EEVblog #480 - 300MHz Owon SDS Oscilloscope

[EEVblog]

Dave takes a look at the Owon SDS Series "Smart" oscilloscope.
The 300MHz model SDS9302. But apart from bandwidth and sample rate, the same as the SDS8302, SDS8202, SDS8102, SDS7102, and SDS6062.

[LaurenceW]

Come on, Dave, get off the fence.

Did you like it, or not?

(Sorry, i dozed off)

[Hydrawerk]

When Dave touches the LCD, it's colors change, so there is no protection glass in front of the LCD!
And see this: https://www.eevblog.com/forum/testgear/review-of-owon-sds7102/

[Hydrawerk]

Frozen. http://youtu.be/RmJ2pZloW18?t=35m6s
Well, the Owon SDS series was introduced in October 2010. This particular SDS9302 was introduced in 2012-06-06.
http://www.owon.com.hk/Newshow-en.asp?id=79&BigClass=Company%20News
Well, Owon scopes are for those people who want a cheap scope with large bandwidth but nothing else...

[marmad]

Thanks, Dave.

You might have been having a problem with the Alternate trigger level on CH2 because you were essentially creating a loop with the Trigger Out. It did make me stop and wonder what logic they use to determine which trigger (CH1 or CH2) is driving the Trigger Out when in Alternate mode.

BTW, you don't actually have to use the alternate trigger for the update rate - just running the Trigger Out back into Channel 2 and using the counter is enough. You can do that on the Rigol as well (although, of course, you'll get the waveform update rate of both channels running).

[Hydrawerk]

And yes, thank you for the review. Whoever wants a cheap scope, buy rather Siglent, Rigol or maybe Unitrend...

[don.r]

More Chinese scope fail.... as if the world doesn't have enough. My Tek CRTs look better and better each day. 

[marmad]

Well, the Owon SDS series was introduced in October 2010.

No, the SDS series first started being sold in the West in July 2011. The one I had in August 2011 was gotten via a group buy organized by an equipment shop direct from the Owon factory. 

[Hydrawerk]

was introduced in October 2010.

Maybe for China market. http://www.owon.com.hk/aboutus-en.asp?id=44

Milestones In Development:
Mar 2006  HDS1022M—the first high quality 2 in 1 handheld DSO made in China with high resolution color LCD display
Sep 2006  PDS5022S—Big 7.8 inch color LCD display bench DSO, ideal for education
Nov 2006  HDS2062M—60M handheld DSO follow up to the success of HDS1022M
Jun 2007  HDS-N series—the upgraded version to the original HDS series
Nov 2007  MSO5022S—Mixed LA-supported DSO
Apr 2008  PDS7102T—100M bench DSO
Dec 2008  OWON is named as “the most competitive price with high quality products” by “WIRELESS” magazine in China
Jan 2009  MSO7102T—Mixed LA support DSO with 100M and 1G real time sample rate
Apr 2009  Innovate independent for Auto measurement and Max. 20 measurements to apply for all items
Oct 2009  HDS3102M-N—first 100M handheld DSO made in China
Oct 2009  HDS1021M—1 channel 20M low cost handheld DSO for niche market
Jan 2010  MSO8102T—Mixed LA support DSO with 2GS/s real time sampling rate
Feb 2010  MSO8202T—200M Mixed LA support DSO
May 2010  PDS8102T—100MHz DSO with 1GS/s sample rate, 2M record length
May 2010  PDS8202T—200MHz DSO with 2GS/s sample rate, 2M record length
Oct 2010  SDS series in Ultrathin design with 10M record length
Oct 2011 The company passed ISO9001 quality system certification.
Aug 2012 SDS5032E-2nd Generation of PDS

[Dave]

The timebase knob is marked with siemens and nanosiemens.

[c4757p]

Wow. Dave, you need to put that back in its box right now - I can smell this turd from all the way out in New York. It's really stinking up my living room.

[Hydrawerk]

What's that "None is saved" message?

[c4757p]

What's that "None is saved" message?

It's a message from Owon, actually. They've kept all the good oscilloscopes to themselves and saved none for you.

[Eonir]

What's that "None is saved" message?

I think it means "All are Doomed".

[Corporate666]

Why don't these companies bother to have someone who actually knows English help them with the text?  It's one thing to have Chinglish in a user manual, but on the outside of the box?  It would cost them so little to pay some high-school English major (English speaking native) from any University on e-lance to get it right... but they let Ding Dong who works in the packing room and claims to know English do it instead.

Releasing a product to the world is a *big deal*.  Skipping out on the part of that project that everyone will see first is like getting dressed for your wedding and wearing only your underwear down the aisle because you couldn't be bothered to put your pants/trousers on. 

The first thing it says to me is "if they don't even give a shit about getting one of the most visible things right, why would I think they'll spend any time on the important stuff inside that I'll never see?".

[quarros]

Dave. The link to the forum, in the blog post is not working. You misstyped something 

[rf-loop]

I see one odd thing.
This individual scope was made one year ago.
(how did you find this old unit? I understand if you have made this video nearly one year ago but... )

First I regognize it from back feets. There was no rubber pieces installed at all.
Then I look more carefully and it looks like manufacture lot have been 122? (year week)
Perhaps old FW? (I can see some bugs in video)

In zoom mode it also increase samplerate. Many oscilloscopes with windowed zoom do not use higher samplearate in the zoomed area in window mode.

In test with glitches. It looks like (but too small amount of time in viedo so it is now only "feel") it sometimes trig to glitch more fast than if think update rate is always around 35 or less per second with long real visible blind time. (also this is proofed in tests using "marmad's double pulse method")
There is not any kind of tests for front end noise levels, no frequancy response test, no rising time test, no measurements accuracy tests, not trigger jitter tests, no FFT mode dynamic range tests. Total lack of any kind of lab tests for signal capturing quality. Only some waveform update rate test but also no any comment why it show this glitch more frequently that if it is really always under 35wfrm/s or less.  Perhaps secret is in table what is published after tests with marmad method.
But, as all know it is not made at all by DPO principle. It IS pure DSO.  (ref. Tektronix XYZ)
This is not for glitch hunting.
Time when this scope have made there was not example Rigol DS2000 series afaik.
So, comparing to much more new designed scopes is not apples to apples.

About vertical fine adjust. Most economy level digital scopes do not have real fine adjust. Just only "pixel zoom" what do not give anything more than magnifier class over TFT. Example measurement accuracy/resolution do not grow. If voltage fine adjust is done in analog front before ADC, it may give real advantage.

And nearly same for horizontal fine adjust. What advantage it give as long as it is just only same sample points positioned differently on the TFT. Just nothing more. In analog oscilloscopes these fine adjustmets are really different. In most digital scopes they are just cosmetics.
Why there is nearly total lack of any real lab tests
(and not only in this video, in nearly all videos, nearly total lack of any real lab tests with real data. I can see reason is not lack of good test equipments. So, why do not use these?)

[mikeselectricstuff]

And nearly same for horizontal fine adjust. What advantage it give as long as it is just only same sample points positioned differently on the TFT. Just nothing more. In analog oscilloscopes these fine adjustmets are really different. In most digital scopes they are just cosmetics.

Horizontal fine adjust can be very useful, e.g. to align the graticule marks with bit-cell positions when looking at async serial data.
I could probably live without vertical fine adjust, but lack of horizontal fine adjust on a scope in this price range is unacceptable.
 

[Christe4nM]

Why there is nearly total lack of any real lab tests

A real comparative lab test would take a lot of planning beforehand, and an enormous amount of time altogether. Although from a technical point of view a comparative test might be more fair, the technical aspects of equipment are only one aspect of what makes a good instrument. Look and feel is one for example, and ease of use is another. Not all aspects have equal weight in what makes an instrument good, but they all contribute in one way or another.

In this case it's very easy to get a negative feeling towards this scope, if only from ease of use or lacking thereof. The glitches and other stuff Dave mention contribute to a feeling of just bad design. If this first impression is already this bad, why would anyone spend time delving into the specific technical aspects. It just doesn't give an impression that it could be any good, even if it were.

[marmad]

In test with glitches. It looks like (but too small amount of time in viedo so it is now only "feel") it sometimes trig to glitch more fast than if think update rate is always around 35 or less per second with long real visible blind time. (also this is proofed in tests using "marmad's double pulse method")

Just a couple of thoughts on the waveform update rate stuff:

@Rf-loop - I think we have to assume, unless definitive proof turns up otherwise, that the Trigger Out of the Owon is the real continuous waveform update speed of the DSO.

OTOH, an observational glitch test is a very simple method to get a very broad idea about waveform update rates. As mentioned in other threads, it can only tell you how slow a DSO isn't - not how fast it truly is (because of probabilities). Since waveform update rates are not precise frequencies (e.g. it may be 53.355 wfrm/s), it can happen coincidentally that the update rate lands sporadically on a multiple of the glitch frequency (if it's repetitive).

The update rates of the Rigol DS1052E have recently been measured and posted here and here by EEVBlog member Harvs.

So, the waveform update rate of the DS1052E @ 10ns/div (which Dave was using) is apparently 60.7 wfrm/s - and the update rate of the DS2202 (AUTO/10ns) was 9,400 wfrm/s.

[xrunner]

Just watched this review.

Ick.

That thing is just ... awful. I just don't understand why a manufacturer wouldn't bring in a consultant, let's say it was Dave for example, that would put it through it's paces before releasing it. Why get such a bloody nose for no reason? 

[Hydrawerk]

About vertical fine adjust. Most economy level digital scopes do not have real fine adjust. Just only "pixel zoom" what do not give anything more than magnifier class over TFT. Example measurement accuracy/resolution do not grow. If voltage fine adjust is done in analog front before ADC, it may give real advantage.

And nearly same for horizontal fine adjust. What advantage it give as long as it is just only same sample points positioned differently on the TFT. Just nothing more.

Is this the case of my DSOX2002A?? I really don't know, it's hard to find any info on that.
Well, probably not. See the parts of my user manual.

[Hydrawerk]

I see one odd thing.
This individual scope was made one year ago.
(how did you find this old unit? I understand if you have made this video nearly one year ago but... )

No problem. My DSOX2002A was manufactured in 6/2012 and i bought it as new scope from Farnell in 4/2013. 
Anyway, here are photos of Dave's Owon. The serial number is SDS93021229286.

[casper.bang]

Nice review, definitely enlightening... will look elsewhere for my first scope. How about a review of a same category Siglent? They have to be better than the Owon and Rigol entries?

[Rasz]

The first thing it says to me is "if they don't even give a shit about getting one of the most visible things right, why would I think they'll spend any time on the important stuff inside that I'll never see?".

At least we know they are honest because ... they DIDNT 'spend any time on the important stuff inside'

Why there is nearly total lack of any real lab tests

Because there is no point wasting time testing a piece of shit like that scope. Fricking UI LAGS more than 500 milliseconds and eats into update rate. It reminded me of GEOS on C-64.