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

[TomC]

The next test will be with the whole scope and probe inside a Faraday cage, because I can not get off of that tone on 106.5MHz, I think that comes from outside.

That's a very interesting experiment. My feeling is that you are right and it does come from the outside. But the experiment should prove this beyond the shadow of a doubt. I'll be looking forward to the results!

[Carrington]

I agree with lemon, with or without the back cover in my case not much difference too, the Zplane does its job. However I think that is better if the back cover is installed. In my case the shielding of the front cover is very effective. I just added a few sheets of aluminum inside (look at attachments) and see the result. It joins the buttons with internal shield. Sorry for the quality of the images. I can't turn up the brightness of the screen.

The area isolated with plastic was on top of the input stage shielding, in my case is better if this part is not isolated.

The test with the scope and probe inside the Faraday cage, not work as expected. I'm still wondering why.

[lemon]

The noise of the two first is something 25mV (I can't to see clearly). The last isn't clearly to me (I must to see to my other display at my office that it have better color gamut) but I see that there is no any periodical noise.
Have you got any photos about your works with aluminum sheets, I can't understood why are you doing? Have you joined the gnd area of pushable buttons together?

[Carrington]

The noise of the two first is something 25mV (I can't to see clearly).

In previous images, respectively, 12, 24, 16 and 16 mV.

Have you got any photos about your works with aluminum sheets, I can't understood why are you doing?

I've been doing more tests, and I think that the previous improvement is due more to DSC04397.jpg / DS04398.jpg.

Have you joined the gnd area of pushable buttons together?

This is what I did, see DS04393.jpg, but in fact profits nothing.

Anyone know what this is (DSC04395), comes from the lcd, psu ...?

[lemon]

If it could be, because then is not properly a Faraday cage. I'm thinking that the Zplane can be like an antenna deflector, and therefore the shielding makes it worse, and would be necessary, to isolate the different zones inside, and this is very difficult.
....

It is done and this test without success. Now I am thinking what is the best solution to removed without damaged the plastics and logos.
Carrington with the last yours photos, I understood complete what you have done.

I repeat the conditions of your last photo, it seems that this pattern exist always but a little, magnify a lot of on lcd.
I make a video on that.

[lemon]

I done the same test to my lab pc tft display. The result has 100% same pattern and causes from tft display.

[lemon]

I done a test with independent main psu.
For this reason I used a Rigol DP832 Triple Output PSU. This psu has a low noise threshold (3.5mVp-p)
I gave it a +8.4V/2.5A and -7.6V/0.5A power supply to adapter board.

From what I saw, the ampere consumption doesn't bigger than 1.43A with 0% backlight until 1.55 with 100% backlight for 8.4V.
The -7.6V has stable ampere consumption of 0.160A

All the test was with the two case of scope closed.
There is no any improvement about noise, exactly the same.
It seems that Owon psu as is now, does the job well.

By the way, I make two videos, the first one with the output psu the second one with normal ac power supply of Owon.
The only difference was the effect of ac wave of 50Hz to the Owon psu, this phenomenon of course doesn't exist with the output psu.

[Carrington]

Is interesting confirm that all have this source of noise. In my case the highest peak occurs in the connector between mainboard and DC/DC adapter. Strangely with the probe on the screen barely appears this noise, may be because I have the screen completely attached to Zplane, I don't know.

I leave it, this instrument no has solution, my oscilloscope have this noise (lcd) as common mode noise everywhere (with or without shielding). I discovered this noise inside the Faraday cage, with 5ms time base, previously I focused only on 500 and 100us time base. I tested too placing a ferrite above on this connector (see attachments), but that place (connector) does not support it due of its shape.

Before the main source of noise was the PSU and DC/DC adapter, now that they have corrected their novice failures on these boards, we can see the influence of other novice error even worse to correct. As I said there is no solution, requires a total redesign. Fortunately the influence of this common mode noise is not very high, and the instrument is usable.

Owon should have used a more appropriate connector or at least should have spent more pins to gnd, but not did, and that says a lot... 

[lemon]

I think that before with the old psu and adapter, the main source of noise was there.

Now, that major noise decreased, we see and the others common noises from the other boards.

I think the same that you, that it is at their limit - let it as be!

I am very sure that you have seen that but I upload one more video that appears how sensitive is!

Please, if someone know a good method for removing Nickel shielding,  it'll be welcome.

[SeanB]

If it is a plate on plastic a soaking in a concentrated strong acid will strip it quite fast. Aqua Regia will strip it in a very short time, though the plastic under it may be damaged a little. Other methods are to electrodissolve it by using it as an anode in an electrochemical bath, as is used to do electropolishing of stainless steel parts, though you will have to gradually immerse it so the anode connection is placed in last to be the final part to dissolve. If you are just trying to cut it into sections then use a small sanding disc in a dremel or similar tool. If you want to leave  sections paint them to make them non conductive ( or cover with wax to repel water) then electrodissolve the remaining uncoated area.

[Carrington]

Using hydrochloric acid and hydrogen peroxide, and mixing all in equal quantities, should work.

[lemon]

Thanks for the info SeanB and Carrington.
I am afraind the plastic cases a lot of. Also I am afraid the logo, letters e.t.c

[Carrington]

Thanks for the info SeanB and Carrington.
I am afraind the plastic cases a lot of. Also I am afraid the logo, letters e.t.c

Well, wait I test it first to see what happens.


After all afternoon, see attachments.

[TomC]

Thanks for the info SeanB and Carrington.
I am afraind the plastic cases a lot of. Also I am afraid the logo, letters e.t.c

Well, wait I test it first to see what happens.


After all afternoon, see attachments.

The powdery substance on the second attachment, does it rub off clean, or did you get damage to the lettering and logos?

[TomC]

Thanks for the info SeanB and Carrington.
I am afraind the plastic cases a lot of. Also I am afraid the logo, letters e.t.c

I can't be sure, but looking at your last video it seems that the front panel acts as a shield for the main board. As you pull it away the noise increases, but if you pull away a shield you could expect something similar to happen. Why do you think that removing the cabinet's conductive paint would improve this situation? I'm probably missing something, but I'm not convinced these are related.

[Carrington]

I can't be sure, but looking at your last video it seems that the front panel acts as a shield for the main board. As you pull it away the noise increases, but if you pull away a shield you could expect something similar to happen. Why do you think that removing the cabinet's conductive paint would improve this situation? I'm probably missing something, but I'm not convinced these are related.

1. Conductive paint has cracked and can break off.

2. Each zone should be shielded/isolated one from another, i.e. PSU, DC/DC adapater, mainboar, from panel, etc... (See the DS2072 teardonw as good example). In the SDS7102 due to its iternal construction is very difficult to achieve this.

3. Do not forget the screen, I'm not willing to spend more money on the material needed to shield it properly.
I used a aluminum foil as a substitute using an external monitor.


"I can't be sure, but looking at your last video it seems that the front panel acts as a shield for the main board. As you pull it away the noise increases, but if you pull away a shield you could expect something similar to happen."

Yes, but this panel attenuating the field generated by the DC/DC behind it, in fact that is the predominant noise when removing it. However, the shielding (including the aluminum foil) not eliminates or attenuates the common mode noise "radiated" by the connector that joins the adapter board with the mainboard. For all this and the above exposed (1,2 and 3) I decided remove the conductive paint.

[TomC]

Carrington,
I get your reasoning, specially item 2, which I think would be needed to control reflections, particularly where SMPS circuits are present. Is the front of your scope's cabinet, lettering, logos, OK after the removal process? I saw a powdery substance on your last images and couldn't tell if it was damage or just easily cleanable residue.

[Carrington]

Carrington,
I get your reasoning, specially item 2, which I think would be needed to control reflections, particularly where SMPS circuits are present. Is the front of your scope's cabinet, lettering, logos, OK after the removal process? I saw a powdery substance on your last images and couldn't tell if it was damage or just easily cleanable residue.

No damage, but takes a long time, the process is slow. The acid concentration is only 25%.

[lemon]

I''ll do a trial to a test Nickel painted plastic, with a mix of HCL (conc. 7%) and H2O2 (conc. 35%).
From wikipedia read that the Nitric oxid is very well to Nickel dissolution, populate product about this is the "aqua forte".
It seems that yours shielding go away without problems...

TomC, the mainboard acts like as you said. I don't know if the exist a small flat ferite on the flat cable that joined the two boards could to be help.

The independent isolation of boards is very difficult to this. Agree with Carrington that "in the SDS7102 due to its iternal construction is very difficult to achieved this".

[Carrington]

I''ll do a trial to a test Nickel painted plastic, with a mix of HCL (conc. 7%) and H2O2 (conc. 35%).

 


After all night, the saturation of copper in the acid must be very high, if it gets into the pores of the plastic will be difficult to remove. So I think that is better change the acid when this becomes very green.

At the rear side:

  - I believe that the off-white residue/color is due to paint solvent, I doubt that it can be eliminated. 
  - The greenish yellow should be residue/color of copper oxide, after another bath clean acid should look better.

At the front side:
 
  - I can not see any damage.

[TomC]

I''ll do a trial to a test Nickel painted plastic, with a mix of HCL (conc. 7%) and H2O2 (conc. 35%).
From wikipedia read that the Nitric oxid is very well to Nickel dissolution, populate product about this is the "aqua forte".

Good luck!

[TomC]

After all night, the saturation of copper in the acid must be very high, if it gets into the pores of the plastic will be difficult to remove. So I think that is better change the acid when this becomes very green.

At the rear side:

  - I believe that the off-white residue/color is due to paint solvent, I doubt that it can be eliminated. 
  - The greenish yellow should be residue/color of copper oxide, after another bath clean acid should look better.

At the front side:
 
  - I can not see any damage.

The results on the images look good!

[AndrejaKo]

Wow! Great work on removing that conductive paint!

[Carrington]

Thanks. 

The acid is for use in swimming pools (conc. 25-35%), and the H2O2 can be found in pharmacies (conc. 10%).

Lemon should be HCL (conc. 35%) and H2O2 (conc. 7%).

I will continue with the process along this week. 

[AndrejaKo]

Small update from me: After reinstalling the firmware several times, it seems that the drift of channels when scope is off is now gone.  Also now it seems to be able to determine that there is nothing connected to probe end if autoset is pressed. I don't usually use this feature, so I don't know if that's something they added in a new firmware or nor, but I remember that this time last year, it couldn't detect if the probe was connected to BNC and the tip is floating.

Also now for some strange reason, I can detect the waveform of the main transformer of the PSU with a probe next to the PSU vents with back cover on. I think that this wasn't the case before, but I don't really remember holding the probe to the back of the PSU.

It seems that just as I fix one issue, two new appear.