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

[Siri]

I did not mean noise as I told that notice about battery. I mean if think linear regulator for 5.5V

Sorry, not very well know English.

I thought about it. In SDS used 2S2P 8000 mAh Li-Po battery (series connection of two parallel cells of 4000 mAh). According to the specifications of these batteries have a lower voltage of 3.0 volts per cell. Lithium batteries do not like deep discharge, and in order to save batteries it is better not to discharge low, at least up to 3.3 volts per cell. By the way, in my SDS7102 low level indicator (flashing battery icon) runs from 6.79 volts and below.

Thus, we assume that the lower the operating voltage of the battery is about 6.6 volts. If instead of a noisy step-down DC/DC converter to use a linear regulator to 5V low dropout voltage then such a change seems possible.

My pre-selection is LM1084-5.0: an output voltage of 5V and a current of up to 5A voltage drop no more 1.3 V.
But there will be an increase in the total supply current oscilloscope is estimated about 1.5 times.

[rf-loop]

I did not mean noise as I told that notice about battery. I mean if think linear regulator for 5.5V

Sorry, not very well know English.

I thought about it. In SDS used 2S2P 8000 mAh Li-Po battery (series connection of two parallel cells of 4000 mAh). According to the specifications of these batteries have a lower voltage of 3.0 volts per cell. Lithium batteries do not like deep discharge, and in order to save batteries it is better not to discharge low, at least up to 3.3 volts per cell. By the way, in my SDS7102 low level indicator (flashing battery icon) runs from 6.79 volts and below.

Thus, we assume that the lower the operating voltage of the battery is about 6.6 volts. If instead of a noisy step-down DC/DC converter to use a linear regulator to 5V low dropout voltage then such a change seems possible.

My pre-selection is LM1084-5.0: an output voltage of 5V and a current of up to 5A voltage drop no more 1.3 V.
But there will be an increase in the total supply current oscilloscope is estimated about 1.5 times.

Owon oskilloscope is watching batt voltage only for display.  Oscilloscope do not control and shut off battery! It is battery control logic inside Hangzhou Wanma made battery module. This main switch is  inside battery.  After module itself shut off  then battery module output go to  high impedance state but current can flow to inside battery using same terminals.

This shut off treshold is around 5.2V if measure it so that scope run with battery until battery module internal control shut off. (in these battery modules what I have tested.)
With this voltage it is still not "deep discharged" (becouse this voltage is of course measured with quite high load and there is battery internal source resistance - after loading stop (battery internally shutted off) of course internally battery voltage is higher after load is zero.

----
In new Adapter card there is 5.5V converter and TFT back light control. Both of these produce high noise.  3.3V converter is acceptable.   

Owon need  find solution for this problem what they have done. Also they need modify main Power unit for lower EMI.

Before Spring Festival I have talked with Owon about this problem and I have promised to Owon that also I will send full test documentation to Owon monday 18.Feb (they arrive back to work).
And "I have feel" that there are also some others doing same work.

Other problem after solutiion is: How to arrange repair for these units what allready have sold around of world.

This "engineer" who have made these modifications in Owon - there is not easy days waiting in work  - if he arrive back to work from holiday.

[Siri]

rf-loop, thanks for the information about the features of the battery.
I wonder if all batteries present more protective electronic module. Outwardly, it does not. Maybe it does and the desired balance in charge of cells? .. Interesting point. But in any case to bring the battery to such a low voltage and power off is not necessary. Oscilloscope itself has no off the low voltage, certainly a disadvantage.
On the main ADC/CPU board is squeaker. Let the developers though, in the new firmware would make a nasty sound alarm reminder low battery discharge. It will be a complementary solution to protect the battery.

The problem of "gnd-noise" in the SDS is, as they say Russian, a fly in the ointment. Trifle spoils in all an excellent device.
Unfortunate "needle" that is present on the ground surface of the device - it fronts 480 kHz switching step-down converter +5.5V. Need to change the scheme of this converter. Perhaps, on a scale the world has sold instruments some solution would be to add the electrolytic capacitors with ultra low ESR an input and output. We must try. But such necessary (and quality!) components are relatively expensive and scarce. This itself upgrade will be in this respect more difficult to install, for example than nylon column close to the first channel.

[akis]

Can we not simply disable the +5.5V completely and add a linear regulator or some such to provide the +5.5V ?  That would get rid of the 480KHz noise generator.

[Siri]

Can we not simply disable the +5.5V completely and add a linear regulator or some such to provide the +5.5V ?  That would get rid of the 480KHz noise generator.

That's what I plan to do in my SDS7102. This will be the fundamental solution to the problem. The problem is that so far have not found LM1084-5.0 in the local stores The voltage in the circuit of the oscilloscope in this place can be reduced to 5 volts, and the chip in my opinion would be the best choice.

The following quote method or identify GND-noise.

Disconnected and removed the back cover, the device is powered by a battery.

As a local probe is used without the ground wire-crocodile. Short (5-6 mm) needle of the end of a high impedance probe - good mini-antenna for local search. Bringing the needle probe to different segments of the device can see the local emission.
For example, the inverting converter -7.6V on board the PSU (chip MC34063) operates at a frequency of about 63 kHz. Emission at the surface of the chip:




And near the inductance of this converter:



Step-down converter +3.3V (it is set in the following scheme of the problem after converter +5.5V and powered by it):




Emission near inductance converter +5.5V 480 kHz (note the sharp current spikes):



Compare this timing (match the cursor and time stamps) with matching fronts pollution on the ground surface of the device. More significant for this would be the use of two channels, but I limited myself to one:



(Here running triggers on the rising edge and the waveform is slightly shifted with respect to current spikes in the converter +5.5V. But the hero of our "celebration" became known).

[akis]

LM1084-5 ? Do we really need 5 amps ?

[Siri]

LM1084-5 ? Do we really need 5 amps ?

Of course not. 5 A is the current limit for this regulator. In the scheme SDS7102 bus +5 V requires 2 A (+/-). I do not know exactly, but I assume this figure based on the total current consumption.
In 1084 when this current is less than the voltage drop (about 1,0V), unlike other, less powerful series regulators, where the voltage drop will be greater if we desire a current.

[rf-loop]

SDS7102 
It was RuiFeng ADC...
now  there is:
(picture is from SDS7102 1240xxx)

[Hydrawerk]

OK... What is that ADC IC? I could not find the datasheet anywhere.

[rf-loop]

OK... What is that ADC IC? I could not find the datasheet anywhere.

Old was: MXT2002 (RuiFeng China)
MXT2002 is National ADC08D500 "clone". (exept case. ADC:128pin, MXT:144pin))

Now there is National logo and (looks like customer part number?) as can see there is 128pin but....
Who knows.

Functionally looks like  as before.

[Evi]

My spoon of tar in a barrel of honey:
http://owon.forumup.com/about153-owon.html

[Siri]

In addition, the threads about ground noise bring their latest research results.
Once again, sorry for the English may incorrectly.

SDS7102V SN 1234XXXX
Viewed the “ground noise” correlations with switching power supply uses two channel oscilloscope.
Content is done at the same SDS7102, power produced by the battery.



The second channel is connected to ground and seen polution noise. Trigger synchronization starts on the first channel. Connected to the first channel "empty" probe. End (short needle) is used as a capacitive local antenna. When approaching the needle to the schema components can be seen local emissions. Both probes have a bandwidth of 250 MHz.




"GND noise" is already known to many owners of problem party of 7102. This chaotic sharp bursts, on closer inspection it is decaying burst with a frequency of about 100 MHz.





Below view of GND-noise emissions from various DC / DC converter power supply circuit SDS7102.

1. Converter -7.6 V. Board PSU, IC3 (chip MC34063), the frequency of about 63 kHz




Conclusion: there is a very weak ringing high-frequency of the rise edge.
I recall that in my 7102 this step-down converter is modified, reduced pollution from him (https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/msg192647/#msg192647).


2. Converter +3.3 V. Adapter board, U21 (chip ACT4060A), frequency of about 380 kHz:





Conclusion: there is a very weak overlap at the front switch. Impact is minimal.


3. Converter +2.8 V. Adapter board, the chip labeled "1780 TB51», frequency of about 1200 kHz:





Conclusion: does not bring influence.


4. Converter +8.8V. Scheme pulse LCD display backlight. Adapter board, U8 chip labeled "5126Q".
The emission of this schema has an interesting look. Short burst repetition rate of about 30 kHz. Frequency of the pulses themselves about 860 kHz (in this example, the brightness of the display is minimal - 20%).








Conclusion: The significant ringing overlap is from the fronts. Tangible impact. Because of the instability frequency converter introduced them part of the noise is a random character.


5. Converter +5.5V core supply. Adapter board, the chip is marked Z1094AI.
Synchronizing GND-noise with this converter explicit. Unfortunate "spikes" of noise is exactly the same with the rear switch fronts. There is also a previously identified high-ringing generation at frequencies around 100 MHz.







Conclusion: this converter provides the main pollution and creates "GND-noise”.

[rf-loop]

Nice work Siri.


Here can find link to these old .It is 1.17M zip including some images and part of explanation txt file. This is small public part from full documentation between me and Owon during inspection and for solve this terrible design error(s) what someone have done.  Also many other peoples have done lot of work for resolve this problem.  (but most of my recommendations how to handle this whole terrible case they just did not listen - they select road for destroy Owon reputation for long time. Pity.)

Still with or without this problem, with what ever oscilloscope it is very important how we use probes and how we design  how to connect to devive under test for reasonable clean signal. It need knowledge and it need experience.

[Siri]

Here can download some part of tests.
...

Big and excellent work, rf-loop!

Our results are very similar. Apparently the main sources of noise is a +8,8V TFT Backlight LED power supply and +5.5V step-down dc/dc converter.

[DrBob]

Siri and RF-Loop

A very big thank you for the detailed investigation into the power supply noise problem on the Owon SDS7102. I have done similar things, so know you have both spent many many hours investigating it, then more time to document it for the forum and as an assist to Owon to get a "fix" for it.

I sincerely hope that Owon reacts quickly and thoroughly to what may be a significant field repair. It is very hard to imagine that they have not known about this problem for a while, as they see the feedback from the worldwide distribution.

I would imagine that sales of new scopes will be substantially reduced until this is resolved.

In my case, I was about to order a SDS7102 from a US distributor, but there is no way that will happen until this is resolved. In my opinion, it would be difficult for any seller to service a product with this kind of a problem until the factory can provide him with a fix. It does not look like "just tighten these screws" will fix it - but we can always hope...

[tjille]

Is this a problem that all SDS7102 scope from Owon has or is it certain hw versions? And from what version in that case? Thanks.

[akis]

Are all these "faulty" components all situated on one replaceable card or is it more complicated than that?

Is there any merit in trying to replace just the 5.5V DC step down converter with a linear regulator or it will not achieve much?

[rf-loop]

Are all these "faulty" components all situated on one replaceable card or is it more complicated than that?

Is there any merit in trying to replace just the 5.5V DC step down converter with a linear regulator or it will not achieve much?

It is good to note that "Siri" tests have made with older version where adapter  This new version of Adapter board have better 5.5V.  There is no any need for linear power supply, exept of course if someone want do it for himself.  There is now specified 4 hour time with battery (after quite short time period after charging). No one want reduce this time. SMPS is only solution for this.

Then it is also good to note that probing method may reduce this noise in most cases. When people use oscilloscope it is also good to learn many things about probing devices/signals under test. In real world today there is lot of many kind of EMI noises all around what may externally add noise to signal.  Same methods reduce these and also oscilloscope own EMI.  It is good to note agen: There is not any extra noise in oscilloscope signal pathway! This Owon front end is still one of best in this price class if look its noise. Take some halfway of 2011 made SDS and new SDS there can see that analog front end noise is reduced really dramatically.

I have get one old SDS for compare to new one.

I have add 5 small capacitors and cut one trace in main PSU and with normal "bad method" probing whole noise is better than old SDS. With good probing practice/methods difference is big. This "noisy" new Owon beats this old "no noise problem" SDS just like 1-0.

But one note: Old owon front end "pink"noise level is so high that these kind of very small peaks can not detect at all becouse they are there under front end base noise level.

This all do NOT mean that nothing need do.

It means that noise level can reduce (and need reduce) to normal acceptable level by adding agen all noise related components what they have dropped out slowly after version to version and finally forget all out from new version (ecxept 5.5V supply is littlebit different case).
They need return  + add some new components and/or change some components type and in later versions change also PCB topology for reduce inductances and also some capasitances. Aftere then maybe these noise filtering components can reduce. (If do not make noise then it do not need isolate/attenuate! It need keep as main principle.) 

Special critical is inside every single SMPS. There need keep current loop(s) as short and low impedance as possible. And take care that keep "GND" current loops  tightly separate from each others and then use "one point" together method in final. Before "out".
 

Someone may think that low ESR electrolytics  --- yes of course need but they are not enough at all, they are for audio frequencies (and it is not this case), they have no role with 20 to 200MHz.  Adding some ceramic capasitors. Yes but types what really can handle RF.  And  200MHz in mind it is also critical where they are becouse PCB traces have inductance. In wrong place they may do situation even more bad.


Later it can do even better if littlebit change adapter board and main supply PCB routing / topology.

Here also very fun picture.

[akis]

I am not sure I agree about the "probing method" as you call it. This is not a user's fault.

I have a 20 year old 20MHz scope that has no trouble at all locking on and displaying very clearly the same traces that the Owon cannot lock on and display.

It is like buying a brand new car that can do 120mph but cannot take a corner at 20mph whereas your 20 year old car cannot do 120mph but has no trouble taking a corner at 20mph. You cannot possibly say this is the driver's fault.

So what we need is a way to fix the Owon. I would rather the Owon became a 20MHz scope that works than a 100MHz scope that does not. I do not think Owon will accept that there is a fault and I doubt they will start replacing all the old scopes. Imagine the costs of shipping all the old units back to China.

I think the best solution would be to find a way that we can DIY fix this by adding capacitors, maybe cutting out tracks like you did and so on.

[rf-loop]

I am not sure I agree about the "probing method" as you call it. This is not a user's fault.

I have a 20 year old 20MHz scope that has no trouble at all locking on and displaying very clearly the same traces that the Owon cannot lock on and display.

It is like buying a brand new car that can do 120mph but cannot take a corner at 20mph whereas your 20 year old car cannot do 120mph but has no trouble taking a corner at 20mph. You cannot possibly say this is the driver's fault.

So what we need is a way to fix the Owon. I would rather the Owon became a 20MHz scope that works than a 100MHz scope that does not. I do not think Owon will accept that there is a fault and I doubt they will start replacing all the old scopes. Imagine the costs of shipping all the old units back to China.

I think the best solution would be to find a way that we can DIY fix this by adding capacitors, maybe cutting out tracks like you did and so on.

Of course it need repair without any doubt!



---------------
Probing methods understanding is other important case. It is NOT solution to this problem. But also with it, after understand it, it may help now markable amount.  Good probing practices and knowledge is usually very very underestimated specially if we look entry level.

[Siri]

Hello rf-loop,
Good and useful notes.

Today I replaced the old electrolytic capacitor 470 uF 16 V (directly to the output of the converter to +5.5V) --> low ESR 330 uF 16 V by Hitano. Ground noise is not reduced. It seems the whole thing is really a circuit topology of pulsed power converters. I have told earlier that eliminate defects in the inverter topology -7.6 V on MC34063. There it gave some result.

Apparently it is (has become) a big problem for Owon. Requires effective and speedy solution. While each new oscilloscope sold Owon expands the number of potential claims. Apparently simple solution of recommendation such as "replace a capacitor(s) on this" will not work here, and it smells like at least the replacement adapter.. Will this step Owon?..

[rf-loop]

Hello rf-loop,
Good and useful notes.

Today I replaced the old electrolytic capacitor 470 uF 16 V (directly to the output of the converter to +5.5V) --> low ESR 330 uF 16 V by Hitano. Ground noise is not reduced. It seems the whole thing is really a circuit topology of pulsed power converters. I have told earlier that eliminate defects in the inverter topology -7.6 V on MC34063. There it gave some result.

Apparently it is (has become) a big problem for Owon. Requires effective and speedy solution. While each new oscilloscope sold Owon expands the number of potential claims. Apparently simple solution of recommendation such as "replace a capacitor(s) on this" will not work here, and it smells like at least the replacement adapter.. Will this step Owon?..

Even if there is as low as 1nH  and frequency is 100MHz, inductive reactance XL is 1.25ohm
But in practice there may be many times this inductance and some frequencies go up to 200MHz or even over. These "low ESR" electrolytic capacitors are good as long as there need handle low frequencies but they can not do nearly anything in this frequency area.  4mm long, 0.25mm copper wire is around 2.5nH. ( 100MHz XL is around 1.6ohm) Now I ask what is low ESR capacitor real impedance with 100 to 200MHz.)   How about PCB GND trace impedances with this freq?

(btw: I know Owon is now working for this)

[akis]

As customers, what in your opinion should we be hoping for from Owon? A complete replacement of the scope (very expensive due to shipping costs to China),  repair by local dealers (I am not sure if there are any local dealers in the UK), or maybe DIY instructions of how to repair it?

[Hydrawerk]

  Well, in 11/2012 I was thinking about buying an Owon... Today not at all. 

[Siri]

As customers, what in your opinion should we be hoping for from Owon? A complete replacement of the scope (very expensive due to shipping costs to China),  repair by local dealers (I am not sure if there are any local dealers in the UK), or maybe DIY instructions of how to repair it?

If it comes to the need to replace the adapter card or PSU, and the dealer does not (or can not carry out repair / replacement), both version will be sending everyone a new card for itself replace it. Owon save face and will gain more trust/respect.
You need only reflect on the mechanism to confirm the possession of the problematic device: such as sending them a foto of the oscilloscope with serial number, etc.