember someone else that bought an SDS6062 prior to that serial and had good results. Perhaps someone else can confirm this?

Edit: This is the post I was thinking about: https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg212002/#msg212002
However, I'm not sure that he was performing the noise test correctly.

Thank you for the info.
I already sent an email to the seller asking him the best route to tackle this issue since I already payed for the scope. On monday he should send it out to me but I might wait for a newer batch since he said that if I need a carrying bag he is finishing another order at Owon next week.
I already saw that user's comments and I also am not sure if he did the test with the proper settings. I sent a pm to him and hope he comes back with good info.
Is there anyone with this 6062 model that cares to share some info? Maybe some pictures with an opened one?
Is the revised psu+adapter card as expected? Something in the range of the other similar prices scopes?
Also I'm planning on using this for audio work so I'll need to look for low levels of noise in smps power supplies. I'm a beginner in this field but I would like to not be limited by this issue further on as I progress.

[TomC]

Congratulations TomC.
From what I see, the old psu with only two mods (belly strap on TR1 & new L2) is at least identical to new psu.
I don't know what else you are thinking to try about old psu but I am thinking if there is a reason to put the decoupling capacitors.
From the average measurement seems that the ferrite absorbs the noise almost 4mV. From this aspect the decoupling capacitors perhaps to help more.

It is interesting at the sample measurement the capture of old psu with frt hasn't any periodical noise that the new has. I don't know if this is stable all the time like here but if it is true is more important from the improvement of value (26->22mV).

Yes, in fact, as you said, the old PSU with the mods works a little better than the new one, and the periodical noise is less. The captures don't show all the dynamics, except for the Avg16 capture that is almost perfectly consistent. But even looking at the signals dynamically it's obvious that the old PSU+ferrite has a little less GND noise than the new PSU+ferrite.

At this point I don't plan to pursue further reductions  in GND noise via mods to the PSU. I feel that the remaining GND noise is inconsequential to measuring low level signals while using the long ground lead. I just wanted to see if it was possible to reduce the GND noise on older PSUs with easily implemented mods. At least to a level where this type of measurements can be accomplished unhindered. I think these mods may be sufficient to accomplish that goal, what remains to be seen is if other members can duplicate this results.

[lemon]

TomC, when I received my new psu, I'll do the same mods like yours.

[lemon]

I would like to find more information about Backlight IC control at the adapter board on version 3.2/3.3

From what we know, it is a chinesse clone of TPS61161A with OBT marking. This same marking OBT uses and TI, too but the image of IC is different with other socket.


The better photo that we have is from member Flash2light here



Have anyone member has any information about this ...rf-loop maybe?

[pullin-gs]

RE: Noise:Why not just get the $50 battery and be done with it??
I only plug it in to charge the battery.
I find it much more convenient to use with a battery also.

[TomC]

RE: Noise:Why not just get the $50 battery and be done with it??
I only plug it in to charge the battery.
I find it much more convenient to use with a battery also.

The battery only takes the place of the 8.4V supply, so it only precludes the noise generated by this part of the PSU which doesn't generate the bulk of the GND noise. Most of the noise comes from the -7.6V supply in the PSU and several other DC/DC converters in the adapter board. So for users with the older PSU and adapter board a battery doesn't resolve the problem. In fact, many of the users involved in the GND noise discussion have the battery option.

[rf-loop]

So much about Owon noise, nearly like it is only Owon special problem and all others do not have.
Here looked Owon, Siglent and Hantek. With as same settings as can. All tests probe comp is adjusted for scope and in all tests exactly same probe.
Vertical 50mV/div with probe setting 10x in scope and in probe. (real scope input is of course then 5mVdiv) Coupling DC.
Horizontal 500us exept Hnatek 400us / div.
Normal sampling, and memory 1M Owon, 2M siglent, 1M Hantek.
Display persistence 5s.

All scopes first image input open, next probe in input but shorted as in pictutre. Then Probe connected to Comp Out GND as see in image. Last probe connected normally to Probe Comp output.

Environment: all switch mode supplies shutted off, computrers off. Only this scope connected to mains what is under test. Even Lights are linear  DC supply powerered LED. Not strong TV, FM or AM stations nearby. Also building WLAN shutted off. 

Owon, Siglent and Hantek all are 100MHz models without modifications.

Just for fun and for realize what is what in this heavy talk about just Owon noise.

[Carrington]

Great job rf-loop. 

Another way to look:

[TomC]

Great job guys!

I really enjoyed the individual pictures and the side by sides of the three scopes being tested under nearly identical conditions. This should also be posted on some of the other threads discussing DSOs.

Maybe this will help dispel the myth that Owon scopes must be noisier than all others since the Owon thread has so much content dealing with this subject.

[lemon]

Realy great job, rf-loop 

You have right that there is wrong impression among the members of this forum about Owon's noise.
It was wrong of Owon, cause wrong initial warranty coverage.
In my country there is a saying that says "better to your eye goes despite the name" ... has to do with the fame.

Anyway, 5-6 members of us keep this thread with our investigation how to eliminated the noise to old noiser SDS7102 not new that all we know that there is no noise, more.

At the start of this month, for the same reason like your thought, I have start a new thread with the title "A method for evaluate your oscilloscopes ground noise?" but there is no results and some members have different aspect about this.
But your comparing results are to the same place and same probe and here is the true!

Congratulation for me!

[Filip Jukic]

I bought SDS7102 few weeks ago and my MATH button is light blue. I saw pictures on the internet and the MATH button is white. Is that sign of new version, i can see rf-loops sds7102 has light blue button too. BTW rf-loop very nice job  .

[tinhead]

Another way to look:

/image/

ehm no, never ever like that. I think rf-loop made images to have an "feeling" how something looks ike on specific model and not to compare directly pictures.  When compare then scale first screens to have same resolution/pix per DIV/data points per pix/intensity full on/etc. Much better is to measure Vpp and/or Vrms and compare results, and even here one should be carefull with buffer depth and methodology of measurment (which didn't matter for these 3 models however).

[lemon]

Your observations have true base and the Vpp or Vrms will be the best for comparative results, but here we have the same methodology, the same position, the same time, the same probe, the same vertical resolution (50mV) the same or very closed horizontal time (400/500usec) and the closer memory buffer 1-2M).
Additional from what I see there is stable optical depth of camera.
I think that they are comparable.

This that is important to see, is if the Carrington's captures have the same resolution and not have zoomed. But in general, if we see the vertical and horizontal spaces are almost idenctical to the three images in series.