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

[Donziboy2]

He does not work for OWON.

He may be a supplier of OWON's but it is not his job to troubleshoot or give fixes to OWON, he does it because he wants to.

If he gets time and feels like doing it im sure we will get pictures of the new boards.

[TomC]

Excuse me if my question probably is stupid:

If with a standard stock owon probe there is this problem, there is on the market a "particular probe" that can reduce this problem without use "a short ground" ?

I' am not an expert but the problem is clear for me but I serching for a solution, if exist, a "short ground" is a big compromise is not a "pratical solution".

Other scope don't need "a short ground" becouse the difference is near zero.

I hope exist an alternative  ...

If someone have some alternative , pratical solution , (not other informations about the problem that is clear for me), please share

The short answer to your question, at least in my case, I've seen some improvement with a different passive probe while still using a 6" ground wire. This is possible if the probe's input capacitance is lower than the stock Owon's probe. With the stock Owon probe my SD7102 displays 100mV pp of noise in Average 16 acquire mode. The noise generated by my scope is mostly around 100MHz as I have verified with the FFT display. I checked the input capacitance of my probe and the inductance of my 6.5" ground wire with an LC meter and got 15.2 pF and 141nH. The resonance is then around 109MHz which leads me to believe that the noise I see is exaggerated by the Owon stock probe ground wire combination due to its resonant point.

I then tried a Tek P6061 which I had around which has 9.5 pF input capacitance and uses a 6.2" ground wire which measured 125nH. In this case the resonance is around 146MHz and although the ground wire is almost as long as the Owon's the noise displayed by my scope went down to 75mV pp. Note that the P6061 is rated at 60MHz, however, I checked it against the Owon's stock probe and its response up to 110MHz was nearly identical.

I've also tried using a snap on small ferrite core on the Owon's probe stock ground wire. This also causes the noise displayed by my scope to go down to 75mV pp. I calculated the resonant point for this combination and it came out to be around 39MHz. However, this solution will cause exaggeration if you measure signals in that frequency range or if some other device emits common mode noise in that range. I happen to have a laptop PS which emits noise in that range and when I plugged it in the same outlet I was using for the Owon the noise went of the screen.

The best results I've had so far is using a custom made 3" long ground wire with the Owon stock probe or a stock 4.25" ground wire with the P6061. In both cases the  noise displayed by my scope goes down to 45mV pp.

What I've learned from all this is that when working with high frequency signals or fast rising signals, the shorter the ground wire the better accuracy you will obtain. In these cases an effort to use the spring ground clip is well justified for me. A long ground wire is tolerable for low frequency work and slow rising signals, and if the Owon's noise is getting in the way because the signals are small, I either use the 20MHz bandwidth limit or the X1 probe setting (6MHz bandwidth limit) depending on the frequencies I'm interested in. If I'm looking at recurring signals I often use the Acquire average mode to get a cleaner more stable display.

[Armageddon]

I have reviewed the mainboard of this oscilloscope, and I found some atrocities. 

The A/D (ADC08D1010 on my oscilloscope) is powered by a TPS4022, entirely. OWON engineers are a little beast! But there is more.

The ADC08D1010 datasheet says (Pag.33): "The Va and Vdr supply pins should be isolated from each other to prevent any digital noise from being coupled into the analog portions of the ADC. A ferrite choke, such as the JW Miller FB20009-3B, is recommended between these supply lines when a common source is used for them"

Q: Where is the ferrite?
A: Very simply, does not exist. Why spend money on this? LOL...

Probably -7.6V at the PSU generated with a MC34063 switching at 100KHz, is used to power a portion of the input stage. And to do this before goes through a 79M05A (RR ~40dB @ 100KHz), ie in all probability the input stage is using +-5V. With this configuration -5V should have a good amount of noise, but surely +5V plane/line have more noise even.

I not know the new changes introduced by OWON, but unless OWON made a miracle I can not believe the results shown in the previous images. This problem has been corrected? Only with a new PSU and DC/DC board?

OWON engineers are as good as their customer service 

[rf-loop]

Hello.

I have a SDS8102V and, today I made the "test of noise" using the description of "TomC".

With a 200MHz probe (P6200) I get:
  -Peak:       Vp=24mV, Vk=9.2mV
  -Average:  Vp = 14mV, Vk = 7.8mV

With a 100MHz probe (T5100):
  -Peak:       Vp=19mV, Vk=8.5mV
  -Average:  Vp=12mV, Vk=6.5mV

I suppose that I don´t have the noise problem. What do you think?

Without probes I have measured a level noise of ~1 mV (BW = 20MHz) [x1] and ~2 mV (Full Band) [x1] with or without 50 ohm load.

I hope that these data be useful. 

Best regards.

Default scope.
Run scope half hour. Do selfcal.
Push CH1 to vertical center.
set trigger level somewhere out from signal, example to + or - 20mV (there is autotrig on)
Acquire mode Peak
Memory lenght 1M.
CH1   5mV/div  and DC, and BW limit off.
CH1   BNC covered with 50ohm terminator without center pin connection. (no matter if you think no effect or some, just do it)
(CH2 off)
Turn horizontal speed to 1ms/div
Turn Measure on and Pk-Pk CH1

Show copy from screen.

[Armageddon]

Hi!
According to your instructions here is the result, with 50ohm terminator on CH1: 

[rf-loop]

Hi!
According to your instructions here is the result, with 50ohm terminator on CH1: 

This result is quite good. It tells that analog front end itself including ADC noise level is ok.
But just small amount more that typically in SDS7102, perhaps due to slower adc 7102 have some less noise but also this kind of difference may be just differencies between individual units.

Of course it can be better if really start do all things for lowest possible noise without changing ADC around of whole front end circuits.  But, my opinion is that now there is all ok.
Take some other oscilloscope... Agilent, Tektronix, and stay in 100MHz models and not any "high end" or middle end model... I believe that result is around same or worse.
After sad "CH1" noise issue in time ago Owon front end have been good if look noise level.

[Armageddon]

Tell the same to all who have problems with noise. 

[rf-loop]

Tell the same to all who have problems with noise. 

I have told many many times  it.

Noise problem is NOT in signal pathway inside oscilloscope after BNC input.

But now there just before you try tell that there is ADC on board and power linbes and this and that...  your picture show exatly that problem is NOT there.

This noise GND problem is exactly in adapter board and PSU SMPS circuits. And I can proof it without any holes. Other but much lower and also not SO serious noise source is TFT databus. Also This is inspected and proofed.

If I connect clean signal to you scope input, it is clean and it stay clean. If I connect clean signal to nousy GND unit there can see just barely small extra noise in some very selected tests. Also this is inspected and this can test in any lab where is real enough good tools for this.

It is also inspected that what is source for single noise component in whole total noise "chaos".
I have even told how this regognization can do very easy.

It is not question anymore that what is this noise and where from it come. This is fact, not mutual.

So, what is your point with this ADC talking related to this noise problem?

[TomC]

Tell the same to all who have problems with noise. 

I have to agree with rf-loop on this, the noise I see on my SDS7102 is common mode noise converted to differential noise by the 6.5" ground lead. As evidenced by the test you just did, my scope also has a clean signal path, maybe just a little bit better than yours. Excessive noise in the signal path would be a deal breaker, for example, noise injected via the ADC as you imply. But this is just not the case with the current GND noise situation. I view this problem as an inconvenience not a deal breaker. That said, I would be very happy if Owon decided to bite the bullet and offer to fix or provide upgraded boards to those of us that have to put up with this inconvenience.

[Armageddon]

This noise GND problem is exactly in adapter board and PSU SMPS circuits. Yes, and the rest of the design also contributes (bad design). Probably common mode noise is converted to differential noise by the 6.5" ground lead.

One moment I have the solution: using a ground lead of 0.00001". 
Do you think we're idiots?

A message to OWON: "Fool me once, shame on you. Fool me twice, shame on me."

You dont remember the beginning? See Attach.
OWON has improved this oscilloscope at expense of all that have purchased it.

Quote from: TomC
"That said, I would be very happy if Owon decided to bite the bullet and offer to fix or provide upgraded boards to those of us that have to put up with this inconvenience."  Totally agree. 

Q.C. passed 

[rf-loop]

Can you tell us what was year you buy this old version or did you buy second hand, perhaps oldest versio or nearly  prototype?
2010 or? 

All what is inside have changed many times if compare to this old version what can see in pictures.
What is same today and example last year is TFT panel. There the similarities with this old version.

OWON has improved this oscilloscope at expense of all that have purchased it.

Do you know how much money I have send to Hewlett-Packard they can develop they some mistakes and develop better machines. And same for Mercedes Benz or even Seat..  they have sold crap to me and later they have also, using my and others money, developed some bad things better.  This is how this train go forward - I do not want stop it.

But now, you can think what is Mr Henry. I can tell you... it is not new soft drink brand and it is not new candy.  But he do some magic tricks if it meet fast changes...fun guy.

No one have told that making Mr Henry more small it solve this primary problem. But it may reduce its effect in measurement. Also it do not mean that with this trick do not need repair primary problem. 

 

[TomC]

But now, you can think what is Mr Henry. I can tell you... it is not new soft drink brand and it is not new candy.  But he do some magic tricks if it meet fast changes...fun guy.

 

Ch1 is connected to the signal source via a probe test point, Ch2 is connected to the same source using the 6.5" ground lead.

[lemon]

Hi, everybody.

I think so that the fan of my SDS7102 has gone.
I can't hear it (and I don't remember if it is hearing in past) and the front part of oscilloscope is warming!

How can I see, if the fan is working, without open it?

(Sorry, for my english...)

[TomC]

Hi, everybody.

I think so that the fan of my SDS7102 has gone.
I can't hear it (and I don't remember if it is hearing in past) and the front part of oscilloscope is warming!

How can I see, if the fan is working, without open it?

(Sorry, for my english...)

Place a piece of tissue over the large vent on the back of the scope, you should see it gently vibrate if the fan is working.

[lemon]

Thanks for the reply.
There is no any vibration of paper or something light tissue!
It is explained now why this reach over 45°C in front of case!

[Hydrawerk]

(My Agilent DSOX2000 stays cool all the time...  )

[digsys]

I always use the pointy end of a small cable tie .. you can tell what rate it's spinning by the clacking,
just like we used to put on bicycle wheels :-)

[luca1000]

The short answer to your question, at least in my case, I've seen some improvement with a different passive probe while still using a 6" ground wire. This is possible if the probe's input capacitance is lower than the stock Owon's probe. With the stock Owon probe my SD7102 displays 100mV pp of noise in Average 16 acquire mode. The noise generated by my scope is mostly around 100MHz as I have verified with the FFT display. I checked the input capacitance of my probe and the inductance of my 6.5" ground wire with an LC meter and got 15.2 pF and 141nH. The resonance is then around 109MHz which leads me to believe that the noise I see is exaggerated by the Owon stock probe ground wire combination due to its resonant point.

I then tried a Tek P6061 which I had around which has 9.5 pF input capacitance and uses a 6.2" ground wire which measured 125nH. In this case the resonance is around 146MHz and although the ground wire is almost as long as the Owon's the noise displayed by my scope went down to 75mV pp. Note that the P6061 is rated at 60MHz, however, I checked it against the Owon's stock probe and its response up to 110MHz was nearly identical.

I've also tried using a snap on small ferrite core on the Owon's probe stock ground wire. This also causes the noise displayed by my scope to go down to 75mV pp. I calculated the resonant point for this combination and it came out to be around 39MHz. However, this solution will cause exaggeration if you measure signals in that frequency range or if some other device emits common mode noise in that range. I happen to have a laptop PS which emits noise in that range and when I plugged it in the same outlet I was using for the Owon the noise went of the screen.

The best results I've had so far is using a custom made 3" long ground wire with the Owon stock probe or a stock 4.25" ground wire with the P6061. In both cases the  noise displayed by my scope goes down to 45mV pp.

What I've learned from all this is that when working with high frequency signals or fast rising signals, the shorter the ground wire the better accuracy you will obtain. In these cases an effort to use the spring ground clip is well justified for me. A long ground wire is tolerable for low frequency work and slow rising signals, and if the Owon's noise is getting in the way because the signals are small, I either use the 20MHz bandwidth limit or the X1 probe setting (6MHz bandwidth limit) depending on the frequencies I'm interested in. If I'm looking at recurring signals I often use the Acquire average mode to get a cleaner more stable display.

What you think about these probes ?

Are better then the standard Owon Probes (and reduce the noise) ?

What are the best of these probes ?

[luca1000]

I have forget the other attach.

[rstoer]

What you think about these probes ?
Are better then the standard Owon Probes (and reduce the noise) ?
What are the best of these probes ?

Probe Master's oscilloscope probes are much better than OWON's or any you'll get with most Chinese scopes. I have an OWON scope and a Probe Master probe, so I've seen both side-by-side and there's no comparison. However, a better probe used with the typical 6" ground lead will NOT reduce noise. If anything it's wider bandwidth might pickup MORE noise. Better construction, less circuit loading, and more options for shortening your grounds (I.E... 3" ground leads, HF ground clips, BNC adapters, etc) are the advantages of a better probe.
Lastly, I can't find a PM2817 probe listed on Probe Master's website so I assume it discontinued or bogus and you're seeing it sold by some third party. If your going to buy a Probe Master probe I'd deal directly with them to be sure of what you're getting.

[TomC]

I have forget the other attach.

From the specs you posted, it appears that these probes have a lower input capacitance than the Owon's stock probes. Based on my experiments, I would expect them to reduce the amount of common mode noise produced by the Owon's SMPS circuits that gets converted to differential mode noise. This is because the resonant frequency with a similar 6" ground wire will be higher and at least it won't exaggerate as much the Owon's 100MHz noise. But remember, I have only tested this with one scope, mine, which mostly produces noise in the 100MHz range. Since you don't have a scope yet, there is no sure way of telling whether you'll even have a noise problem or whether the bulk of the noise will be in the same frequency range. However, should you decide to buy these probes anyway, as far as quality, they will surely beat the stock Owon probes.

Another thing to keep in mind is that a 6" ground wire doesn't provide accurate readings for high frequencies and fast rising signals. This is true regardless of whether the scope produces noise like the Owon or is squeaky clean. So the 6" ground wire should only be used for low frequency work with any scope. If you plan to stick to these rule, you can dramatically reduce the noise from the Owon when using the 6" ground lead by turning on the 20MHz bandwidth limit. If you are working with even lower frequencies and small signals, you can also set the stock Owon probe to X1, which limits the bandwidth to about 6MHz.

[lemon]

Except the Owon probes, I have a Hameg HZ-36 Probe with 12/57 capacitance and another germany one (no name) with 11/46 capacitance (1:1/1:10 both).
Neither one has a better behaviour on gnd noise than Owon stock. For my probes, as the capacitance is lower as worst be the gnd noise!

[TomC]

Except the Owon probes, I have a Hameg HZ-36 Probe with 12/57 capacitance and another germany one (no name) with 11/46 capacitance (1:1/1:10 both).
Neither one has a better behaviour on gnd noise than Owon stock. For my probes, as the capacitance is lower as worst be the gnd noise!

I'm curious, have you used the FFT to check the frequency range of your Owon's noise. On mine the bulk of the noise is around 100MHz. Given your results, I suspect the bulk of the noise on your scope is at a higher frequency.

[lemon]

Here are two captures, one of them is "gnd noise" and the other the same time what is appearing the FFT...almost 100MHz.
The test was done by Owon probe.

[TomC]

Here are two captures, one of them is "gnd noise" and the other the same time what is appearing the FFT...almost 100MHz.
The test was done by Owon probe.

Thanks a lot for the images. From your FFT image I can see that the noise on your scope is a little different than mine. Yours has almost even noise from about 80MHz to 140MHz. That explains to me why you don't see much difference using lower input capacitance probes. Here are some similar images from my scope using both the stock Owon probe and a 9.5pF input capacitance Tek probe. As you can see the noise improves with the lower capacitance probe, but the noise in my scope peaks at 100MHz unlike yours.