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

[bertchai]

Hi Tinhead,

Any suggestions on how to mod the Owon?   

[tinhead]

Last pic is Owon 435MHz test, not 145MHz as read in picture file name. test. With 5MHz freq mod and with 100kHz modulation freq (not meningful here.) Zoomed 10x. Middle is 436MHz (1MHz wrong if my signal did not change),whole pic area is 75MHz.

oh yeah, i like that picture. No chance to get anything similar on Hantek due autopersistency, time to patch?
Probably ...

Any suggestions on how to mod the Owon?   

not yet, currently i'm waiting for an S3C2416 board to continue my work.
There are some on ebay, but they bad designed. I found two companies producing what i need,
one of them not responding (as they probably on holidays), another one out of stock (have to wait extra week).

[Janne]

Thanks for the analysis on the 'scope. I've been thinking of buying the same owon model.. I could really use the battery option, which many other seem to lack. So I'll definitely read through the whole thread once I get the time =)

[0xdeadbeef]

I wonder: is it clear at this point if the SDS7102 is essentially the same as the SDS8102 and the SDS8202 (or even the SDS8302 or SDS9302) from a hardware point of view? Or do the models with 2GS/s and more have also different ADCs and input circuitry?

[tinhead]

I wonder: is it clear at this point if the SDS7102 is essentially the same as the SDS8102 and the SDS8202 (or even the SDS8302 or SDS9302) from a hardware point of view? Or do the models with 2GS/s and more have also different ADCs and input circuitry?

ADC on SDS8102 and SDS8202 must be different than on SDS7102, it should be MXT2001 instead of MXT2002.

On SDS8302/SDS9302 we can only guess what inside, but probably (in worst case) it is/will be still MXT2001
with overclocked clock to 1.25GHz/1.6GHz or maybe ADC08D1500 (in best case) - which i really doubt because
of the price (ADC08D1500 costs 500USD at 1000pcs and SDS8302 only ~1000USD)

Regards the frontend, the firmware looks universal, so you can assume the frontend will be the same.
You can actually even simulate based on the schematics that this works fine for 300MHz.
 

[duke3k]

There was a comment in another thread that a new SDS firmware should have been released end of last week or so.
Is anything known about it? Like: what has been fixed/improved and if it's freely available somewhere?

I believe the firmware update for the SDS7102 fixes the problems described in these posts:

https://www.eevblog.com/forum/index.php?topic=4597.msg62352#msg62352

https://www.eevblog.com/forum/index.php?topic=4412.msg62228#msg62228

[alm]

You can actually even simulate based on the schematics that this works fine for 300MHz.

You trust simulators on this? Where parasitic behavior may be the major contributor?

[rf-loop]

There was a comment in another thread that a new SDS firmware should have been released end of last week or so.
Is anything known about it? Like: what has been fixed/improved and if it's freely available somewhere?

I believe the firmware update for the SDS7102 fixes the problems described in these posts:

https://www.eevblog.com/forum/index.php?topic=4597.msg62352#msg62352

https://www.eevblog.com/forum/index.php?topic=4412.msg62228#msg62228

FW update have two things repaired. (also maybe some unknown minor things?)

1. Repaired bug in FFT frequency cursors if FFT window is x2, x5 or x10 zoomed and/or horizontally shifted. Cursors frequency info was totally wrong. Repaired.

2. Repaired waveform save to USB if used oscilloscope slow speeds modes.
This problem was only in slow speed modes, not high speed modes. (1)

Note: Current version of PC software can not open and/or convert these to csv.
Captured datapoints are ok in saved file. File header have some conflict with PC software somehow related to timebase and scope mode (slow speedmode/high speed mode).
Header is 69 bytes long. After this header start captured datapoints 1k, 10k, 100k, 1M or 10M bytes lenght. Every datapoint is 8bit signed value. File can open if littlebit mod header with editor. But there need change t/div value so then PC software displayed timebase is wrong. And also if convert to .csv file, there is then wrong timing info.
Owon need repair this PC software.

(1) Scope have two major time band. High speed and Slow speed.

High speed is 50ms/div -- 2ns/div (1ns/div in 8202 and higher)
Slow speed is 100s/div --100ms/div

High speed band works normally. Not affected in this memory, samplerate, save to USB issue.

Slow speed band have two different modes. Scan mode, Normal mode.

Scan mode is used if 100ms/div to 100s/div speed setting AND trig is "Auto"
- scan mode is like roll paper penplotter. It use slow samplerate. It is feature not bug. Display info about mem and samplerate is wrong, it is bug and still not repaired.

Normal mode is used if 100ms/div to 100s/div speed setting AND trig is "Normal" OR "Single"
- in this mode you need wait it have meet trigger and then capture rest of signal. (before trig id do of course hidden pretrig sampling loop) After whole data is captured it is displayed.
In this mode it use highest possible sampling speed related to selected amount of memory.
In this mode you can also stop shift and zoom using "Horizontal position" knob and "Sec/div" knob. (time shift is more fast if first zoom out then shift and then zoom in) In this mode you can also save waveform to USB (as mem selected 1k - 10M) but this mode saved wave file PC software can not open. (waiting fix)

----------------------------

About 8202 and higher.

Front end from BNC to ADC is enough for all these frequencies.

Maybe there is some minor adjustments in some resistor/capacitor value (or maybe not - maybe they use different front end, this can know only after inspect these models, but if they have done it with sanme front end, it is well possible).
This front end desig have not problems upt to 300 or even 400MHz if we are talking related to these things what low end  or even middle level oscilloscope need. (maybe shielding need be littlebit better between channels). This is so clean three voltage band  one road attenuator front end.
This can easy see also my pictures. Example in pics where is sweep up to 500MHz in FFT mode.
Also there is set of pictures with different frequencies sinewave, and pulses. (of course I have more)

With my own inspection I can tell (opinion) that changing ADC and littlebit adjust/reduce higher frequencies attenuation it can easy be 300MHz oscilloscope just with this front end. (in theory and in practice and with just this PCB) 

But what I hope is: in all models they change PCB design just around ADC. GND's layout need change and Power to analog and digiltal part of ADC chip need RF isolation. (it works but it can be better without money) Also they can littlebit develop some HF/RF shielding also it may do it better without money.  Also body where BNC's are assembled can easy be littlebit better.

Front end / ADC system meets very easy all frequancy specs without any changing. You have loook these Rigol/Atten/Siglent/Uni-Toy pictures around internet. This Owon is in captured signal quality far over these.
Also before I have tested example Rigol 50 and original 100MHz models. Owon is far above these in capture quality.
Also frequancy response is better. Response shape is better, this need also remember, not only look -3dB point.

Now this 100MHz scope step response is around 2ns rising time and -6dB point is around 270MHz.

And becouse there is not interlevaed style ADC there is not ADC timing and level match problems.

Only what need do is reduce higher frequancies attenuation!

If want 7102 freq response -3dB point over 200MHz it is maybe possible with small HW mod?
No need software knife for LMH6518. It looks that it is not set 100MHz by FW. It is maybe set to 200 or 350 becouse freq response is what it is. Look whope front end and LMH datasheet and response curves... my opinion is that it is set for 200 or 350MHz(?).  If it is so and as we know total curve it can think that around 200MHz it can rise some other way.

[tinhead]

You can actually even simulate based on the schematics that this works fine for 300MHz.

You trust simulators on this? Where parasitic behavior may be the major contributor?

no, by default i don't trust any simulator.

This time it is easier, when you look on the PCB picture you will
see that components from ch1 and ch2 are placed in different locations. So that means even with not exact
same mechanical design (and resulting parasitics) both channels are working smooth enough.
Additionaly the measurments done by rf-loop confirming my observations.

I haven't measured directly on PCB how the signal look like (let say up to 300MHz) right before LMH6518,
maybe rf-loop can do this (as i send back Owon).

[Chinakoch]

Hi there,
my SDS7102 arrived just yesterday. As I do not own a Windows machine, I decided to implement my own version of the PC software using Python and Qt. I found some specifications about the USB protocol here. Unfortunately, either when using PyUSB, or using the driver provided in the link above, the connection times out when trying to read from the scope. I guess OWON might have changed something about the USB protocol in the SDS series. Before I try to contact OWON and get some information regarding the (potentially) new USB protocol, it would be great if someone could replicate this finding.

If you own an SDS scope and a Linux machine, please try to run the already existing Linux driver. Alternatively, you could install PyUSB 1.0 and run the following code:

Code: [Select]

import usb.core

VENDOR_ID               = 0x5345
PRODUCT_ID              = 0x1234
WRITE_ENDPOINT          = 0x03
READ_ENDPOINT           = 0x81
CONFIGURATION           = 1
INTERFACE               = 0
START_COMMAND           = 'START'
RESPONSE_START_LENGTH   = 12
USB_TIMEOUT             = 2000

dev = usb.core.find(idVendor=VENDOR_ID, idProduct=PRODUCT_ID)
if dev is None:
    raise ValueError('Sorry, mate! No OWON scope found...')

dev.set_configuration(CONFIGURATION)

ret=dev.write(WRITE_ENDPOINT, START_COMMAND, INTERFACE)
if ret is not len(START_COMMAND):
    raise ValueError('Writing failed...')

data=dev.read(READ_ENDPOINT, RESPONSE_START_LENGTH, INTERFACE, USB_TIMEOUT)
print data
If you find any read-timouts, please get back to me!

Best,
  Chinakoch

[rf-loop]

This "guidance" manual is old version what can find there.
I will send later link to you for more new guidance manual if it helps anything. (updated manua know also SDS models)
Also in new manual have maybe something wrong specailly in part of data structure.
IMHO this driver is for old models. No one have been interest to update these becouse they are these "open procect stuffs"

[Chinakoch]

This "guidance" manual is old version what can find there.
I will send later link to you for more new guidance manual if it helps anything. (updated manua know also SDS models)
Also in new manual have maybe something wrong specailly in part of data structure.
IMHO this driver is for old models. No one have been interest to update these becouse they are these "open procect stuffs"

Thank you very much, rf-loop! My first guess was, that endpoints or the 'start command' might have changed. Anyway, I can not confirm this for the data given in the new manual. So it is maybe just my dilettante programming or other reasons I can not foresee right now...

[Chinakoch]

Using some magic software tools, I can now confirm, that my first guess was right and OWON has indeed changed the 'start command' in the USB protocol for the SDS. I am now able to get data from the scope. As I do not know if OWON wants details about their protocols made to be public, I will not post any details here.

[rf-loop]

Using some magic software tools, I can now confirm, that my first guess was right and OWON has indeed changed the 'start command' in the USB protocol for the SDS. I am now able to get data from the scope. As I do not know if OWON wants details about their protocols made to be public, I will not post any details here.

I think it is not problem publishe and share what you have your self find and know, I think it can make public.
Owon have been somehow open mind as they have be very helpful for develop this old linux driver.

Only what I ask is that do not publish/share this "new" guide as Owon published official  material becouse it is not officially give for share and it may have obsolete or wrong information. ( I think this new Guide is not yet ready...  and as we all know wrong information is more bad than no info...)

I think Owon (this is only my individual citizen opinion) is happy if someone help other peoples example so that linux driver can use also with this SDS

If someone want do full develop for open driver, I think if he/she do seriously this project also Owon want give help or some cnowledge for this work just as they have done before.

[0xdeadbeef]

Is it possible that the SDS series is a little picky about USB sticks? I received my 8102 today and tried to store screenshots and waveforms to an USB stick which failed more or less completely. Only in slow acquisition modes and for a 1k record length, the SDS wrote at least a (very small) BIN file to the stick, but the Windows software showed it was empty. For any higher record length in faster acquisition modes or when trying to save a screenshot in any mode, the saving never finished (or even started).
I tried to reformat the (16GB) stick in Win7 and XP (Fat32), but with no success. I also tried a 2.8" USB-HD (also Fat32) which behaved the same. Also another 8GB USB (again Fat32) stick failed.
Only when I reformatted the 8GB stick on my XP netbook, the SDS would finally store screenshots and waveforms to this stick (and only to this stick). I'd suspect the SDS is limited to certain block sizes or whatever, but this is a bit frustrating.
My FW is v1.0 btw. which is obviously not the newest version (there was an v1.0.1 on devices bought in August) . I already contacted Owon and hope that they can provide me with a firmware that has a little more stable USB support.

[rf-loop]

Is it possible that the SDS series is a little picky about USB sticks? I received my 8102 today and tried to store screenshots and waveforms to an USB stick which failed more or less completely. Only in slow acquisition modes and for a 1k record length, the SDS wrote at least a (very small) BIN file to the stick, but the Windows software showed it was empty. For any higher record length in faster acquisition modes or when trying to save a screenshot in any mode, the saving never finished (or even started).
I tried to reformat the (16GB) stick in Win7 and XP (Fat32), but with no success. I also tried a 2.8" USB-HD (also Fat32) which behaved the same. Also another 8GB USB (again Fat32) stick failed.
Only when I reformatted the 8GB stick on my XP netbook, the SDS would finally store screenshots and waveforms to this stick (and only to this stick). I'd suspect the SDS is limited to certain block sizes or whatever, but this is a bit frustrating.
My FW is v1.0 btw. which is obviously not the newest version (there was an v1.0.1 on devices bought in August) . I already contacted Owon and hope that they can provide me with a firmware that has a little more stable USB support.

Also there may be "slow speeds" bug (USB save bug) in your scope, same as old 7102 FW have and what is later repaired. (also same time repaired FFT frequency cursors). Also you know that slow speeds have dividet to two different modes, scan mode and "normal" mode. (scan mode is default).

Btw, where from you buy "old" scope? Why you do not contact to seller for update?

It is maybe very frustrating to Owon that there are sellers who do not care anything and then lot scope owners need direct help from factory. I think soon Owon need make some rules for dealers - who is certified dealer and who is not. It can not be long time so that sellers sell without any customer care  selling instruments just like milk bottles, only get money and then all customer care need come directly from Owon.

Sellers need do customer care work also, before and after sales. Lot of sellers do not care anything but get money  - easy money.  Then we see lot of scope owners who hang around internet finding who can help me...who can help me.. this is "bad sellers problem" and also these bad sellers push prices down -  maybe they are brainless.
There are also sellers who care and give after sales customer care service, but they are not maybe cheapest ones.
If you just get your scope from seller and it is very old FW - why seller did not update it?


 

[0xdeadbeef]

Well, my main interest was if there is a known bug with Owon's SDS line regarding USB sticks in general (i.e. support only for small sector sizes or whatever). Since to me it looks as it couldn't store files >8kb on my larger (>8GB) USB drives.

Regarding my "old" scope: I guess the 8102 sells much less frequent than the 7102, so it's not too surprising that it doesn't have the newest firmware. Please also take into account that the version numbers between 7102 and 8102 seem to differ since the SOP of the 8102 was later AFAIK. E.g. at the time v2.1.1 was out for the 7102, I understand the 8102's firmware was at revision v1.0.1. Let's see what revision the newest firmware will have (if I get it), but I would guess it's something like 1.0.2 or 1.1.0.

I also don't really expect a seller to open every box and reprogram the newest firmware. This is not common for any consumer product, so why should it be different for a low cost scope? Besides people would complain about opened boxes. IMHO You could only expect that kind of service when buying a >10k€ scope directly from LeCroy or similar.

I'd gladly update the firmware if Owon had a public download site, but as they don't, I have to bother them. So this clearly caused by Owons strange strategy of giving you updates only if you demand them and not the seller's fault. So maybe bothering puts them back on the right trail.

Besides, I also contacted the seller (which claims to be the only official Owon seller in my country), but they didn't react yet (as didn't Owon). I sure hope though that one of them will finally responds and provide me with a newer firmware revision...

[0xdeadbeef]

Since I don't want to mess the "good choice for first oscilloscope" thread, I'll answer this here:

Local authorized Owon dealer is good place to ask update. Everyone have buy it from some place. Very simple, ask seller. And more simple, before buy, look littlebit wherefrom buy.  It is not so bad if buyer also use his brains when select wherefrom buy.

As mentioned before, I bought from the only official Owon dealer in my country and neither did they respond to my mail nor did Owon up to now. To be fair, I have to mention that I contacted them last Wednesday, so it's not like I already waited for weeks, but they could have answered on Thursday and Friday and didn't. Furthermore I have to mention though that the Owon customer service answered another of my mails some time ago, but they needed a week or so, the answer was in pretty bad English and it answered only half of my questions. As this is more or less the industry standard, I wouldn't go so far to say that the customer service is completely absent, but it's not really perfect either.
So neither do they react quickly nor is it good practice to keep SW updates from the customer. Heck, large comanies as Agilent and LeCroy encourage you to update the firmware even for their highend equipment, as do smaller companies as Rigol, so why can't Owon?

Owon customer support and customer care is one of most good what I know. Always nearly immediate answer and you can fax, email, phone call etc.

Then you obviously never tried the service of LeCroy or even Agilent. They usually react within hours.

There is no any reason for public open dowload place for FW updates or repair patches.

Yes there is. If they release buggy software, they should also give you the updates without letting you beg for it.

Buy from grey ebay seller who sell teddybears and oscilloscopes and parfumes... you really think it is clever to buy. Then waiting customer care.

Dunno how often I need to repeat that I bought from the only official Owon dealer in my country. And no, they don't sell anything but measurement equipment. So why do you keep insisting it was my fault that Owon delivers buggy firmware and neither Owon nor the dealer offered me an update up to now? It's not like I already gave up. But a good customer service is something else.

Still, even taking into account the existing bugs and the problems to get a firmware update, I don't regret that I bought the SDS8102. I don't think you can get a much better scope for the money at this moment and there's the chance that it can be patched to 200 or even 300MHz in the near future. But even at 100Mhz, the great screen and the 10M record length IMHO make it a clear winner.

[0xdeadbeef]

Another thing just for the sake of completeness. At the begin of this thread, marmad said that ...

It takes exactly 4 minutes to save the full 10Mpts to the USB hosted device (during which, of course, the scope is doing nothing else).

This doesn't seem to be true for my 8102, and thus it shouldn't be true for the 7102 either. The saving of one 10M ch needs about 27seconds on a relatively lowish class 4 microSHDC (in an USB adapter). The created file is 9766KB (or 10000069 bytes) in size, so the transfer rate is about 361 kByte/s or about 2.83 MBit/s.

Since a class 4 card should reach a minimum of 32Mbit/s when writing, the card is obviously not the limiting facor. So yes, saving to an USB stick is very slow and doesn't even nearly reach the limits of USB1.0, but it's still about 8 times faster than stated by marmad.

BTW: it's a bit weak that you can't save both channels at the same time. Also the "copy" button always saves the waveform, even if you defined saving to bitmap in the "save" menu.

[rf-loop]

In this case I did not at all compare to Agilent or Tektronix. LeCrou  have never touch.
I have repaired Agilent and Tektronix equipments long time and I know something about they service (but mainly only local).
Well... today they have _some_ updates downloadable. Still I hope get some updated FW to my nice spectrum... oh well but but product is designed to be obsolete... so I can not get. I need copy it from some other. I know Agilent have it... but no one there want find it and give it.

Chinese companies are just started learning how to arrange customer care and after sales services.
They are just started this long travel for learn how to make better. 

But what I mean, Owon is good also for customer care if look many Chinese factories.
I have try example Rigol. No any answer. Same Atten (Siglent), try to get UNI-T FW update...

I have seen timetable how Owon answer. I know there is also language problems and also, some workers are more loosy than some others so quality of answers may vary.  I think they marine radar customer service is better..  maybe.
Example short time a go China have several days holiday. Now it take some time to work with this queye of emails and faxes.
If I try Agilent answer in Christnmas... maybe I need also wait some more hours.. 

After soon (some days) I know if SDS8102 or 8202 have  factory updated FW.

In this price class 7102, 8102, 8202
they  are amazing good scopes. Not so much features in FW but, quality of singnal sampling is good. Display is special good. (and if someone really want it can also make more bright becouse backlight is driven well under specs)
Full speed 10M capture memory is  really good. If look slower t/div it give really big advantage in sampling speed if compare what ever this price class scopes or even much more expensive Agilents.
Turn Agilent expensive 100k memory scope to 1ms/div and look what is realtime samplerate, it is like bicycle compared to phantom.

Btw, my language is bad. I did not mean at all you. In most things what I have write I mean universally peoples who buy oscilloscopes or other things from some sellers around of world. There need be careful. There need use braind and not only paypal pay now button. Example in China, there are lot of cultural things... buyer is responsible... not seller (caricatured). Seller is clever if he can sell bad things with good price... buyer is just stupid as pawn if he buy this. Buer need shame... seller can be broud. 

If people buy bad things, this is not manufacturer... people need go to mirror and ask who is guilty. There is common that if someone sell something it need always think (or suspect) that all is lie. You buy chicken. Ask they first empty stomach...  there may be cup of rocks... if you buy electric wire... ask seller to show that inside really is copper wire... not only plastic tube... if you buy ten eggs, do not give seller select them... and ask you break one random.. so you can see that they are real chicken eggs and not human made lie eggs..   these old things are still affecting... in global world and buying without posibility to real look, touch and test this old system do not work.
So, many Chinese companies need do extremely hard work for learning how to get peoples trust. If there is 1000 satisfied customer they do not make noise over internet every corner... if there is one unsatisfied he may do big big noise what are readable in each mud hut anywhere in the world.

I can ask: Why Hantek have need publishe  "many" updates and repair patches inside less than one year? Then sharing this bullshit... this we want? Public bullshit sharing. If Hantek do not immediately stop this stupidity they really can loose all!
Please hantek, do buglist, do changelog and start do work in good order. I think these young couple of boys there in Hantek need really someone watch over them. Give commands what to do how to do and when to do. In good order and with tight programming rules. Do not any change before it is a proofed it need do, tested before implement, do final tests... carefullu and keep good test diary. This game need whistle break now - or vthen newer.

I hope no one try push Owon to this road. What is right road, it need carefully design. Today this system (updates, repairs, customer care, after sales services etc) is not mature, but they still have done amazing good work and all signs show that they also learn and they are interest to learn to be better.

[0xdeadbeef]

Sure, I agree that the Owon's firmware - despite of the USB saving issues - is very stable. If I believe marmad's video reviews (and I tend to do so), Owon's firmware is even much more mature than that of Rigol and Hantek/Tekway.
Up to now, the SDS never crashed on me, always showed plausible waveforms and I didn't find any major issue apart from the USB saving. So of course I favor a stable firmware update now and then over a beta patch every 2nd week were each patch causes new issues.

Also, honestly I didn't expect the firmware to be completely bug free. Indeed out 8k€ Agilents in the job crash now and then and e.g. suddenly don't respond to the vertical position knob any more. Or they switch to AC mode (displayed via LED) though they claim (in the menu) they are in DC mode. Even the 14k€ LeCroy crashes in some (admittedly high endish) analyzing menus and after the last firmware update, the front panel was dead and I had to fiddle around in the system registry to make it work again.

Still, I would really appreciate if there was a public firmware download with a good history on Owon's web site. There is nothing wrong with making mistakes. They just need to be documented and fixed. Not offering public downloads and change history might have other reasons (like trying to avoid fw hacking), but usually it stems from the idea that making bugs is something to be ashamed of and therefore trying to conceal it. This would be a pretty bad attitude from a quality management point of view.

[hacklordsniper]

Thats better than my DSO-X 3000 which never works as it should. (If i even manage to get it to boot)

[somlioy]

I've been thinking of buying this scope for a while, but I can't find a proper place to buy it. Most places I find seems dodgy and yeah, I'd really like a Owon Certified dealer. Does anyone know good dealer that I can trust?

[0xdeadbeef]

You could ask Owon directly for a dealer in Norway.

If there isn't one, you could import it from another EU country.
E.g. www.batronix.com has a pretty good reputation, though they are no official dealer AFAIK.
In the UK, Owon has an own online shop: http://shop.owon.co.uk/
There is also the guy from Finland in this forum: https://www.eevblog.com/forum/index.php?topic=4597.0
In the Netherlands, you could try e.g. http://www.eleshop.nl/nieuwe-serie-owon-sds7102-p-480.html
Or in Germany there's e.g. this dealer: http://messgeraete-chemnitz.de/tisch-oszilloskope/index-2.html

[tinhead]

Up to now, the SDS never crashed on me

as you have SDS8102, try this one:

setup you DSO to whatever settings (default or something specific, didn't matter)
open FFT
do STOP
do Autosetup

and you scope should now "play guitar together with Kurt Cobain"

It was always killing SDS7102 (no matter what firmware), would be interessting to hear what on other models.

EDIT: to play fair, Owon knows this bug since maybe two weeks max, so be patient if your scope have same issue.