OWON Oscilloscopes

Owon Electronics and test equipments. Discussion area is now open!
It is currently Sat Jun 24, 2017 11:47 am

All times are UTC + 2 hours [ DST ]



Welcome
Welcome to owon

You are currently viewing our boards as a guest, which gives you limited access to view most discussions and access our other features. By joining our free community, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content, and access many other special features. In addition, registered members also see less advertisements. Registration is fast, simple, and absolutely free, so please, join our community today!


Post new topic Reply to topic  [ 13 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Some tests with new SDS7102V Version 2.6 and 2.8.2
PostPosted: Sun Jun 03, 2012 3:27 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Owon have done some improvements in hardware.

I will do later some tests.

What I know now is that:

- frequency BW is littlebit reduced. (this is, in most cases, good change!)
- 100MHz is still around -1dB as before
- noise level is reduced (also before it have been very good in this class)
- thermal stress in front end is reduced due to changes in circuits (front end run markable lower temperature)
- cosntruction is also now better for service
- power supply is new version (some advantages in battery use)
and thermal design is better. (but also before I have not meet any thermal problems but maybe this give better longevity (same as front end new design)

After I have made tests and repeated these enough times with different test settings I will put here some results.



Before this, it is good to read these for better understanding about BW, samplerate and noise things.
These all are from Hewlett-Packard/Agilent (just only becouse my nickname is aghp):

Evaluating Oscilloscope Bandwidths for Your Application

Evaluating Oscilloscope Sample Rates vs. Sampling Fidelity

Understanding Oscilloscope Frequency Response and Its Effect on Rise-Time Accuracy

Evaluating Oscilloscope Vertical Noise Characteristics

Sin(x)/x Interpolation: An Important Aspect of Proper Oscilloscope Measurements
By Chris Rehorn, Agilent Technologies



As far as I know 2.6 UI / FW features and functions are same as 2.5.1. (or 2.5.1.6)
(if there is some differencies in FW they are not visible to user.
I have not seen/find any changes) Inside system there may be differencies what are realted to HW (HW/FW surface may be different as some circuits have changed and driving these mey need some differencies)

Also Owon answer for UI/FW question was that number do not mean any chage to user level and 2.5.1 is same (no update need this time).

Test results coming soon.....

-aghp


Top
 Profile  
 
 Post subject: Re: (Coming soon) Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Sun Jun 03, 2012 11:40 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
First some tests about frequency response and risetime.

Image

Here is linear frequency sweep from 1 to 250MHz. Sweep time is 100s.
Signal source is HP8544B and its level is well enough accurate for this test.
Signal come with around 1m 50ohm coaxial and scope input have Tektronix 50 ohm feed thru terminator.
(this is not perfect match due to oscilloscope input capacitive reactance)
Oscilloscope settings: Peak mode. (displayed sampling speed is not important)
Vertical 50mV/div. Signal adjusted for around 8diVpp @ 1MHz

Nice shape! No big up-down throws that could have come from a poor front-end.
Very good result in this price class!
(Note: test signal impedance mismatch. Light wavy structure in shape perhaps just from this mismatch.)

Oscilloscope front end (analog) circuit is divided to three voltage "bands"
||1. 2-100mV/div ||2. 200mV-1V/div ||3. 2V - 10V/div ||
Bands 2 and 3 have around same BW. Littlebit higher freq BW (-3dB point) if compare to band 1.

Band 1 have this measured BW. Note: 2mV/div have fixed 20MHz BW limit.




Image

Test signal risetime is around 1.3ns. It is "Edge" from Tektronix CG5011 (small modification).
1m 50ohm coaxial and Tektronix 50 ohm feed thru terminator.
V/div 200mV
Test signal maximum edge level is this so it can not give 8 div display with 200mV/div

Oscilloscope front end (analog) circuit is divided to three voltage "bands"
||1. 2-100mV/div ||2. 200mV-1V/div ||3. 2V - 10V/div ||
Bands 2 and 3 have around same BW. Littlebit higher freq BW (-3dB point) if compare to band 1.

This can see if look this risetime (~1.8ns) using band 2.





Image

Test signal risetime is around 1.3ns. It is "Edge" from Tektronix CG5011 (small modification).
1m 50ohm coaxial and Tektronix 50 ohm feed thru terminator.
V/div 100mV

Oscilloscope front end (analog) circuit is divided to three voltage "bands"
||1. 2-100mV/div ||2. 200mV-1V/div ||3. 2V - 10V/div ||
Band 1. have littlebit reduced BW (-3dB point) if compare to bands 2. and 3.

This can see if look this risetime (~2.2ns) using band 1.






Image

Test signal risetime is around 1.3ns. It is "Edge" from Tektronix CG5011 (small modification).
1m 50ohm coaxial and Tektronix 50 ohm feed thru terminator.
Scope set for infinite persistence for show all variations on the display (it is nice feature for collect changes)
V/div 100mV ((2)-100mV/div is lowest voltage "band" in front end circuit and it have littlebit reduced freq bandwidth if compare to 200mV - 1V band and 2V - 10V band.

Here can not see markable aliasing.
This is natural becouse samplerate is 1GSa/s and signal itself do not have so much frequency components over Nyquist frequency (500MHz) and also at Nyquist freq , front end attenuation is something like over 20dB.




Image

Here can see one form of alising. "Corners wobbling".

Nyquist freq is now 250MHz (samplerate 500MSa becouse two channels in use)

Test signal risetime is around 1.3ns (just same as before) so it have lot of freq components over Nyquist freq. Front end attenuation is roughly 10dB @ 250MHz.

There is used infinite persistence for collect and show this aliasing. This aliasing is natural becouse test signal risetime, samplerate and analog front end bandwidth shape.

It is now good to remember here test "100MHz" oscilloscope with 1.3ns risetime edge and realtime capturing.
If signal is repetitive here can use averaging or just this persistence (not only infinite, also 1 - 5s is useful)
But how if we capture just single shot? Look picture, single shot result may show what ever risetime you can imagine in picture collected with persistence.

This aliasing is not becouse oscilloscope is poor. It is just as waveform and sampling mathematic fundamental. Digital oscilloscope user need know and understand some amount of these basics.

Of course it is nice IF oscilloscope have expensive and complex analog front end where is analog adjustable low pass fileter with nearly perfect "brick wall" shape. If its corner frequency is adjusted together with current samplerate so that always it have tight low pass from zero to Nyquist limit we have very nice and expensive digital oscilloscope what do not have aliasing problem at all.
Owon solution with front end BW and samplerate etc is some kind of "good" compromise now.
before this 2.6 version there was really wide open front end.
In some cases it it is nice.. but but...
On the other hand, it can easily lead to incorrect measurements especially if you have little experience or knowledge. Also, for the experienced user, it is sometimes a tricky thing when you consider the signal is completely unknown.

My opinion is that this BW is now (littlebit) better than before. Someone may tell that ~250MHz -3db is better than 160 - 200MHz -3dB.

Yes it "looks like" better IF we test oscilloscope with well known signal! But do you need oscilloscope for exatly known signal - no you do not need oscilloscope at all if you really know signal. (exept if you like play and use oscilloscope just as toy). Oscilloscope need for look more or less unknown signal. And with unknown signal you need always know something about what may lead wrong results. Aliasing is one.
(there is also other type of aliasing... more common in peoples mind maybe. If Nyquist is example 100MHz. Then I connect 110MHz sinewave to oscilloscope. Oh...what is this 90MHz? But what if I connect 201MHz to scope input and agen it trawel to ADC.. through front end circuits. Oh.. what is this 1MHz.
Specially this need know if you go down to slower horizontal speed so that samplerate drops.


Here Owon is very strong! Becouse really big sampling buffer memory.. up to 10Msample with full speed in SDS7102. This result you can find in samplerate table. Look then competitors. (example Hantek and specially Rigol and many Rigol copycats)


Top
 Profile  
 
 Post subject: Re: (Coming soon) Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Mon Jun 04, 2012 11:39 am 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Next some words about noise. (not yet ready)

This Tektronix tiny answer is just fun but least partially true.

On the other hand wondering why Tektronix does not give even a little better answer.

(yes if they are honest they take pictures from Tek low cost models and... well after then they do not sell any. -- Exept to peoples who buy even ******** with high money if there read Tektronix (This adhesive front panel label can buy with one cent).

This is maybe becouse before years (in history) Tektronix have been leader in oscilloscopes, analog oscilloscopes. )


Continued ...


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Mon Jun 04, 2012 1:03 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Noise is now reduced some amount due to new design in front-end (V2.6)

But first this old.
(later, soon, I will try same tests with new version)

Here is this OLD original "rf-loop" test with earlier oscilloscope HW version (example same front end as V2.5).

rf-loop in eevblog forum wrote:
Not very bad.
Measured with peak-peak using 1M memory and 1GSa samplerate with 50us/div.
Persistence infinite. Then stop. Zoom out so whole capture is on the screen. Then with cursors.
(with this setting Owon capture is 20div lenght so whole capturing is easy to look after one step zoom out and so that it is littlebit comparable to Agilent paper. (not exactly but no big difference in this noise meaning. Of course other kind of setting give lot of better result but this is not AD, so...nearly worst case realism. But this I do not know how much Agilent paper have bias... )


Image
Table made by: "rf-loop"[/quote]


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Mon Jun 04, 2012 4:58 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Here done as rf-loop described and added to table.
(it need some explanation due to measurement resolution with cursors)
(cursors value is in () ) first value is oscilloscope automatic measured p-p value.

Image

Over 100mV/div measurement with cursor is not ok for this low values due to cursor resolution.

Clear improvement! Look these under 50mV/div values.

Here is good to note that example 1V/div there is 10div vertically and it is around full scale of ADC.
I'm not exatly sure but it looks like full ADC scale is in this case 10.24V (10div on the display visible)
10.24V/256 = 40mV resolution)
5-20mV/div tell more about analog front end noise (becouse cursors and measurements resolution)
and THIS noise is really super low in this class of oscilloscopes!

Low noise is meaningful also for example trigger stability.


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Thu Jun 07, 2012 10:34 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Correction: Previously reported test signal rise time, 1.3ns was wrong. (it was from Tektronix datasheet where it is: <1.3ns not 1.3ns

As I began to suspect that it is more fast, I did research. Exactly the same cable and terminator connected to the HP oscilloscope. The result: about 850ps or less.

This influence will be quite small in this case but it give littelebit more alias than expected. If the signal rise time is greater it would reduce the alias effect. (in this case, if the lower corner wobling should be longer than the rise time)


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V (2.6 to 2.7)
PostPosted: Sun Jun 10, 2012 9:38 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Owon SDS7102 - Tektronix TDS520A
Noise

Setup is made so that scopes are as close comparable as possible.

Note: Tektronix 100MHz BW limit is more low than Owon full BW (100MHz)
Signal from HP function genrator using 50ohm feed thru terminator. (Tek have built in 50ohm but here used same as with Owon)


Owon set here 10k memory and Tek 5k memory and "fit to window"
With this they are more comparable becouse then Owon have 10k 20div Tektronix 5k 10div.

Image
Owon 100MHz full BW just as it is (-3dB far over 150MHz)


Image
Tektronix with 100MHz rejected BW


Image
Tektronix with full BW







Noise without signal.
Oscilloscope 1M input with 50ohm terminator.

Image

(cursors show Owon OLD version noise peak level.)



NOTE: Owon have used 10M memory. Tektronix have only tiny sampling memory so also trace looks different.
(if drop Owon memory it looks same but, these tests was done different time and I did not use Owon small memory becouse with it, noise looks "too good" visually in display (of course it do not reduce real noise p-p).

Image


Image


Need realize that Tektronix is old but totally different price class oscilloscope.

Owon result is very good.

In this price class I do not know any other what is as good as owon if look noise.
Still Owon have noise, of course, and it also affect to triggering stability. But also so do this Tektronix.

Due to lack of time this was only small test.
(specially becouse need add floppy disk driver to this old Tek for images and it need some improvisation becouse not original parts.)


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V (2.6)
PostPosted: Wed Jun 13, 2012 10:29 am 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Littlebit more with face to face with Tektronix 520A.
Scopes get exactly same signal (parallel).
Settings so that highly comparable.
Tektronix 100MHz BW becouse Owon is 100MHz oscilloscope.


Image
Image


Image
Image


Image
Image


Image
Image


Looked more deep this triangle corner. Just after voltage range change over 100mV p-p level there appear this ringing. (proportionally most high ringing with this level).

It is problem in this function generator - need service adjust.


Image
Image


Image
Image



Image
Image


Image
Image


Owon front end signal quality is amazing good as we look 0.5kEur price class. Maybe becouse they have designed it themselves, not try make cheap copy from some other.
If do copy it also may mean that copycat do not understand fully this circuit what he is copying. Then they may do some changes for more cheap without understanding how it affect and then we see "Siglent" type result.

If design and test in own lab, it need understand something and result may be this what Owon show here.



Why this noise is important. Of course it mean what resolution you can look your signal. (if you take example single shot, how you know what is oscilloscope made and what is really in signal under test - I think it is good to ask)
Then, noise affect also to Trigger and this is very important. Part of trigger instability come just from noise - naturally.



It is also good to imagine that these pictures are still pictures.
In practice with live screen human eye-brain system make some effects and it may mostly look better. This effect is extremely clear if you run video with fast movements and then you stop it. Did you really know from how bad image your eye-brain complex produce image what you "see". We have high amount of image processing. This help also with oscilloscope in many cases.
If you have long experience with analog oscilloscope it may need some time before you learn how to look digital oscilloscope screen and what you look and what you "filter" away (you see but with intelligence you do not care)



Add:
Here one image what may tell something how BW affect.
It is just with same signal as before but now with full 500MHz front end BW. This is why I use 100MHz BW in Tek (becouse also Owon is 100MHz. But Owon BW is higher than Tek 100MHz filter. (Owon BW is around over 50% higher)

Image



Next some other things tests... soon.


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V HW version 2.6
PostPosted: Tue Jul 31, 2012 11:31 am 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
Today oscilloscope version is 2.7.x
FW is functionally exactly same as 2.5.1
There have been only some changes in HW.

These tests are also valid for oscilloscope version 2.7 and 2.7.1 of course including random variations between single individual units.


Top
 Profile  
 
 Post subject: Re: Some tests with new SDS7102V HW version 2.6
PostPosted: Thu Dec 27, 2012 12:50 pm 
Offline
Site Admin

Joined: Wed May 16, 2012 7:54 am
Posts: 187
December 2012 newest factory version is 2.8.2
(I can not find FW differencies to 2.8, to 2.8.1 so mayb ethere is what ever change in HW and this time I do not know what it is.)

Roughly it seems about same if look BW etc.

I heve made tests but I have not yet double checked data for public share.
Soon I will show some new test data and pictures.

But it seems new version is as good as before or better.

Exactly BW vary littlebit between individual units due to components tolerances and also how it have final adjusted in factory.

One unit I test -3dB was around or over 200MHz using V/div middle band. (whole range from minimum to maximum is divided to 3 voltage bands (selected with inside relays)
It looks like lowest V band have littlebit lower -3dB limit (as usual and lowest 2mV/div of course becouse it have fixed 20MHz low pass)

It seems also that noise is also least as good as before.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 13 posts ]  Go to page 1, 2  Next

All times are UTC + 2 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
 cron
Donate Now
Donate Now



Hosted by © 2017 FreeForums.org | Create a free forum | Powered by phpBB
About FreeForums | Legal | Advertise Here | Investors | Contact FreeForums.org
Report Violation

subSilver+ theme by Canver Software, sponsor Sanal Modifiye
suspicion-preferred