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

[marmad]

Agilent DSOX3000 is sampling with much LARGER memeory depth, don't compare like that, when you look DSOX2000
you will see for example 4500 wfrm/s (20us/div) where DSOX3000 "only" 2000 wfrm/s  - but DSOX2000 is sampling
into 5k memory and DSOX3000 into 1M i think.

According to the Agilent literature you posted, they did this for measuring the waveform update rates:

"Only one channel of the scope was turned on for these tests.  Memory depth was optimized at each timebase range by selecting the minimum amount of acquisition memory that would also provide the maximum available sample rate."

Also, again, according to your figures, the Hantek is faster than ALL of the scopes in those tests at those sec/div ranges - mostly 10x faster.

[rf-loop]

Not directed to any separate person here, but who ever is interest.

I will highly recommend to read  this R&S aplication note. www2.rohde-schwarz.com/file/1ER02_1e.pdf

 It give really good knowledge about "how stuff work" and what is wfms/s and blind time to read this linked R&S aplication note. Starting from firts row, ending to last. Agen and agen. There is lot of "good to know" things and specially "goot to understand" think. Even if it is handling mainly high-end scopes but there are lot of basic information. Also it helps understand better how digital scopes work yesterday and today.

[marmad]

As Agilent said "Memory depth was optimized at each timebase range by selecting the minimum amount of acquisition memory that would also provide the maximum available sample rate" - but this also means the lowest possible memory depth!!!, so you can't compare 50k depth to 4k depth - just calculate back depth * wfrm/s and divide by depth * wfrm/s from competitor and you will see how to read such tables.

Sorry, my friend - I think you are missing my point - look at the figures for EVERY scope in EVERY chart in that literature at, for example, 20ms:

Timebase - Agilent2000 X-Series - Tek DPO2000 Series - Tek TDS2000 Series - LeCroy WaveJet
20 ms/div                               ~5                               ~4                              ~4                        ~4
                 Agilent 3000 X-Series - Tek DPO3000 Series - Tek TDS3000 Series - LeCroy WaveSurfer
20 ms/div                               ~5                               ~4                              ~5                        ~3
                     Agilent 7000 Series - Tek DPO4000 Series - LeCroy WaveRunner
20 ms/div                               ~5                               ~3                              ~3

The Hantek, according to you, at 40k sample depth and 20ms/div, does 20 wfrms/s.  That is a factor of 4 to 7x faster than ANY of those scopes - and I'm sorry, but that just doesn't make any sense.  The chart is showing what some of the most expensive scopes can do,  so I would have to assume that your figures are wrong.

Anyway, I wasn't planning to create a disturbance - I know you are a big supporter of the Hantek.  I think the Hantek is a great scope for the money (even if those rates are /10) - better hardware and firmware than the Rigol - and I would own one now if I had found a good price before I heard about the new Owon.

But if we want to continue this discussion, it's perhaps better suited to the Hantek thread.

[rf-loop]

if direct true capuring memory is 1k and sampling rate (as long as all samples are pushed to capture memory)  is 1Msa/s can not go over 1000wfms/s and also it is impossible becouse blind time can not me zero. And if it is 1M and 1Msa/s maximum with zero blind time is 1wfms/s (even what ever is display t/div) This is for 100dollar scope and it is for 100k dollar scope.

This is good discussing also with Owon.

marmad, I'm sure we are technical peoples here. Not selling oriented peoles. Technical facts and tests are most important. All scopes have goods and bads. My interesting is truth about every pros and cons. Some people want mango, some people need banana, some people need mandarin. Also these are somehow related to price.

I hope soon know Owon tested values and also Hantek tested values. Sometimes help also some my friends etc find right scope for him. Some may need hantek, some may need Owon. King is customer or this friend and his need. I do not want never push some people buy test quipment what do not meet his needs. First need also priorize needs. (including price/budget) I want always people get what he need. Not what I like sell. If after this study it may be that people need soething what I have not. I tell it to him. I can teach him what he need, what I recommend.
One key for this is as much knowledge about machines details and characters. Many times anufacturers datasheets are not enough - need know machine itself. Car datasheet is dimensions, power, speed etc. Who buy car only knowing datasheet?

[marmad]

if direct true capuring memory is 1k and sampling rate (as long as all samples are pushed to capture memory)  is 1Msa/s can not go over 1000wfms/s and also it is impossible becouse blind time can not me zero. And if it is 1M and 1Msa/s maximum with zero blind time is 1wfms/s (even what ever is display t/div)

Yes, this is what I was trying to get at before when I mentioned the 'math' in my post above.  Without special dedicated hardware (like Agilent's new ASIC), there is a theoretical limit to wfrms/s - and it's not very high.

[marmad]

I re-measured trig out frequency using 100k pull-up resistor and frequency counter.  The figures were more or less the same as I posted back at the top of page 7 of this thread.  The frequency seems more closely related to refresh rates then to wfrms/s, but in any case, the frequency never climbs higher than about 25Hz.  The duty cycle time MIGHT be an indication of 23% dead time.

[rf-loop]

I re-measured trig out frequency using 100k pull-up resistor and frequency counter.  The figures were more or less the same as I posted back at the top of page 7 of this thread.  The frequency seems more closely related to refresh rates then to wfrms/s, but in any case, the frequency never climbs higher than about 25Hz.  The duty cycle time MIGHT be an indication of 23% dead time.

Memory depth 1k
example 1us/div
1MHz signal to CH1 input. CH2 off.
all extra functions off, like measurements, cursosr, dots connection etc. (I mean minimize all processing what scope do blind time)
Trigger edge.
What is trig out freq?
I hope it is not so that capturing is  related to display refres time. (just as -80's digitals... )  It can not be...
But it can be that trig out is not real trig out... becouse there is not clear explanation what is this trig out...

[marmad]

ok let's calculate for DSOX3000, 50000 * 5 = 250.000, now Hantek 4000 * 20 = 80.000, so which one is higher?
This means DSOX3000 would do 62.5 wfrm/s when there would be an option to enable 4k, or vice versa
Hantek would do 1.6 wfrm/s when it would be able to capture 50000point at given sample rate.

I posted a response to this over in the  Hantek - Tekway - DSO hack thread - where it makes more sense.

[House91320]

This scope is probably the one I'm going to get for college mostly because its portable its cheap and the screen is nice. My question is is anyone aware if owon will updated the software because the current software in his reviews are a possible deal breaker?

[bertchai]

Just received my SDS8102. There will be a lot of things to check.....

[rf-loop]

This scope is probably the one I'm going to get for college mostly because its portable its cheap and the screen is nice. My question is is anyone aware if owon will updated the software because the current software in his reviews are a possible deal breaker?

Xiamen Lilliput overseas sales department have promise (to me after I ask) that this FW is not "final" version. They know it need some development and bugfixes.

They  "promise" that new FW come later. FW do NOT come (afaik) as public downloadable.  Main stream looks that they give it directly to distributors and customers.  Now there is not ready FW update. They can not answer exactly timetable. But I somehow understand that "soon". (what is soon? 3 weeks? 3 months? I do not know - yet)
------

I will make some tests soon with SDS7102 w option VGA

[House91320]

Cool as long as there's updates for the scope coming I feel comfortable spending my money.

ps excited to see the teardown

[marmad]

Still tabulating some tests and trying to work out the results... I'll post the third (and final) installment of the SDS7102 review sometime in the coming week.  Bear with me 

[House91320]

Is there going to be a treadown because I,m interested in the quality of the construction?

[FreeThinker]

Is there going to be a treadown because I,m interested in the quality of the construction?

I Don't think so . I asked the same question earlier in the thread a the reply was he is still in the 'Golden warranty period' ie he's not going to ( publicly ) void his warranty. (Bet he has a sneaky peek though )

[rf-loop]

maybe soon (inside two weeks) I have something to tell. There is coming some lot of these and I will take one for my deeper tests. This I can not promise what results and examinations are public and what not. But some "opinions" I can promise later.

[marmad]

Is there going to be a treadown because I,m interested in the quality of the construction?

I can't exactly promise a teardown at this point, but I'll have a lot of extra info - some of it pertaining to what you're interested in - in a couple of days.

[rf-loop]

Only overall look after arrive.
DSO7102  (Vga)
V 2.1.1

First feel is: nice, not feel like cheap plastic, finishing good. (if I give points about look and touch feel: Hantek  get 6, Rigol 7, and this . In hand it feels more expensive than it is. And this display...

Good TFT. Also if dim baclight. Owerall feels well made.
separate on/off's for mains and scope on/off (becouse battery option)
Fan is - just extremely silent.

This machine looks and feels well made (outside).

Now I only have time to fast look freq response.  This I do later better. Now need study littlebit menu and features.
I have seen many kinds of scopes, old professionals and high-ends and some these entry level scopes.
This is better than I expect. (this is only first feel)

Freq response - It is good! Really good. Nice gaussian reasponse without highs and lows. Well made.
Noise looks reasonable and  ok in this class. I have not tested more deep but fast mutual feel was.. this is of course better than Rigol ( and also maybe better than Hantek. maybe specially some class of noise  is lower)

Compared to 500kHz pure sinewave signal as reference. (300mV p-p)
Scope 50mV/div, 2ns/div

-3dB frequency 135MHz, -6dB  230MHz    also I only fast turn  Signal gen f up to look simply where it loose trig... after 380MHz without any trying to find trig by any settings. Signal level same, only scope changed to 20mV/div))

Signal termination in scope end with Tektronix 50ohm  terminator. Generator well leveled HP8657B.

[Mechatrommer]

Freq response - It is good!
-3dB frequency 135MHz, -6dB  230MHz 

and how it is compared to rigol ds1052e hacked to 1102e?

[marmad]

This machine looks and feels well made (outside).
I have seen many kinds of scopes, old professionals and high-ends and some these entry level scopes.
This is better than I expect. (this is only first feel)
Freq response - It is good!

Yes, it's true.... and all of this I've mentioned in the videos and the pages of this thread.  Everything that Owon claims in their specs (except, of course, for optically-isolated synchronous Trigger out) appears to be correct.  But the problems lay elsewhere - in the unreported stats.

But now you can compare the Hantek with the Owon in this regard - quite simple to test them against each other, as Dave did with the Agilent and Rigol.

[rf-loop]

Only shortly and not exactly compared. So this is mutual but I know something and it as "wise quess".

Hantek have goods and bads also. But Hantek is littlebit like DPO. It can capture more wfms/s to display.
Owon is bsic sequantially capture-display-capture-displ. conventional DSO. But in this it is very good in this price class imho.

Later more, some days now extremely bysy.

[bertchai]

hmm....  my SDS8102 arrived in a version  : V1.0.1.  May be i need to check for updates from Owon.  See if there is any update.

[tinhead]

DSO7102  (Vga)
V 2.1.1

good point, marmad what's is on your Owon ?

[marmad]

V 2.1.1

Same.  I did hear on samenkopen.nl about a bug with Chinese characters appearing when you use FFT - that was on units shipped in late June (early July).  But 2.1.1. doesn't do that - so I'm guessing that might be the 0.0.1 in 2.1.1 

[rf-loop]

Shortly about FFT:
First opinion without enough testing, only just as "hello" test: normal FFT what I have seen this price class DSO.
Can not window with waveform. Seems like they have put more development for dB scale than Vrms scale.