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

[AndrejaKo]

I took some pictures of modifications from yesterday when running on battery in room with no electronics devices. They are the first two attachments and I was quite happy with the results. Ringing did remain!

I decided to try to implement all modifications from the lemon's picture, but I got some pretty bad results. Image is the third attachment.

Next, I removed the bar and the capacitors going from the bar onto the connector, but I forgot to take a picture. Next, I removed all of today's mods, but I couldn't get near the results I got this morning with yesterday's mods. That's the fourth attachment.

I also noticed some new bursts of noise coming and they're in the fifth attachment. They are completely new and didn't exist until today. I'll carefully compare my picture old PSU mods to see if I did something and forgot about it.

[TomC]

I took some pictures of modifications from yesterday when running on battery in room with no electronics devices. They are the first two attachments and I was quite happy with the results. Ringing did remain!

I decided to try to implement all modifications from the lemon's picture, but I got some pretty bad results. Image is the third attachment.

Next, I removed the bar and the capacitors going from the bar onto the connector, but I forgot to take a picture. Next, I removed all of today's mods, but I couldn't get near the results I got this morning with yesterday's mods. That's the fourth attachment.

I also noticed some new bursts of noise coming and they're in the fifth attachment. They are completely new and didn't exist until today. I'll carefully compare my picture old PSU mods to see if I did something and forgot about it.

The first two pictures look quite good! If you can get back to that, that level of noise is acceptable and you should be able to work with low level signals while using the long ground lead.

What I find puzzling is the low frequency high amplitude spikes. Since you get them running from AC or battery, I don't think the 8.4V supply can possibly be involved. Where are they coming from? That's a hard question to answer!  Do you have any other scopes available to help troubleshoot?

[lemon]

Something is going wrong, sure!
The result on third capture is bad!

I searched this thread to find my older posts about same mods. All there are to page 88 and then. From this page with the same mods (not to adapter board, only to psu board without any ferrites) my results was:


Before the adding smd decoupling capacitors, I have changed all the electrolytics capacitors. Look at the following image (right is what I had before, left what I take with the new caps)


I don't believe that the results with the smd decoupling capacitors are caused of electrolytic capacitors. All the rf-loop's result are without any electrolytic capacitors changed.

Unfortunately, you didn't take a capture of all these mods that you done. Perhaps, someting goes wrong!
Did you believe that you helping if I take a photo capture from my board with the mods?

[lemon]

I don't know exactly what values of smd caps Owon used to the last version, I supposed that used the rf-loop's suggestion.

While I had the scope apart I measured the SMD caps on the new PSU board. Most of them didn't give me a true reading because of the surrounding circuits, but the few I was able to measure were around 220nF. These were of the thicker type. There are also thinner caps I couldn't measure, I suppose those are the 1.5nF ones.

TomC, can you to mark with the arrows on PSU Print Side Circuits Photo what smd are 220 and 1.5?

[TomC]

I don't know exactly what values of smd caps Owon used to the last version, I supposed that used the rf-loop's suggestion.

While I had the scope apart I measured the SMD caps on the new PSU board. Most of them didn't give me a true reading because of the surrounding circuits, but the few I was able to measure were around 220nF. These were of the thicker type. There are also thinner caps I couldn't measure, I suppose those are the 1.5nF ones.

TomC, can you to mark with the arrows on PSU Print Side Circuits Photo what smd are 220 and 1.5?

Here they are!

[lemon]

You are really fast like Bolton! Many thanks, I am going to examined vs what we proposed.

[AndrejaKo]

@lemon
If it's not a problem for you, please post image of your PSU with mods. Maybe it will help. I suspect (maybe it would be better to say hope) that the source of this noise is some minor thing that I'm overlooking.
 
@TomC
The spikes are coming from the power supply itself. I've just connected my computer's microphone input to the PSU, disconnected the adapter cable and turned it on to run from the battery. The noise was there even with the rest of the scope powered down. I don't have another scope (this is my first one), but there is one guy I know with TDS210 who owes me a favor. I'll try to arrange the use of his scope, but that may take a while.

[lemon]

No problem, tomorrow you will have it.

[AndrejaKo]

Ok, great! I'm too tired to do anything today anyway.

[TomC]

@TomC
The spikes are coming from the power supply itself. I've just connected my computer's microphone input to the PSU, disconnected the adapter cable and turned it on to run from the battery. The noise was there even with the rest of the scope powered down. I don't have another scope (this is my first one), but there is one guy I know with TDS210 who owes me a favor. I'll try to arrange the use of his scope, but that may take a while.

What happens if the Pwr Sw (J2 ) is not pushed in. I'm asking because with the adapter cable disconnected, the only difference between pushed in or released (on or off), is that in the first case the -7.6V DC/DC converter is energized, and in the second case is not.

[Skimask]

If I can interrupt here for a second...then I'm eat a whole bag of "step out of the way and let the pro's go at it"...
I've been following this thread for awhile, trying to decide on a decent digital 'scope that's a notch or seven better than the old 1052 at not much more cost.

All this talk about ground noise and such, does it have any effect on probing logic level signals (e.g. above a volt or so), triggering on them, and the like?

[AndrejaKo]

@Skimask
Even with all of this noise problems I'm having, I had no problems working with logic level signals at all. This scope won't catch very fast glitches, but if you need to take a look at general purpose I2C, SPI,RS-232,RS485, low or full speed USB and other slow protocols, then it's OK. Of course, it doesn't have protocol decoding, so you'll have to work out by hand what's going on and there's no memory segmentation either.

UPDATE:I made a bar again and this time used 1 µF capacitors between all lines and the bar. Now I have results which are relatively close to the good ones I posted few pages back, but only when the probe is in the fixed position as shown few pages back. If I move it away, then I get the new spikes back, but they aren't as bad as before. This was made when running on battery. I'm a bit concerned how the AC trigger signal is going to work with such a big capacitor on it, so I'll probably go back and change that one capacitor to something lower.

[Carrington]

AndrejaKo: These results are not bad!
But "1 uF capacitors Between all lines and the bar" What do you mean? Bar?

Now to prevent that the probe position affect to the measure -> shielded all the oscilloscope.
http://www.opticalfilters.co.uk/emiclare-ito-film.html
http://www.adafruit.com/products/1309#Description
http://uk.farnell.com/kontakt-chemie/emi-35-200ml/coating-conductive-emi-35-200ml/dp/2142398

[AndrejaKo]

By "bar" I mean the thing referenced by lemon in his list of recommendations few pages back. Basically, new PSUs have a ground track running near the board edge close to the power connector. There are capacitors running from that trace to the power connector. You can see that in TomC's post with a photo of new board. Lemon recommended that I make a bar (would that be the right term to use here?) and that I connect it to the ground lines at both sides of the power connector and then place capacitors from the connector to the bar. I made the "bar" from a piece of 1.5 mm² solid copper wire and then connected capacitors to it. I hope this clears things up a little bit.

[lemon]

Here I am and upload my psu with mods.

If you would like some more explanation, don't hesitate to ask me.
Carrington, the Farnell denied to send me the EMI Shield Spray (it is courier rule for airport transfers).

[Carrington]

AndrejaKo: I understand, thanks for the clarification.
lemon: Try Amazon or eBay.
  http://www.amazon.es/KONTAKT-CHEMIE-EMV-Abschirmlack-200/dp/B00154KONS
  http://www.ebay.es/itm/ws/eBayISAPI.dll?ViewItem&item=261076709207

[lemon]

@Skimask
Even with all of this noise problems I'm having, I had no problems working with logic level signals at all. This scope won't catch very fast glitches, but if you need to take a look at general purpose I2C, SPI,RS-232,RS485, low or full speed USB and other slow protocols, then it's OK. Of course, it doesn't have protocol decoding, so you'll have to work out by hand what's going on and there's no memory segmentation either.

UPDATE:I made a bar again and this time used 1 µF capacitors between all lines and the bar. Now I have results which are relatively close to the good ones I posted few pages back, but only when the probe is in the fixed position as shown few pages back. If I move it away, then I get the new spikes back, but they aren't as bad as before. This was made when running on battery. I'm a bit concerned how the AC trigger signal is going to work with such a big capacitor on it, so I'll probably go back and change that one capacitor to something lower.

If you straighten the problem cable, you catch everything like antenna.
I don't have my scope functionality now, if another member with low level gnd noise test with the straighten the probe cable you'll see that the noise is increased.

[lemon]

AndrejaKo: I understand, thanks for the clarification.
lemon: Try Amazon or eBay.
  http://www.amazon.es/KONTAKT-CHEMIE-EMV-Abschirmlack-200/dp/B00154KONS
  http://www.ebay.es/itm/ws/eBayISAPI.dll?ViewItem&item=261076709207

Carrington, thanks for the links. I had order the other with coating by Nickel, was 400ml vs 200ml of Kontact Chemie. Had you search them more and decided to order the Kontact Chemie?

[Carrington]

Carrington, thanks for the links. I had order the other with coating by Nickel, was 400ml vs 200ml of Kontact Chemie. Had you search them more and decided to order the Kontact Chemie?

Where have you bought the coating nickel spray?
With 400ml on Farnell I only find this:
http://es.farnell.com/kontakt-chemie/graphit-33-400ml/coating-conductive-400ml/dp/836606
I bought the EMI 35 on Amazon.

[lemon]

From Farnell, look at http://export.farnell.com/jsp/search/productdetail.jsp?SKU=1675347
If you can't to open search for order code with 1675347

Also, just came from Farnell this flat ferrite and fits perfect to flat cable of adapter. Look at:
http://export.farnell.com/jsp/search/productdetail.jsp?SKU=3538369
If you can't to open search for order code with 3538369

[Carrington]

From Farnell, look at http://export.farnell.com/jsp/search/productdetail.jsp?SKU=1675347

Also, just came from Farnell this flat cable and fits perfect to flat cable of adapter. Look at:
http://export.farnell.com/jsp/search/productdetail.jsp?SKU=3538369

Ok, I see now.
http://www.farnell.com/datasheets/31877.pdf
http://www.farnell.com/datasheets/1674774.pdf

I bought the ferrite on ebay. 
http://www.ebay.es/itm/190837205693


The SSH-33.5-20 is better, but I do not believe there is much difference.

[TomC]

Now I have results which are relatively close to the good ones I posted few pages back, but only when the probe is in the fixed position as shown few pages back. If I move it away, then I get the new spikes back, but they aren't as bad as before. This was made when running on battery. I'm a bit concerned how the AC trigger signal is going to work with such a big capacitor on it, so I'll probably go back and change that one capacitor to something lower.

AndrejaKo, try ferrites on the probe cable, one at the top close to the BNC, and one at the bottom, close to the probe itself. See if this helps the spikes when you move the probe cable and when you get it close to the screen.

The ferrites for the flat cable that Lemon and Carrington ordered may also help. Check the links on their posts.

[TomC]

I'm a bit concerned how the AC trigger signal is going to work with such a big capacitor on it, so I'll probably go back and change that one capacitor to something lower.

There may be a slight phase delay, in most applications that probably wouldn't matter. If you look closely at the new PSU board, you'll notice that C23 is missing. That's the equivalent of the capacitor you are talking about. Evidently Owon didn't think a capacitor there would make very much of a difference as far as EMI is concerned, so they left it out even though it was part of rf-loop's recommended changes. So it probably will be OK whichever way you decide to go.

[lemon]

AndrejaKo, if you haven't the work of rf-loop, you can see that at:
http://imageshack.us/photo/my-images/571/pic2psu2b.jpg/

The thread for that image is held at: http://owon.forumup.com/viewtopic.php?t=163&start=0&mforum=owon

[AndrejaKo]

With ferrites in place on the probe, I get around 100 mV of noise, but now it doesn't matter how the probe is positioned. I still see the spikes, but they are reduced. Also the large spikes remain.  In some positions, I can get as low as 50 mV of noise.

Removing the two insulation washers from under the top ground contacts make the noise look a bit "prettier" on the screen. Now I get a 150 mV thick line and the new peaks are almost invisible. The large peaks are still here.

As for the adapter cable ferrite, I already tried with few additional ferrites (both flat and round) and they didn't help much. I also added ferrites on the display cable and the LED cable, but I'm pretty sure that at this point they aren't doing anything useful.

I did try to register for owon.forumup.com once before, but it seems that administrator didn't approve my registration. I tried one more time now. Hopefully, I'll have more luck.