!!

Welcome to ITTSB Forum, extension and content sharing platform of ITTSB.EU Blog

Registration this is Free, to support this Blog please use our donation links available in our Blog page and forum.

Copyright Notice:  All content of this web site this is copyright protected.

Author Topic: Announcement: GW Instek GDS-2102A oscilloscope purchase, March 14/2013 ITTSB.EU  (Read 36398 times)

0 Members and 1 Guest are viewing this topic.

marmad

  • Guest
We have  2048 segments, 1K point size per segment at 8 ns resolution, you did mentioned 20 ns resolution for your Rigol (rearm time ).

Hi Kiriakos,

I think you misunderstood my email. All scopes that use segmented memory use triggering to capture segments - but the fastest they can do that is limited by the wfrm/s (or perhaps a bit faster by using lowest possible sample size). So, for example, the GDS-2000A at 500ns/div could be calculated as approximately (1/80000) + (500ns * 10) = ~17.5us minimum segment time.

Cheers, Mark
« Last Edit: April 05, 2013, 01:33:33 PM by marmad »

marmad

  • Guest
Hi Kiriakos,

Curious about your numbers with the higher frequency, I ran a quick test on the Rigol today at the 2ns/div setting - sending it an input signal at the following frequencies: 100kHz, 1MHz, 10MHz, 20MHz, 50MHz, and 100MHz.
The Rigol's rate was always 25,400 waveforms per second, plus or minus 200Hz, at each frequency. So it's curious that the GW-Instek's rate varies with the frequency.

Best,
Mark

Offline Kiriakos GR

  • Administrator
  • Hero Member
  • *****
  • Posts: 2163
  • Country: gr
    • ittsb.eu
  • job title: Industrial Maintenance Electrician
Hi Mark, 

I did read all your questions very carefully and I have collect data which they will
help you and also help all my readers, so to further understand what it happens on my workbench.
The Why it happens? is another topic all together, and my guessing will be that the GDS-2000A is using
some " Engine Parts "  from the GDS-3000 Series which is an advanced and mature product line.
I will post in few hours those new information at the ITTSB page.


Mark regarding the segmented recording, this "dead time detection" under waveforms capturing,
it gets confusing when you describe it as Triggering.
By my opinion it is an simple recording with some resemblances to what my HD Video camera does at a fixed rate of 25 frames per second.
At the scopes even the maximum number of frame per second it totally relevant with the Time base + those 8 ns separator between frames. 

GW Instek very nicely described the capturing process as film movie,
in which every new frame it is separated by the previously one  by an 8 ns in duration barrier.
I will agree about the fact that the time length of its one frame is 100% relevant to the time-base,
lets not forget that its time-base increment serves best an specific range of wavelengths.

The duration which every wavelength needs so to be recorded as frame it varies.
What it does not varies is the file size, the scope is actually waiting for the waveform to arrive so to be measured, and then it takes a picture of it.

My guessing about the 1Kpoint per segment, that seems as 1000 points with 8-bit information,
is that it works very similar to:  The 8-bit color palette (256 Color )
That is 256 free parking seats (storage for points) so to serve its one Input Chanel of the scope,
for one four channels scope the sum of 4X256 = 1024 points, that is the true size of this 1Kpoint.
And in conclusion this 1Kpoint serves all channels but always divided by four.

I would expect from your Rigol to have a larger segment size due it 10 or 12-bit resolution that is dependent to specific frequencies range.
At 12-bit Color Codes you have 4096 colors
In conclusion your current scope it using 1024 points (per channel) x4 channels = 4096 total of points as one complete segment.


My logic tells me that your scope needs this larger memory because the sample chunks are larger,
and not as benefit for super long recordings. 

Here is where my IT training pays out too.   ;)
WWW.ITTSB.EU   Industrial Test Tools Scoreboard  (Product Reviews Blog) / Editor in Chief.

marmad

  • Guest
Hi Kiriakos,

I think you are getting confused about the way that segmented memory works because of the badly-written description of the process in the GW-Instek manual.

From the GW-Instek GDS-2000A User Manual (page 111):

"The advanced segmented memory utility allows the scope memory to be divided into different segments. Each time the scope is triggered, it only acquires data for one segment of memory at a time. This allows you to optimize the scope memory to only perform signal acquisition during important signal events."

This description is correct - and the way that all DSOs with segmented memory work. They record a segment when a trigger happens - and not like an HD video camera, recording frames based solely on time.

But later in the GW-Instek manual, when they try to describe the process in detail, they make it sound almost as if it's unconnected to triggering, although that wouldn't make any sense for practical use. But you only have to test it yourself to see that segment recording is related to triggered events (although you can also, at least on the Rigol, set a time delay before the next triggered segment).

Also, the GW-Instek's "time-tag resolution of 8ns" does not refer to the time between segments. Instead, it refers to how much resolution (i.e. number of digits) the display of the time has. For example the Rigol DS2000 seems to have a time-tag resolution of 10ns (or 7 digits - whichever is greater). According to published Agilent information:

Time-tag resolution:
3000 and 4000 X-Series: 10 ps or 6 digits (whichever is greater)
2000 X-Series:                 10 ps or 6 digits (whichever is greater)

Re-arm time (minimum time between segments):
3000 and 4000 X-Series:  1 μs  (1/1000000 wfrm/s)
2000 X-Series:                20 μs  (1/50000 wfrm/s)

So you see, the time-tag resolution and the minimum time between segments is a very different thing.

As far as the Rigol goes, it is much different than the GW-Instek in selecting segments. On the Rigol, I choose the segment size and then the DSO reports how many segments I can use. For example, if I choose a segment size of 1.4MPts, I have 63 segments. If I choose 14kPts, I have 8128 segments. If I choose 1.4kPts, I have 65000 segments.

marmad

  • Guest
P.S. And yes, you can do super-long recordings on the Rigol  :)  When using the maximum 65000 segments - and the maximum re-arm (delay) interval of 10 seconds - you can record a time-lapse picture of a triggered waveform over more than a week: at least 7.5 days of time in length (with a trigger of at least once every 10 seconds). With 700ns being recorded at each segment, that would be ~45 milliseconds total of recorded time - spread evenly over 7.5 days. You can then run it through the analysis function - and get a histogram of deviation of the waveform over that entire length of time.

Offline Kiriakos GR

  • Administrator
  • Hero Member
  • *****
  • Posts: 2163
  • Country: gr
    • ittsb.eu
  • job title: Industrial Maintenance Electrician
Mark you are correct that for few things I am confused,
this is why I pushed the damn HELP button on that scope. ;D
From the Help menu ....

About the record length of 1000 points ,
in Short Mode  the length is limited to 500 Points.
in Auto Mode the length could be 1000 Points for a single Chanel,
or half with both channels active (2 Channels).
I will just make here the assumption that if this GDS2104 comes also with a 1K record length,
with four channels active it would have 250 points per channel.
   
In the next hours I will upload some charts and pictures regarding the Wfms/s story,
I did confirmed again that the scope behaves that way at a higher frequencies range,
which is not commonly used,
By my humble opinion you may see some benefits,
if you are designing low-pass filters which works at 10MHz to 100MHz range.  ;D

Lets set aside for now the segments recording,
which is a topic that I need to educate my self further.  :)

WWW.ITTSB.EU   Industrial Test Tools Scoreboard  (Product Reviews Blog) / Editor in Chief.

marmad

  • Guest
Hi Kiriakos,

In at least one way, the DS2000 and GDS-2000A are quite similar: their manuals, translated from Chinese to English (or other Western language), have some funny mistakes and strange language use  ;D

I measured the Trigger Out on the Rigol and it is 3.5V into 1M Ohm (1.5V into 50 Ohm); so it seems the low voltage on the Instek makes it harder for you to read frequencies.

Although I haven't used the GDS-2000A, from what I can tell from the information available, it clearly beats the DS2000 in terms of waveform update rate, expandability, and some features which are missing from the Rigol, and the DS2000 beats the GDS-2000A in terms of memory. Which scope is better for a potential buyer, of course, will always depend on their individual needs.

Best,
Mark

Offline Kiriakos GR

  • Administrator
  • Hero Member
  • *****
  • Posts: 2163
  • Country: gr
    • ittsb.eu
  • job title: Industrial Maintenance Electrician
Mate I did not had the time yet to read the user manual, but I am going to eat this book page by page.  :)
The GW Instek User-Manual regarding writing quality, it is very good but those people they assume that you have some previews education regarding scopes functionality.

Thanks God that is not a F16 with a 10,000 pages flying manual.  8)
Those Agilent webinars are great help for basic understanding of what a scope is all about,
and also of which specifications are important.

And I will publicly admit that what I liked up to date about Agilent and GW Instek is the fact that those brands they have very common technological goals (targets), what ever Agilent points out as important GW Instek haves it too listed in the features of the GDS-2000A.
 
Yes GW Instek looks a bit as to be closed behind high walls, regarding reaching up the end user directly,
but as long the public complains level is zero, I am satisfied.
 
About RIGOL even if the pricing looks friendly and the product quality is almost acceptable, they do not pay much of attention in the detail.
Those constant Firmware bugs is unacceptable, for me as professional repair man, I did not get an Oscilloscope so to act as nurse to it.

And so my vote it did go to Simply reliable;)   
WWW.ITTSB.EU   Industrial Test Tools Scoreboard  (Product Reviews Blog) / Editor in Chief.

Offline Kiriakos GR

  • Administrator
  • Hero Member
  • *****
  • Posts: 2163
  • Country: gr
    • ittsb.eu
  • job title: Industrial Maintenance Electrician
I am aware that you love adventures as I do, here is screen shot regarding the noise of this scope with out probes conected to it.

use it as comparison to your RIGOL and share your thoughts.  ;)




WWW.ITTSB.EU   Industrial Test Tools Scoreboard  (Product Reviews Blog) / Editor in Chief.

marmad

  • Guest
use it as comparison to your RIGOL and share your thoughts.  ;)

My friend, low noise is certainly one area where the DS2000 beats the GDS-2000A and DSOX2000 very nicely  ;)  It has a nice low-noise front-end designed to go down to 500uV/div.

BTW, in your picture you have your DSO set to 1mV/div. That means the GDS-2000A is automatically limiting the bandwidth to 20MHz, as written on page 250 of the manual:
"When the vertical scale is set to 1mV/div, the bandwidth limit will be set to 20MHz automatically."

Here is a picture from my DS2000 at 1mV/div at full 200MHz bandwidth (no limiting filters)  ;)

 

ITTSB Links

Blog YouTube facebook twitter PayPal Donate Link

ITTSB Blog Sponsors

priniotakis.gr - Electronics division protosnet.com - Internet solutions

2016 Review samples received by

HIOKI EE Corporation - Japan Times Electronics ltd UK Vishay Foil Resistors POWERTRON resistors - VPG DER EE Electrical Instrument Co., Ltd.