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

[rf-loop]

2. waveform refresh rate improved
- Seems to be some improvements cause changes at Autoset.
For that reason I repeat the test of Mark Madel (Owon SDS7102 review. youtube) with the trigger output capturing at Chanel II.
The Mark has found that this DSO varies the refresh rate depending time base & memory depth. He had recorded 22.96fps at 1MHz sinus signal (1K depth memory, 1us time base) that dropped to 16.43 when the memory depth was 10M.
With the same conditions, I repeat the test and the results seem to be very faster. With 1K depth memory the refresh rate was 119.07 and with 10M was 36.95. Nice to be good!
Please look at my attachment to verify.

This old test give a little indication that the speed can increase, that there is room for improve wfms/s speed some amount but of course it is still conventional  DSO, not DPO.
Nice that they have finded some solution for add speed by FW.

For rare peaks etc it is more fast than may think if look only continuous wfms/s speed.
It can rearm trigger faster and do one capture to buffer. But it can not of course repeat this cintinuously. But, for rare extra glitch this may find it more fast that may assume if look only continuous average wfms/s speed.

Here is this old test agen:

(http://owon.forumup.com/viewtopic.php?t=183&mforum=owon)



I can see also nice new work and thinking about noise.

[jaxbird]

Haven't read the whole thread  (sorry)  .. Any quick links to the firmware update?

Mine is SDS71021303xxx, I guess that means I should benefit from the latest firmware improvements.

Any high risk of bricking or anything I should be aware of?

(haven't done any updates yet as I've had bad experiences with firmware updates on other hardware, requiring days of poking around with JTAG to recover)

[TomC]

Haven't read the whole thread  (sorry)  .. Any quick links to the firmware update?

Mine is SDS71021303xxx, I guess that means I should benefit from the latest firmware improvements.

Any high risk of bricking or anything I should be aware of?

(haven't done any updates yet as I've had bad experiences with firmware updates on other hardware, requiring days of poking around with JTAG to recover)

Link to Owon is below, click on Download Center:

http://www.owon.com.hk/main.asp

Done several upgrades on my unit without any problems.

[TomC]

Hi lemon,

Thanks for the excellent analysis of the new firmware features!

Doing something like that was next in my agenda, but you got it covered!

[TomC]

I can see also nice new work and thinking about noise.

Hi rf-loop,
Nice to to see you active on this thread again!

Any chance you can do some testing on the below when your schedule permits?

In theory ferrites on the probe cable act as a common mode choke and shouldn't affect the differential signal at the probe's input, I've been able to prove this to my satisfaction to 110MHz (the max of my RF generator), but beyond that I can't be certain. I'd like to see someone with more sophisticated equipment test this further.

[TomC]

Arrghh! Some of those features would have been useful to me. Too bad Owon decided to replace my 1211 main board with a 1209 one when I sent it for repairs.  After installing the update, I got firmware version 2.8.1.6 and check-sum o B0BB1638, same as before, so no new features for those with older units.

Hi AndrejaKo,
Nice to hear from you again!

Have you tried writing to Owon about this?

I think their warranty states that they will replace defective parts with equal or equivalent parts. Using an older board that is no longer upgradeable doesn't seem to match that promise!

[Sarasir]

Hi TomC,

just as a first reply to your long post above:

Yes I noticed that there where ouside sources of noise as well, e.g. my new LED bulbs (changed all my illumination to LED recently - it's great but would be worth it's own thred as there are different qualities on the market and some with and some without switching converters).
I minimized that by putting ferrite cores to the mains cables of the lamps as near to the bulbs as possible and not just one, 3 on each. (Fortunately I had a box full of that cutted cores types which are  inside of a plastic enclosure that you can open and close again.)

I think it's best practise is to have an emission free working desk. 

As to the terms Common mode and differential mode concerning Owons input I'd like to mention that it's not a perfect common mode input at all (at Owons  BNCs). Only if you had a very strong generator with a floating output with zero impedance over the whole spectrum connected without any wire in between to the BNCs you cold assume a perfect -/+ signal fed to the input. And by using another scope with differential input you could get an idea what the Owon does to this signal common modewise by it' s emission and impedances. And if the characteristics of the generator's output (and a probe connected to the input is a generator as well if connected to a signal source) are changed various effects will be experienced. The most important to me is, that dependent on the generators output characteristics part of Owons emission will be changed in phase and level and in a way sent back to the input's center connector and then and only then of course will be displayed on it's screen.

And your cores no matter why you put them on are part of the game.

So far for now and until later...

 

[TomC]

As to the terms Common mode and differential mode concerning Owons input I'd like to mention that it's not a perfect common differential mode input at all (at Owons  BNCs). Only if you had a very strong generator with a floating output with zero impedance over the whole spectrum connected without any wire in between to the BNCs you cold assume a perfect -/+ signal fed to the input. And by using another scope with differential input you could get an idea what the Owon does to this signal common modewise by it' s emission and impedances. And if the characteristics of the generator's output (and a probe connected to the input is a generator as well if connected to a signal source) are changed various effects will be experienced. The most important to me is, that dependent on the generators output characteristics part of Owons emission will be changed in phase and level and in a way sent back to the input's center connector and then and only then of course will be displayed on it's screen.

And your cores no matter why you put them on are part of the game.

So far for now and until later...

Yes, I agree, everything is inter-related. Just like resistors are also inductors and capacitors in some ways, and transformers also act as capacitors, etc., etc. But in this imperfect world everything has a primary purpose!

Cheers!

PS. I thought you meant differential instead of common in your post, correct me if I'm wrong.

[TomC]

Just got my curve tracer working with the SDS7102 in XY mode. I made a special interface cable with adjustable voltage dividers so I could perform the calibration procedure. This of course wasn't necessary with the Tek CRO because I could use the variable controls for that function. The first two attachments are captured images from the SDS7102, on the first one persist is turned off, which was the only option before the latest firmware release, on the second one persist is turned on. The third image is an older photograph of the Tek CRO I've been using with my curve tracer until now.

The persist option as implemented for XY mode is not perfect, the time settings don't seem to have any effect, and regardless which one you select, it seems to operate as if infinite had been selected. But a least I can obtain a usable image if I want to keep a record of curves for particular components. So I think I can now retire my Tek CRO from this job.

[Sarasir]

Yes TomC,

I think indeed somehow I missed the point. I was in a hurry and did not think over what I wrote once more, and by the way, the grass is blue and the sky is green....

But seriously:

I discovered the noise problem by connecting a Tek SG505, a very low distortion sine generator for audio tests with a high level floating output to the Owon after having checked it's frequency response with a Tek SG503 (A sine generator up to 250MHz) and a SW503 (A sweep generator up to 500MHz).

With both the HF generators the noise was not disturbing, I can't even remember if I noticed it, but with the floating output generator it was terrible (filled almost the screen at 2mV/div input position.

I said to myself - sorry - fuck what's happening here and started to do tests and tried to build up a picture of what happens inside and around that Owon.

I don't know actually what I meant in my last post differential-, common mode or both. Now I see that it did not fit together at all and from your view your correction is of course right.

Let me put it another way and please tell me what you think:

1. We know that there is noise generated inside of the Owon that is also present at the ground contacts (BNC sleeve).

2. The noise must be also present at the inner contact of the BNC other wise we would see it on the screen, and we don't (nothing connected yet).

3. So, if I am right therefor this must be common mode noise provided you look at the sleeve as minus and at the center contact as plus, or?

4. The moment I connect a device that's not grounded by itself e.g. the floating output of the SG505 to the input the common mode noise seems to be or is actually converted into a differential one.

5. If I make a short directly at the input no change is to be seen.

So provided that you can agree so far, tell me please how would you explain and also test this effect?

(Only after you tried all roads you can tell which one is a dead end one.)

P.S.: yes, got me some samples from TI already but the controller of the display supply converter is from a smaller company I don't know if they are so generous as TI.

[lemon]

2. waveform refresh rate improved
- Seems to be some improvements cause changes at Autoset.
For that reason I repeat the test of Mark Madel (Owon SDS7102 review. youtube) with the trigger output capturing at Chanel II.
The Mark has found that this DSO varies the refresh rate depending time base & memory depth. He had recorded 22.96fps at 1MHz sinus signal (1K depth memory, 1us time base) that dropped to 16.43 when the memory depth was 10M.
With the same conditions, I repeat the test and the results seem to be very faster. With 1K depth memory the refresh rate was 119.07 and with 10M was 36.95. Nice to be good!
Please look at my attachment to verify.

This old test give a little indication that the speed can increase, that there is room for improve wfms/s speed some amount but of course it is still conventional  DSO, not DPO.
Nice that they have finded some solution for add speed by FW.

For rare peaks etc it is more fast than may think if look only continuous wfms/s speed.
It can rearm trigger faster and do one capture to buffer. But it can not of course repeat this cintinuously. But, for rare extra glitch this may find it more fast that may assume if look only continuous average wfms/s speed.

Here is this old test agen:

(http://owon.forumup.com/viewtopic.php?t=183&mforum=owon)

I can see also nice new work and thinking about noise.

Hi rf-loop.
I was remembering your way to measure the refresh rate but I haven't two channel signal generator for this.
Unfortunately, my Hameg is one channel and the only way that thinking is the test of M.M

[lemon]

Just got my curve tracer working with the SDS7102 in XY mode. I made a special interface cable with adjustable voltage dividers so I could perform the calibration procedure. This of course wasn't necessary with the Tek CRO because I could use the variable controls for that function. The first two attachments are captured images from the SDS7102, on the first one persist is turned off, which was the only option before the latest firmware release, on the second one persist is turned on. The third image is an older photograph of the Tek CRO I've been using with my curve tracer until now.

The persist option as implemented for XY mode is not perfect, the time settings don't seem to have any effect, and regardless which one you select, it seems to operate as if infinite had been selected. But a least I can obtain a usable image if I want to keep a record of curves for particular components. So I think I can now retire my Tek CRO from this job.

TomC, thanks for the testing.
It seems to be not to very fine but simple "well" this XY mode.
But, it seems like a dots type than vector, something like dots with a hi resolution (more dots). Of course the job is done!

I don;t know but 2 years and still we haven't Div/V adjustment.
Is it so difficult for this scope?

[TomC]

Hi Sarasir,

The way I understand it, in theory, the common mode noise stays that way until it encounters an impedance imbalance in its path, the change in amplitude and/or phase at that point results in a difference of potential between the two terminals which is the definition for differential mode. In other words, the common mode noise now can be seen by the scope. For example, that is the effect of the 6" ground lead, in that it causes an impedance imbalance, and what before was in phase and at the same potential at the sleeve and inner conductor of the BNC, now exhibits a difference in phase and/or potential.

At this point I don't understand exactly what in the SG505 setup causes the imbalance. From your post I can tell that it was connected directly to the BNC, I presume with a coax wire and BNC connector. So it's not clear in my mind where the imbalance is taking place at. Maybe a schematic of the SG505 may help. I'll think about it some more, and in the mean time you let me know what else comes to mind.

The wife is impatient wants to go for a walk or bike ride, so I'll get back to you later on today, maybe with a better explanation.

[jaxbird]

Link to Owon is below, click on Download Center:

http://www.owon.com.hk/main.asp

Done several upgrades on my unit without any problems.

Many thanks 

Update went without any issues.

[Sarasir]

Hi TomC,

you made my day man, that's the term - impedance imbalance.
All consequences of that where whirling in my head for weeks, phase delays, frequency peaks and zeros but couldn't think of a common term. And once you have it it's looks so easy.

Thank you very much.

Had a look at the circuit of the SG505 and its basically just a sine generator, a phase inverter for the minus side, and two complementary push-pull power amps that drive the + and - outputs, with a switchable attenuator before each output. All power is supplied by a conventional power transformer through linear +/- regulators. The common of the whole circuit is connected to the front panel's CT connector as they call it and is not grounded but may be switched to ground (chassis/power cord ground)  by a push button. Another Pusbutten (On/Off) disconnects the generator from the output amps, and by means of a switch, 50, 150 or 600 Ohm source resistance can be selected. None of the switches altered the noise displayed on the Owons screen. Only when I shorted the CT connector to the ground post on the front panel with a screw driver it went a bit down.

But don't want to bother you any more with that circuit - unless you are personally interested.

'Impedance imbalance' made everything clear to me, wow!!!!.

Yeah, loved also the bike rides along the river here with my last girl friend a few years ago. Over 4 years from heaven to hell and back untill I learned another expression that was not really new to me, but never cared about what it really means: manic depression. I just needed a break - was close to go down - and now I see it in every woman. Don't know when that will go away again....

[TomC]

Hi Sarasir,

I found an SG505 manual on the web, the only thing I see that may had played a small role in the impedance imbalance is the capacitor with the arrow on attachment #10. But I don't think that can even come close to account for the near 40mVpp noise you described, specially at the 2mV/div setting since it has a 20MHz bandwidth limit. The only thing that comes to mind that may account for that kind of imbalance is an unterminated coax connection to the scope's BNC, because of reflections and standing waves. But I doubt that's what happened because you already mentioned how important proper termination is. So I don't know what else may have been involved, unless your 50 ohm terminator is open and you haven't noticed.

Anyway, I thought the subject of proper termination may be useful to other members, so I captured some images to illustrate my experience. For the attachments I used my audio generator set to around 100kHz, 10mVpp. This is not a sophisticated device like the SG505, but it can operate from a power brick or from 9V batteries. So for these tests I chose battery operation, so the output is floating, and at least we know some of the noise is not coming in via the power brick. Can't say the same thing for radiated emissions from other devices in my lab, couldn't turn off everything else for these tests, so at least some of the noise doesn't originate at the Owon.

Because the noise amplitude varies constantly it's difficult to illustrate with a static capture what is going on, so I decided to set the trigger to normal and set the trigger level as high as I could and still get an occasional sweep. That way it's possible to capture the highest noise examples. Because of this the trigger frequency may not show 100kHz even thought the generator is set to 100kHz.

Some of the captures were taken with a .01uF capacitor connected between the (-) post of the generator and earth ground. I just wanted to see if there was any difference with the capacitor in place and without it. The idea came to mind because the SG505 uses a similar capacitor when in floating mode and I wanted to simulate the effect that it may have.

#1 through #4 - Here I have the Owon connected to the audio generator via an unterminated coax. On the first two images the generator is off, on the next two it is on. Some captures show the noise with no capacitor connection to earth ground and some with a capacitor connection to earth ground. There seems to be a little more noise with the capacitor connected to earth ground. The bulk of the noise, however, is present regardless.

#5 through #6 - These are similar to the last two images but the coax cable is now terminated at the scope's end with a 50 ohm terminator. Don't know what happened to my in-line terminator, so I had to use a T and a cap terminator. Not the best arrangement but it shows quite a contrast with the no termination setup.

#7 through #8 - Here I ditch the coax setup and use a X1 probe instead, the lower bandwidth of the probe makes quite a difference as it rejects most of the remaining noise. Using the spring ground clip seems to give a small extra improvement.

#9 - Same setup as #8 but limiting the scope's bandwidth to 20MHz, seems like there is an additional small improvement in the noise.

[TomC]

Yeah, loved also the bike rides along the river here with my last girl friend a few years ago. Over 4 years from heaven to hell and back untill I learned another expression that was not really new to me, but never cared about what it really means: manic depression. I just needed a break - was close to go down - and now I see it in every woman. Don't know when that will go away again....

Know exactly what you mean and had to learn every detail of what this disease, bipolar disorder, means over the years. My wife suffers from it since childhood, but the diagnosis didn't happen until a major manic episode over a decade ago. After endless therapy and medications, she is now free from psychotropic drugs and successfully controlling the ups and downs with fish & krill oil as well as yoga and exercise.

[TomC]

TomC, thanks for the testing.
It seems to be not to very fine but simple "well" this XY mode.
But, it seems like a dots type than vector, something like dots with a hi resolution (more dots). Of course the job is done!

I don;t know but 2 years and still we haven't Div/V adjustment.
Is it so difficult for this scope?

Yes, even though you can select vector or dots, it's always dots, that didn't change with the new firmware. However, you can now select persistence and when enough dots accumulate on the screen to make a line you can stop the scope and capture the image. The thing is that you have to this manually because the time options don't do anything, dots keep on accumulating forever and eventually you have a very thick line. But it's better than before.

[Sarasir]

Hi TomC,

your pics above show exactly what I had experienced. By the way, the SG505 I used has the high Level output (option 02) installed and in this version capacitor C1720 is not present. I guess the stray capacitance of the additional high level out PCB towards the grounded frame and shields replaced that in a way. I have to admitt that I didn't use any termination with the SG505 because it goes only up to 100KHz and here no reflection in the BNC cable needs to be considered. (Of couse I would use appropriate terms if had to check e.g. a 600 Ohm balanced audio line or 200 Ohm mic input of a mixing desk, but that would be only for correct leveling.) And, now as we know in a way what's going on here my recommendation would be to use a 50 Ohm temination rather on the Audio Generators side of the BNC cable because it's the Owon who is the generator of the HF spikes. Have you thought about that?

Even if I knew from working with the SG505 since decades that it is absolutely clean I simply connected it after experiencing  the spikes over to my Tek 475A (250MHz analog scope) and the display was clean of course. Connecting both the Owon and the Tek in parallel to the SG505 also the Tek showed the spikes.

So I knew that there's something wrong with the Owon.

After a few more tests with different probes and cables I just started with checking through the Owon to find and kill the problem.

Also searched the web but found almost nothing - just one short remark in a review said that it's very noisy.

To the thread here I came by chance much later when searching for the datasheets of the last two switching controllers.

Your results by using the probes just verify the findings of your earlier tests with your HF generator and your probe.

Yesterday I got the Backlight switcher working in another small tin box and today I will start rebuilding the last one - the display supply.

You know as I alredy said I want to get rid of that noise entirely so to be able to connect in which way ever whatever I have to ckeck and do not have to deal everytime with these noise suppression considerations - but we know it's a long and hard way.

Yesterday when reading through the recomendations of the TPS61160A backlight switcher controller for denoising found an interesting remark about ceramic Cs. Didn't know that yet but also had no time yet to read through all denoising literature including the ones you posted.

I will come to that when I have the Owon working again and will focus on minimizing the noise and also component use. At the moment I'm more ore less using brute force to get every switcher clean. 
-------------------------------------------------TPS61160A----------------------------------------------------
Datasheet, page 11:

Care must be taken when evaluating a ceramic capacitor’s derating under dc bias, aging and AC signal. For example, larger form factor capacitors (in 1206 size [3.2mm]) have a resonant frequencies in the range of the switching frequency. So the effective capacitance is significantly lower. The DC bias can also significantly reduce capacitance. Ceramic capacitors can loss as much as 50% of its capacitance at its rated voltage. Therefore, leave the margin on the voltage rating to ensure adequate capacitance at the required output voltage.

-------------------------------------------------TPS61160A----------------------------------------------------

And to the disorderly bahavior, my last girlfriend who has already been hospitalized once before I met her for the first time (what I only learned later through a neighbour who told me the whole story) would refuse any help... So what can you do?

[TomC]

Hi Sarasir,

I debated in my mind with that concept, but since the Owon in this case is not just the source but is also used to view the signal, I looked at the coax as a kind of a probe, so I thought it would be prudent to have a terminator at the scope's end. It's basically what one would do if using a coax as a high frequency (low capacitance) X1 probe.

The highest bandwidth analog scope that I own is a Tek 453A (60MHz), I doubt I can see much of this noise with it, but I may dust it off and experiment with it a little later on.

I'm anxious and very curious to see how clean you can get the Owon with your mods, I hope you have the time to post some pics similar to the ones I posted when you get it as clean as possible.

Not much you can do if they refuse help, I'm glad my story turned out differently!

Edit:
On the other hand, if the objective is to see how much noise the Owon is producing, then of course you don't want to use a terminator at the scope's end. In this case this setup would be similar to the one we've been using in this thread with the probe tip and ground lead connected to the ground lug. Basically just to purposely create an impedance imbalance to make the common mode noise (GND noise) visible.

[Sarasir]

Hi TomC,

Of course I will post them. Looking forward to it by my self.

-----Do not read this - only more preliminary thoughts.-------------------------------------------------------

the proposal I made to put the termination to the generators end was of course only meant for to see the difference concerning the noise displayed (I just forgot to add that), not at all as a general rule for every days work!

Actually I thought that if you put the terminator to the generators side - only in case of a non HF generator - the noise displayed should get less as the terminator would minimize the HF impedance imbalance of the generators output. But as I think about it again now there should be of course another terminator at the scope end as well (as the input otherwise would be 1 MOhm and so not correct for a BNC cable - I forgot that before) and the generators output switched to highest source imedance (like 600 Ohm at my SG505).  But again only meant for testing noise behaviour. But wouldn't say much I think only the influence of the BNC cable in connection with the impedance imbalance of the generator would be cancelled out in a way.
------------------------------------------------------------------------------------------------------------------------

Was just another one of my unfinished ideas I always have on the way to true wisdom...

And  I also forgot that the noise source is on the ground side of the BNC....


But now out of the above rubbish I think I have three statements that should be true (generator here is a non HF audio sine wave one):

1. The transfer of the noise signal from the scope to the generator happens initially through the shielding of the BNC cable (here the preliminary thought above was a dead end 'cause you cannot terminate the shielding - against what?).

2. And the BNC cable acts in its normal way then to bring back the result of the impedance imbalance in the generators output to the scope (only this part will be infuenced by correct termination) (...and not only the output but the entire generator circuitry might be involved more or less)

3. And as the shielding of the BNC cable is not at all a perfect medium for to transfer HF signals part of the noise will already be coupled to the inner conducter way before it reaches the generator including all the consequencies like impedance missmatch and reflection.

Can you agree with that?

And I'm glad for you too, cerntainly something if you stood that through with your wife.

[TomC]

Hi Sarasir,
Trying to visualize exactly what's happening is not easy, I have some ideas, but that's all they are, no concrete proof and only scant evidence that they match reality!
So here is what I currently think, however at this point I'm not sure how to test or verify it!

The following is just a little background of things I think are pretty much of general knowledge:

As we know, most of this originates at the switchers. Due to stray capacitance, inductance, or due to radiation, some of the fast switching current strays from its main path and flows through an alternate unintended path to reach the return. These stray currents can be differential or common mode from the scope's input point of view. Owon has managed to clean up the differential signals with filtering to the point that there is a reasonably low amount of baseline noise, and a reasonably low amount of distortion on the signal path. However, some common mode currents are still present that are invisible from the scope's input point of view.

The following is what I think is happening inside the Owon in more detail but can't prove it:

So I believe that what remains are: a) Currents that are nearly identical in both phase and amplitude at both the sleeve and center conductor of the BNCs. b) In addition, as long as there is nothing connected to the BNCs, I believe there is an even larger number of current paths that are only present on the ground plane but are not referenced to the BNC's center conductor. Therefore these currents are also invisible from the scope's input perspective.

When you connect something to the BNC, for example, a probe cable or plain coax, I visualize a couple of long physically close parallel conductors analogous to long adjacent parallel traces on a circuit board. Since the shield is part of the ground plane, any currents on it that did not originally have a counterpart in the BNC's center conductor will be coupled to it. Assuming a perfect cable, the result will be common mode currents from the scope's input perspective, and as a result, these currents will still remain invisible.

The following tests out experimentally, but the reason why it happens is just my theory:

However, in reality, due to cable imperfections, I would expect a moderate but gradual increase in differential signals as the length of the cable increases. In other words, you would see a little more noise displayed on the scope as the cable length increases.

The following is just a little background of things I think are pretty much of general knowledge:

As the cable length increases propagation delays become more of a factor. This effect prevents distortion free transmission of differential signals, particularly HF signals, from one end to the other unless the cable is properly terminated to match its characteristic impedance. Improper termination will result in reflections that end up being added to the original signal. The resulting signal is a distorted version of the original signal.

The following seems to test out experimentally, but the reason why it happens is just my theory:

However, the signals of interest in this case are common mode, and the concern isn't signal distortion, instead, we'd like to know the mechanism that converts some of these common mode signals to differential mode signals. First consider a coax cable cut at the end with a sharp razor blade. To a signal this cable will exhibit an abrupt transition from 50 ohms characteristic impedance to nearly infinite impedance. However, because of the clean and nearly perfectly symmetrical cut, there will be little if any impedance imbalance in respect to the shield and inner conductor, so the reflections caused by the abrupt change will result in nearly identical currents, in both phase and amplitude, on both conductors. If the scope's input circuit is properly designed and laid out, an analogous effect will take place when the reflections encounter the 1 meg ohm input impedance. So any increase in visible noise given these circumstances should be small. In contrast, if for example, the coax ends in 1 or 2 inch pigtails, the increase in visible noise is quite a bit larger. There is still an abrupt change in characteristic impedance, but in addition, there is also an impedance imbalance because of the rather raw unequal imperfections at the end of the cable. I believe in this case reflections multiply the effect of the impedance imbalance at the end of the cable. Outside interference probably also plays a role in this scenario. Also, in my opinion, any equipment attached to the pigtails will also have an influence on the impedance imbalance and/or the characteristic impedance at the end of the cable.

I also believe that using proper termination, which we know produces the illusion that the cable has an infinite length and as a result minimizes reflections, will also minimize the multiplying effect that I believe an improperly terminated cable will have on impedance imbalances at its ends.

Anyway, that's the way I view the effect of proper termination as it relates to the increase in visible noise. But again, the main reason to use proper termination is to minimize distortion of the differential signals that traverse the cable on their way to another device.

How much of the above reflects the way that things actually work? I'm not sure, it sounds good to me at the moment. The only thing that I know for sure is that the visible noise is reduced when proper termination is used.

[Sarasir]

Hi TomC,

things clear up more and more. Your analysis and explanation is very profound and I read it several times, just great.

A few thoughts and facts I found out I might add:

As to the noise transfer from the switchers inside I think that actually most of it takes place through the ground posts by which the respective PCBs (mostly the adapter PCB) are screwed to the main frame. The adapter PSU is held by 5 screws and one could say, to realy make sure that every switcher has it's direct connection to pass on it's noise.

(Only the little switcher on the main PCB has no direct connection to ground only through the grond plane of the PCB.)

I tested that with my Tek 475A by using a probe with very short ground lead. Connected the ground lead to the frame in different locations and moved around with the tip on the surface of the frame. Close to the ground lead there was no noise or very little, moving further away the noise increased remarkably, of course more or less depending on the location. Most increase per centimeter so to say I encountered in the area where the adapter PCB is connected to the Main PCB. The cut-through here for the connector if far to big resulting in a weak ground conduction.

It's no supprise at all that there is noise as well present on the BNCs. The whole frame is polluted and the return path takes place mostly through the electrical field(s) but also through the magnetical field(s) and the power cord I think. The whole thing acts like a radio transmitter with lots of little antennas all around.

A little detail by which you can see the struggle of Owons engineers is how the Main PCB is screwed to the frame (so in my model). The hexagonal spacer bolt between frame and input side of the main PCB is a plastic type. Only a few centimeters to the side where the trigger input is located they use a metal post. Consequenly there is a bit less noise on the input BNCs than on the trigger BNC.

But they had to connect the frame with the main PCB on the top side again with 2 metal spacer bolts.

You know out of my long time working with Tek and knowing their machines back then in and out, down to the last screw I know excactly how to deal with ground and shielding or in other words how things like this should be done. Noise on the ground is just inacceptable.

And to the use of termination e.g. with an audio generator which would not realy need one: You are of course aware of that a termination put to the output reduces the signal level by half (on a 50 Ohm output, on higher output impedances of course even more reduction) as it acts also as an attenuator in conjunction with the source impedance. So if a certain noise reduction is achieved and the 'good' signal is also reduced by the same amount or even more than there is no improvment finally. Only an increase in signal to noise ratio would give a benefit.


And by the way, I think you mentioned that you've got a second PSU. Was it from Owon directly or from where you purchased it and did you have to give the old one back?

And the last thing for now: Checked the web for other DSOs just to learn more about and learned, that Rohde & Schwarz untill August I think have an offer standing that they give 40% discount on any DSO provided you trade in your old one. So get one for e.g. 10.000$, give them your Owon and just pay 6.000$. It's just a 'bit' too much for me but would be actually not such a bad deal. And they have very fine machines.

Good luck

[rf-loop]

A little detail by which you can see the struggle of Owons engineers is how the Main PCB is screwed to the frame (so in my model). The hexagonal spacer bolt between frame and input side of the main PCB is a plastic type.

This detail have totally different history from year 2011.

Also in this time we did not have at all this level of "noisy GND" sad case what happend after some peoples in Owon, who did not have enough experience and knowledge, start doing modifications to power supply, adapter board and other things where we can see local switch mode sub power supplies/regulators.
In this time they have forget all rules how to design SMPS and also how to isolate them for avoid common mode noise distribution to all around.
Also in year 2011  internal metal base, aka - Z-plate was polluted but not at all like after these sad modifications.

After this GND noise case totally blow up it was total catastroph and also I did hundreds of hours in lab for investigate this year 2012/2013 new show what Owon schoolboy "designer" arrange to us. 

If thinking is that less experienced and educated people get work for design PSU's because of course simple PSU is easy, every schoolboy can "design" SMPS - or something like this thinking. I can tell that just this is perhaps one of most difficult part to desingn and even experienced engineers do lot of shit in this area. But if really know what to do it is "easy".

Afterward modification Owon to "clean" is impossible. It is totally impossible without kick out all exept main board. But still lot of can do. After then it just - what is accaptable and how much is wise to do. If really use lot of hours, investigations, lot of knowledge and experience - it can do quite clean but not perfect. But it can also say that when go to some amount less noise then we have lot of brands what also need "repair" today.

Owon is spread spectrum RF transmitter.  (Owon is not alone but this radiating level is  terrible if it is this worst version and even manufacture lot may affect due to change of components.  Components quality may also vary. One manufacturer 10uf inductor may be totally different as other same looking with same main specs but perhaps very different more detailed specs and result in SMPS circuit may be totally different. Same with diodes, capacitors etc. This may also explain why same looking but different lot may have different noise situation.




But here something about this plastic pillar what was my main reason for answer.

CH1 come noisy after they change CH1/2 front end area lay out / circuit totally. After then this metal pillar connect this some amount polluted Z-Plate to very sensitive GND area inside CH1 circuit.  Owon did these and end users get these. After short time problem was solved by Owon and they tell that solution is change one metal pillar to plastic pillar for stop this noise road from Z-plate to front end circuit.

Before this it was very different.
 

Back to today problem. There is so many noise sources that making it clean is least difficult.
After  shut off 6-7 most bad SMPS noise, then under these can see other noises. These separate noises hide each others, so running scope and inspect noise and what part these are generated  need trivial trick. Oscilloscope what use for analyze need two channels. Other channel is noise chaos pick up and other channel can synch to different sources. With this method can easy pick up and separate every different noise souce produced fraction. Even Owon itself can use for separate these noise fractions so that these can investigate on the Owon screen. But there need walk over some traps with working methods.
Warning!  main PSU prmary SMPS control IC, as other also have told, leaks its fuel - smoke - out very easy. One wrong touch with probe (no need shortcirc) and buff.

After most bad noise fractions are reduced then come data/power share to front panel(s) dependent of version, there come TFT panel itself, there come this terible TFT databus from main borst and thru adapter board to TFT cable. These also are related to version. Some older versions have example display control circuit in totally different place if compare more new versions. (old version have also flat cable from controller to optional VGA output. This time databus to TFT was physically different and afaik less noisy. (now this databus is like mugpie and impedances vary around bus and whole bus radiate around but these frequencies are not very high.
 
After "noisy GND" catastroph happen I (and some others also) try do some analyzing and help work directly to Owon but... result was bad. They did not believe nearly any recommendations for fast change it better and also handle warranty things for unhappy end-users so that this case do not destroy Owon.  I will not show these investigations and final decisions and recommendation for circuits due to some kind of NDA.  After then I stop and I'm partially satisfied it is now better than in worst times. many peoples can use it without problems and it have some advantages and disadvantages.   I can not work for solve some problems and then after I'm ready I can see peoples have shutted of they eyes and ears. So, they get what they order themselves. I have loosed tens-hundreds of intensive lab work for this without get anything, so just - hands up. It more nice to help and support peoles who listen. I can not do any miracles but I have still "silent knowledge" things after over 50year hobby and after over 30 year active work... 

Here link to this some old (and obsolete) "help paper" for  this CH1 noise case when some end users do this repair (change this pillar to nylon pillar) and because many users are entry level some need help.
http://www.box.com/s/pvg1dqtdonx192p4os0m

Btw, who can count all Owon SDS HW versions...and combinations even if talk about same model...  how many? I have not any idea anymore. 2011 and some amount 2012 I know nearly sure but today, impossible to know.

Btw, whole all time what I have sold some Owons, end user warranty time problem/failures around 2%.
Some (name hided)  brand  was over 10% end user side and for me over 40% (difference is because I do 72hour burn in and test for every equipment what I sell exept for Siglent 48h due to building quality and factory better Q.C. in product line. But still these all Chinese factories Q.C. leaks.

[TomC]

Hi Sarasir,

The new style PSU I got came directly from Owon. They wouldn't give it to me for free, so I purchased it for US 50. In addition, since they only accept bank transfer for payment, I also had to pay fees for the transaction. If I remember right the fees were around US 40. So my old PSU wasn't part of the transaction in any way. However, later on other members got new style PSUs from their dealers after complaining about the GND noise problem. In those instances I think they only had to pay for shipping and didn't have to return their old PSUs.

In any case, I'm currently using my original PSU because after my mods it performs a little better noise wise than the new style PSU that I bought from Owon. Of course, I didn't know it would turn out that way, I didn't do any mods on my original PSU until after I received the new style one. I was afraid I'd smoke it and be without a DSO and that maybe my 3 year warranty would also be void.

By the way, on the new style PSU, the PCB's ground plane, I call it GND-C on my schematic, is directly connected to the chassis (Z-plate) through the PSU's mounting posts. On the old PSU, like yours, GND-C first goes through the ribbon cable before being connected to the Z-plate via the adapter board's mounting posts.

I get what you mean about a 50 ohm termination reducing the signal level of you audio generator. What I don't understand, probably because I don't do much audio work, is why is it necessary to connect the scope to the audio generator via a coax. I'm just not familiar with the application(s) where that is needed or more convenient.

Good luck with the continuing de-noising saga. I think you'll eventually clean up all the switchers. Are you going to do something about the TFT as well?