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

[yuyu11uu111]

About Reading MAINS

my scope arrived with 2 pin mains cable, does it mean that with this cable i can test mains? if not, how can i test mains without having any risk and without using fancy equipment like special probes?

thank's

[Hydrawerk]

Always be careful. Your scope should be grounded when powered by AC 230V or 110V. Replace your power cord. And there is other rule. When working with high voltage use only your one hand, so the risk of electric shock is reduced.

[rf-loop]

... how badly does this affect normal signals you are checking on the scope?

This is really good question! 

Does it make small 100-200mVpp signals unreadable, or just butt ugly?

No and yes.  It depends how user use probes or how user connect signal to the oscilloscope input.
And it depends what you are doing.


This claim I can repeat agen becouse this can be some kind of starting point for thinking:

There is oscilloscope input BNC connector. From this point to inside and finally visible waveform on the screen there is no problem. In this oscolloscope class this is cleanest what I have seen. Even if oscilloscope individual is this "GND noise issue" unit or usual "normal" unit. I can compare also these.
Analog front end and ADC is good, and extremely good in this class. If someone have better, welcome with it to my workshop and let's make test for compare in real world with equal tests.

But then there is this dirty side:
First, if you do something very sensitive and you need turn off all swithc mode lights and computer and phone and energy saving lights and if HF/VHF/UHF radio/TV stations RF fields do problems  it is best you forget this Owon and specially unit what have this extra "noisy GND" issue. Becouse it radiate RF from its several SMPS circuits and then other things like TFT is also radiating EM noise but this is not Owon alone.

If you need "EM black room" you need absolutely forget this Owon (good and bad individuals, all) and also nearly all other oscilloscopes exept some very special equipments. I have example one RF generator what can take to "black room" (HP8642B with low EMI options. It is so heavy there read it need two person for carry)
-------

Owon made some changes to main power unit and adaptr board.

After these changes this noise level rise.
This total noise include spread spectrum and from these noise sources inside scope it travel by radiating and conducting and outside of scope it can say it is "common mode" noise. Somewhere near circuits it is differential mode also - of course.

It radiate and it may this way pollute your signal or it can travel also using your probe cable and pollute signal.  Common mode noise things are very complex and it is not easy understand how they affect. There is so many things and also now this is really spread spectrum noise. Some frequency components do other things and some other, they are like group of different animals.
It looks fynny, if you move your probe cable, if you touch it with hand, if you move your device under test... all may do some change what you see.

Now there is several things what can do for reduce this problem.  I do not explain here now these exept...small amount.  But it need somehow understand it.  You can not see anything about this noise as long as it stay common mode. If you see it, you have somehow changed it to differential mode or piced up it as differential mode.

First test you can really see how it works and then think "why":

Take your probe and connect it 10x and oscilloscope 50mV/div(scope setting 10x for probe but it do not mean anything if you forget it 1x... then just 5mV but change it in your mind)

Connect probe GND wire to Probe comp 1kHz output GND. Connect probe center tip to same point.
Turn oscilloscope for quite slow horizontal speed but highest possible samplerate (using max memory) OR turn acquire mode to "Peak".
(in peak mode it use always full samplerate for find highest and lowest sample in one sample "group" (explanation, if displayed samplerate is 1Msa/s and there is 1GSa/s samplerate it means it drop out 999 and keep one. In my text this "group" meant this 1000 sample. But peak mode look these samples for find  "min-max"  )

Then do same but open your probe pouch and find small metal spring. Take probe tip hook away. Insert this spring over tip GND area near center tip. So that all spring turns are over this (for avoid extra inductance).  Now, connect your probe GND and center tip agen to probe comp. out GND. Where is noise? Why?

Ok lets make other test (and remember, probe comp out is NOT clean, it is only made for adjust probe compensation and there HF noise is not problem.

Connect probe tip to probe comp. signal and GND spring spike to probe comp GND.
Do not connect probe black GND wire.  First look 1Vpp and then turn (scope input DC) even to 100mVdiv or 50mVdiv. Look signal bottom. Adjust trig and vertical shift if need. Noisy? Very very noisy?


Then, take GND spring away and just same using probe black GND wire. Noisy? More than just before?


Why there is so big difference and only with this quite small difference in probing method?
It is good to understand least some amount.   
(these are also important things for probe use with high frequancies/fast risetimes signals, always, not only this case, even with best possible oscilloscopes.)


I have made tests using one "noisy GND" individual factory original unit and  units what I have last get directly from factory.
Just due to lack of  enough free working hours these all results images are just raw "in the  basket all together" and then lab hand write paper notebook explanations. It need work before they are ready for publish. And then, if I show these. I'm biased seller - so....  I do not know if I do it here or not. (90% of my customers are local and they can come visit and see with they own eyes if they want)

So, answer for this

Does it make small 100-200mVpp signals unreadable, or just butt ugly?

No and yes, depends how you do and what are you doing.

[yuyu11uu111]

Always be careful. Your scope should be grounded when powered by AC 230V or 110V. Replace your power cord. And there is other rule. When working with high voltage use only your one hand, so the risk of electric shock is reduced.

thank's

i will replace my cable (too bad this is the cable arrived from factory with the scope), why is it important to ground the scope? also, so can i test mains without any special equipment? i would love to get more details

thank you

[Donziboy2]

Thanks for the info rf-loop.
Could I impose on you to take 4 more screens.  First using the same settings on the previous page of this discussion, but connect so you can see the 5v 1Khz signal instead of going to ground and zoom in on the top and bottom portions of the signal using AC and then switch to DC coupling.  I would like to see how much noise is present on the portions of the square wave that should be relatively flat. 
This will give me some idea of how bad it will be.  I am most likely going to get an SDS7201V since it seems to have the features I am looking for.  From what I understand at the worst I could mod it to fix the ground noise issue.  For the price its tough to be picky.

[rf-loop]

Not what was asked but these are from allready made tests.
This time I do not have time for extra tests and document these etc.

I hope pictures are self explanating. (use file name and text and information in pictures.

NoiseBad oscilloscope is SDS7102V original from factory and time ago when this issue come on table.

NoiseNormal: Oscilloscope is SDS7102V original from factory after they have worked for this noise issue. (these scopes arrive from factory 15.03.2013 to me)

In test you can also see how big difference is with probe "long wire" GND or with short GND spring.
(it is becouse this long GND wire have "big" inductance - something between 150-200nH and not only in GND wire, also inductance between this point where it is connected to this point where probe tip is connected. You can see "noise issue" oscilloscope results with better GND practice)

[yuyu11uu111]

nice post rf-loop, very informative, thank you

[electroguy]

rf-loop, I wish i could get mine to look as good as your 2n picture, let alone your 2a!
This is mine, and owon said that this is normal
do you think your mod that you mentioned earlier to the SMPS would fix this?

[rf-loop]

rf-loop, I wish i could get mine to look as good as your 2n picture, let alone your 2a!
This is mine, and owon said that this is normal
do you think your mod that you mentioned earlier to the SMPS would fix this?

Please do some more tests.


1. Do exatly same as your picture is exept: run with battery power. (sidepanel switch off or AC power cable disconnected)

2. Agen exatly same but take whole battery out from oscilloscope and  AC power connected.

3. Do exactly same test as  yesterday 11:31:56 PM message but connect probe using GND spring. (battery assembled and with AC power and battery charging)

[electroguy]

1. Do exatly same as your picture is exept: run with battery power. (sidepanel switch off or AC power cable disconnected)

The picture in my previous post was with battery only.

2. Agen exatly same but take whole battery out from oscilloscope and  AC power connected.

Wow, it works better without the battery!!!
See 2 pictures attached with no battery (one with ground spring the other with long clip).

3. Do exactly same test as  yesterday 11:31:56 PM message but connect probe using GND spring. (battery assembled and with AC power and battery charging)

See 3rd photo.

[Donziboy2]

Thanks rf-loop.

Electroguy, how old is your SDS?  NM found it.

[EIS]

What do the diagonal lines mean in the attached photo?

Thanks...
Earl

[TomC]

What do the diagonal lines mean in the attached photo?

Thanks...
Earl

The way I understand it, the diagonal lines are just filler indicating that there are samples in that area that are not being displayed. In peak detect mode the scope always uses the maximum sampling rate, but only the highest and lowest samples are displayed. With the time base set to 500 micro secs, if my math is right, you get about 500,000 samples per division. However the screen only has 800 x 600 pixels (1 division is about 53 x 60 pixels). The beauty about the Owon's deep memory is that although all the samples are not displayed, they are stored in memory. So if you press Stop, you can change the time base to say 10 micro secs, and see many of the samples that were not visible before. As you do this you'll see the area with the diagonal lines shrink because there are fewer samples that are not being displayed.

[Armageddon]

Any news about the new PSU and adapter? 

[sigsegv11]

Well, here's my scope (my first scope!) and I'm not too happy, though I don't know enough to understand when this sort of error will creep into measurements and affect me. I received the scope from Saelig about a week ago. I think I'm going to ask to either return it or see if they'll fix it. I really like the scope except for the uncertainty of whether this issue will ever be a problem for me. I'm just a hobbyist and still learning and $400 makes this the most expensive electronics tool I'll own for the foreseeable future. I don't want to spend it on a tool that is clearly flawed, especially in a way that many units of the same type are not.

I took screenshots of the noise both when measuring the 1KHz square wave (which works out to about 50mV of noise in a 5v signal) and the noise when the probe is grounded, which is about 200mV peak to peak.

[luca1000]

Can you please share your serial number of your unit ?

[TomC]

Well, here's my scope (my first scope!) and I'm not too happy, though I don't know enough to understand when this sort of error will creep into measurements and affect me. I received the scope from Saelig about a week ago. I think I'm going to ask to either return it or see if they'll fix it. I really like the scope except for the uncertainty of whether this issue will ever be a problem for me. I'm just a hobbyist and still learning and $400 makes this the most expensive electronics tool I'll own for the foreseeable future. I don't want to spend it on a tool that is clearly flawed, especially in a way that many units of the same type are not.

I took screenshots of the noise both when measuring the 1KHz square wave (which works out to about 50mV of noise in a 5v signal) and the noise when the probe is grounded, which is about 200mV peak to peak.

Did you do a Self Calibration before taking the screen shots? I've found that this influences the amount of noise I see in my SDS7102. You may also want to try using the spring ground clip instead of the 6.5" ground wire. If the noise that you are seeing is common mode noise from the Owon's SMPS circuits converted to differential noise by the 6.5 ground lead, using the spring ground clip should decrease the noise dramatically. If it doesn't you may have a malfunction in the signal path that needs to be repaired.

[luca1000]

Have you tested the channel 2 ?

On this unofficial document (I attached only the first page) the problem seem appair only on channel 1 and NOT on channel 2 ...

[Hydrawerk]

Oh, it's like everyone should rather buy Rigol than this noisy Owon... 

[sigsegv11]

The noise is the same on channels 1 and 2. My screenshots were made after calibration.

[Armageddon]

To "rf-loop" answer this pelase: 

   a). What noise level is right for OWON?
   b). Will send Owon spare parts to all with this problem?
   c). What items (s/n) are affected? And do not tell me that OWON do not know.

rf-loop reply:

a)
Noise levels what are here in images I can accept.
here:
https://www.eevblog.com/forum/testgear/sds7102-old-stock-serial-identification-(with-noise-issue)/msg231565/#msg231565

Also here: tested "normal" and "noisy" SDS7102.


https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg226398/#msg226398

This "normal" I can accept. This "noisy", this level I can not accept, but still also in many cases with it can live IF do not need measure very low level signals just with using probe long GND wire. It need also note that this "noisy" scope conducted and radiated RF noise may disturb some sensitive circuits what you are testing.

I can not write rules what noise is accepted and what is limit what can not accept. Also before this need
If someone want define the limit values there ??should first create a definition for it. Who determines what and how to measure.

b) I can not answer this. First people need ask seller. If seller do not want help and or solve this problem then customer need ask Owon. Only Owon make decisions how they handle individual situations.

c) I do not know. Also I have never told that Owon do not know. But, I do not know what Owon know and what not. Who want know he can send fax, call phone or send letter to Owon and ask Then, after get answer from Owon,  he know what Owon want tell. Also in this case you do not know what Owon know, you know only what Owon tell to you. If they give some public information it is as it is. If they give some personal answer it is only just it and nothing more.



-----------------
This is beginning to resemble the ring.
Nature of the problem is quite well aware of when you read and studied here in the forum what all peoples have tested and discussed over long time in many many threads about noise overall and about this case.

Universal noise level threshold value at which the device should reject or accept you're not going to get anywhere.

I thought so, ie we are going to anywhere. Or even more explicit: Probably OWON will do nothing for you if you have a machine with this problem, under warranty or not.   

[rf-loop]

Here some add for this message:

https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg226398/#msg226398

These images are really bad and also in this case whole measurement do not meet any kind of acceptable lab test "good practices".

There Owon probe connected to oscilloscope GND. Also probe GND wire is connected to same GND point. (probe comp gnd)
Probe is 10x for better BW.
Probe is connected to spectrum analyzer 50ohm input. Every people understand that level itself is now "un known". There is high attenuation due to this connection.
Note: in all three images connection is same.

1. Both scopes OFF.
Scope is just as antenna.

2. Probe connected to "normal" Owon and alone it is power ON and running for avoid radiated noise from noisy one.

3. Probe connected exactly same way to noisy Owon and now other is OFF.

Spectrum analyzer adjustments are not optimal, level is undefined but pictures are comparable with each others.
(sorry I did not find time for play with HP-IB for pick-up screen capture from this OLD workhorse HP8568B and also no time for find some camera stand but these images show something about this noise. Note also these commercial broadcasts as example FM stations so they are also in all images and you may find some idea what is level of this Owon GND noise.  Nearest FM stations are over 10km distance. There can see also some other RF signals (look 1. image). If there is nothing, it is just flat noise floor.

[Armageddon]

rf-loop, can you share images of the new boards? 

Thanks

[rf-loop]

rf-loop, can you share images of the new boards? 

Thanks

Of coiurse I can do it as also many other things I can do... but still I do not many things what I can..

[Armageddon]

rf-loop, can you share images of the new boards? 

Thanks

Of coiurse I can do it as also many other things I can do... but still I do not many things what I can..

But you know what. The best that you're doing is not worry about all existing customers!