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

[lemon]

rf-loop this is it.
Many thanks for the link 

[TomC]

Here is a link to the uncompressed version of the Version 3.2 Adapter and New PSU photos. This link has an expiration date of November 2013. Until then the photos will be available for viewing or download.

http://sdrv.ms/1cOVr1M

Please have in mind that it demands account to login.

Sorry for the bad link, I revised it on the original post and here is the correct link also.

http://sdrv.ms/15hxusH

[TomC]

where is the ferrite? I have a ferrite flat cable.
I have not tried its effect.

I don't know if the Adapter Ver 3.2 had a flat ferrite but mine Ver. 3.0 hadn't.

Please look if there are and others ferrites inside.

My scope, with adapter version 3.2,  also has a large flat ferrite on the TFT screen cable.

[lemon]

Now it is not strange, is a thriller!

Can someone to give us a photo of adapter 3.3 from under side?

[Flash2light]

This is ver 3.3 dc-dc board under side.

[TomC]

This is ver 3.3 dc-dc board under side.

That's identical to mine (3.2), but yours had the smd capacitor under C34 installed at the factory.

Flash2light, can you read the markings on U2 on your board, on mine I see OBT 2?K CX10. The question mark is a character that I can't make out.

[TomC]

This is ver 3.3 dc-dc board under side.

That's identical to mine (3.2), but yours had the smd capacitor under C34 installed at the factory.

Flash2light, can you read the markings on U2 on your board, on mine I see OBT 2?K CX10. The question mark is a character that I can't make out.

I think I answered my own question about U2, I just found this on the internet.

http://www.s-manuals.com/smd/ob

You can download the datasheet for OBT marking from this link.

[BBAAHHOO]

Modified AC-DC power supply.

[lemon]

This is ver 3.3 dc-dc board under side.

That's identical to mine (3.2), but yours had the smd capacitor under C34 installed at the factory.

Flash2light, can you read the markings on U2 on your board, on mine I see OBT 2?K CX10. The question mark is a character that I can't make out.

I think I answered my own question about U2, I just found this on the internet.

http://www.s-manuals.com/smd/ob

You can download the datasheet for OBT marking from this link.

Nice TomC.
Now it is clearly, an other big difference btw ver. 3.0 and 3.2 is the IC for backlight led. It is completely differente circuits.

Your mystery is like "Mitos of Ariadnes" (from Greek Mythology) same psu, identical adapter but the noise is differente, but remaining the answer of Flash2light what is marking of U2 in ver. 3.3. Probably, there is a difference there...

[lemon]

Modified AC-DC power supply.

??, in your photos there is a difficulty to see someone clearly, because the pcb had a lot of residual of soldering.
You can removed them with a lot of isopropyl alchool.
I don't know if the flux that containing in core of your solder is water soluble (that means that is removed easy with water) or you used a lot of extra paste of flux, but usually fluxes are RA (rosin acid) and their residuals makes a corrosion in time.
It is better to removed with an appropriate solution.

[Carrington]

The design of version 3.0 is crap, continuous gnd plane cuts, parallel capacitors that resonate at frequencies where noise is generated, DC/DC Boost with input connected to critical places (everyone should know that at the input is where more noise is generating by a boost), etc.

I tried to explain to OWON that for those who have the 3.0 version a 1n capacitor is not enough, but there is no way. 

Now it is clearly, an other big difference btw ver. 3.0 and 3.2 is the IC for backlight led. It is completely differente circuits.

... remaining the answer of Flash2light what is marking of U2 in ver. 3.3. Probably, there is a difference there...

For this reason they changed the G5126 by a TPS61161A. The G5126 don't support more than 5.5V at the input.
And if I'm not mistaken lack a capacitor that in combination produces resonancere. True? Look near C45 in the 3.2 and 3.3 version.

[Flash2light]

I can get better pictures until Monday.

[lemon]

Flash2light, we'll waiting for that because it is not clearly the photo at U2.

From what I can see it marks like:
OBT
24J
PDW?

The TomC reference was:
OBT
2?K
CX10.

But if you search to internet for TI TPS61160A (Led Driver) the icon of IC is differente to these.


It is meaning that they are compatible to TI IC made to china of course.
I don't have in front of me an adapter Ver. 3.2 or 3.3 but you can check the following diagram from Texas Instrument. It is typical diagram for TPS61160A with fixed and dimming network.
I don't see at the pcb routing how they done the dimming network of leds.

[lemon]

The design of version 3.0 is crap, continuous gnd plane cuts, parallel capacitors that resonate at frequencies where noise is generated, DC/DC Boost with input connected to critical places (everyone should know that at the input is where more noise is generating by a boost), etc.

I tried to explain to OWON that for those who have the 3.0 version a 1n capacitor is not enough, but there is no way. 

Now it is clearly, an other big difference btw ver. 3.0 and 3.2 is the IC for backlight led. It is completely differente circuits.

... remaining the answer of Flash2light what is marking of U2 in ver. 3.3. Probably, there is a difference there...

For this reason they changed the G5126 by a TPS61161A. The G5126 don't support more than 5.5V at the input.
And if I'm not mistaken lack a capacitor that in combination produces resonancere. True? Look near C45 in the 3.2 and 3.3 version.

Exactly and this is the answer why they used the output of AOZ1094 for input voltage to G5126 and not the 8.4V direct from PSU like Ver. 3.2/3.3.
But at the second paragraph I am loosing you. The C45 is a capacitor to AOZ1094 circuits (+5.5V). What is relation with the Backlight Led Circuit of Ver. 3.2/3.3?

[Carrington]

https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg264016/#msg264016

But at the second paragraph I am loosing you. The C45 is a capacitor to AOZ1094 circuits (+5.5V). What is relation with the Backlight Led Circuit of Ver. 3.2/3.3?

Ok, it's just that in 3.0 version there is a small SMD capacitor in parallel with C45, these resonate at frequencies where noise is generated by the DC/DC. In the new version it (if I'm not mistaken) is not populated.


To design is better to have a minimum of experience, if not, then these things happen. I keep wondering how many "slip" more ...

The simplest thing that I can think to see how much noise is generated on the main board is power it by a set of batteries (5.5, 3.3, 8.4, -7), except the power of the screen, where we must use the appropriate driver (but is easy also isolate it with a battery).

[TomC]

Flash2light, we'll waiting for that because it is not clearly the photo at U2.
From what I can see it marks like:
OBT
24J
PDW?

The TomC reference was:
OBT
2?K
CX10.

But if you search to internet for TI TPS61160A (Led Driver) the icon of IC is differente to these.


It is meaning that they are compatible to TI IC made to china of course.
I don't have in front of me an adapter Ver. 3.2 or 3.3 but you can check the following diagram from Texas Instrument. It is typical diagram for TPS61160A with fixed and dimming network.
I don't see at the pcb routing how they done the dimming network of leds.

The lines under OBT are not indented, probably batch number but not related to device type. The attachment is from TI website. There is no room on IC for markings in the pictured icon, probably just a picture for show, but not the way actual IC looks like.

[Carrington]

Try here: http://www.ecadata.de/searchnew/index2.html

Bad luck does not give a result, but it is clear, SMD Marking:
  OBV = TPS61160A
  OBT = TPS61161A

[lemon]

https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg264016/#msg264016

But at the second paragraph I am loosing you. The C45 is a capacitor to AOZ1094 circuits (+5.5V). What is relation with the Backlight Led Circuit of Ver. 3.2/3.3?

Ok, it's just that in 3.0 version there is a small SMD capacitor in parallel with C45, these resonate at frequencies where noise is generated by the DC/DC. In the new version it (if I'm not mistaken) is not populated.


To design is better to have a minimum of experience, if not, then these things happen. I keep wondering how many "slip" more ...

The simplest thing that I can think to see how much noise is generated on the main board is power it by a set of batteries (5.5, 3.3, 8.4, -7), except the power of the screen, where we must use the appropriate driver (but is easy also isolate it with a battery).

Yeap, if I remember well this capacitor that missing now on the Ver. 3.2/3.3 was C36 a small one.
From the works of Russian member BBAAHHO seems that the mainboard has a very little noise. With his modifications achieves noise about 30mV that is perfect. 

Seems that this oscilloscope has designed from different groups. The works for the mainboard is quite good. There is no noise at the front end and the bandwidth is better than the specifications. The badly was at the psu and adapter board, here the Owon designs a crap and after one year and a half seems to resolve the badly design of the two boards.

[lemon]

Flash2light, we'll waiting for that because it is not clearly the photo at U2.
From what I can see it marks like:
OBT
24J
PDW?

The TomC reference was:
OBT
2?K
CX10.

But if you search to internet for TI TPS61160A (Led Driver) the icon of IC is differente to these.


It is meaning that they are compatible to TI IC made to china of course.
I don't have in front of me an adapter Ver. 3.2 or 3.3 but you can check the following diagram from Texas Instrument. It is typical diagram for TPS61160A with fixed and dimming network.
I don't see at the pcb routing how they done the dimming network of leds.

The lines under OBT are not indented, probably batch number but not related to device type. The attachment is from TI website. There is no room on IC for markings in the pictured icon, probably just a picture for show, but not the way actual IC looks like.

Unfortunately, it isn't TI IC, it is clearly a copy made in China with the same mark (OBT). If you look at the end of datasheet pages, the TI has drawn the schema of this chip. It is exactly the same like the photo that upload before.
But, for me it no matter about this, they have copy exactly the circuit of TI (perhaps the TI have better characteristics), the important is to check how the Owon makes the routing and what components choosed vs datasheet.
The rf-loop had made a lot of modification to this part of adapter and somebody had writes the phrase "which is design this board, he was drunk!"

[Carrington]

  I was, and the exact phrase was: "Who designed this? He was drunk?"
But who knows the reason... Maybe he did not know what he was doing or did not care...

[Carrington]

Seems that this oscilloscope has designed from different groups.

Wonder!

The works for the mainboard is quite good.

I want to believe! But I think not.

There is no noise at the front end and the bandwidth is better than the specifications.

Better BW, deliberate I think and, very easy to get.
There is no noise at the front end, wow if this is true then OWON you have been very lucky.
To get that Rigol has used lots of filters (ferrite bead / EMI Filter / LDO / shielding) and good design technique. But still have ~ 1mV of noise at full BW and in the worst case.

The badly was at the psu and adapter board, here the Owon designs a crap and after one year and a half seems to resolve the badly design of the two boards.

Okay, but not covered by the warranty.

[rf-loop]

Better BW, deliberate I think and, very easy to get.
There is no noise at the front end, wow if this is true then OWON you have been very lucky.
To get that Rigol has used lots of filters (ferrite bead / EMI Filter / LDO / shielding) and good design technique. But still have ~ 1mV of noise at full BW and in the worst case.

Of course it have noise in front end.
Here you can find data.

It do not loose with Rigol. (exept that Owon lowest full BW is 5mV/div.
Rigol have 0.5mV/div (with more than two div peak peak noise.).
But then, if reduce BW with filter or with averagind (capture averaging or high-res averaging) there can get slow and low signal very nice)

Also this noise level in front end (and BW) is proofed by many FFT tests what also are available.
(there can find test where full not clipped level is displayed and then signal dropped 60dB and signal can still rise very clearly over base noise. If this front end noise level is bad, it really can not do it.

Also we all know that Rigol front end is nearly like as copied from (older and littlebit noisier version) Owon.
Also it is good note that Owon real BW is over 170MHz typically (and over 200MHz typically in voltage band II) when look front end noise, BW need also note. 

My strong suspect is also that maain board designer is different people than PSU and perhaps also adapter board.  Specially PSU design "signature" is so different hat same people can not do this. Or he really have been drunk when he draw PSU.

[lemon]

Carrington about BW, you know (I think) the works of rf-loop, he has show us that this oscilloscope has over 200MHz bandwidth.

If I remember well you have additional the Rigol 2000 Series, I think that is one category over than SDS7102 and it is not fair to compare the two of them. About Rigol DS1052 or DS1102 that is the competition.

But if I remember well, the noise of SDS at full BW (5mV) is under 1mV. At the beginning when the noise gnd is very high (160-200mV) I couldn't to examine a signal with a few mV, it was full with artifacts and not stable - only with my old analoge Hameg I could.
Today with my mods (I have a noise gnd 40-55mV peak to peak) there is no any difference between Hameg and Owon in signals with a few mV, very stable and clearly.

For the warranty about gnd noise as very right you said, all of them give them many 

[lemon]

Oops, me and rf-loop write at the same time...

[TomC]

Your mystery is like "Mitos of Ariadnes" (from Greek Mythology) same psu, identical adapter but the noise is differente, but remaining the answer of Flash2light what is marking of U2 in ver. 3.3. Probably, there is a difference there...

Well that's not too bad, I'll be more scared to look around for the answer if it resembled the story of Medusa, it wouldn't be cool turning to stone.

There are a couple of things I've been thinking about. I took a real good look at my new PSU board and discovered that one of the new decoupling capacitors is probably missing. There is no capacitor where C23 is supposed to be. C23 is supposed to be at the end of the adapter board connector. I see glue residue, as if the component was there but fell off during the soldering process. However, I don't think this would cause such a big difference in noise by itself.

Flash2light, can you tell me if this capacitor is present on your PSU board?

The other thing I've been thinking about is that I don't have the battery option. With a battery installed you have additional filtering for the 8.4V supply. The battery itself can act as a super capacitor. I don't know for sure if the results from dgmoore78@yahoo.com and Flash2light were obtained with the optional battery installed.

dgmoore78@yahoo.com and Flash2light, can you clarify this, and if possible, if your test was done with the battery installed, can you do the test again without the battery?