Author Topic: Ridiculously small Rigol DS1000z series font size  (Read 45314 times)

0 Members and 1 Guest are viewing this topic.

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #150 on: February 01, 2016, 12:22:32 pm »
Well, I'm losing my close-focus ability as I age too, but I can't expect everything in the world to accommodate my vision problems, so I wear glasses when I need to see up close. Pretty strong ones too, up to 3 diopters and even more when I'm doing repairs or sewing my socks or something.

There is a lot of information on the Rigol screen, some of it unnecessary like the always-on left and right menus. So the fonts have to be small to display all that information.  I don't think this is going to change in some "ideal" firmware revision.  If you need a larger screen, like on higher-end Keysight or LeCroy scopes, you'll have to pay the price for it. But their fonts are pretty small too.

Same diopters, same thinking  :)

And when i solder i use, in addition, a desktop magnifying lamp.

Probably the vast majority of the DS1xxxZ scopes is made up of young people with perfect sight, otherwise this thread would have been longer.

I sure do not expect any magical FW revision and i'm starting to wonder if the reason of the forced presence of lateral menus is due to the need to mantain a reducted Z buffer size to limit FPGA computing effort.

Yesterday i was viewing the Dave's HMO1202 review (), i think that is a good example about how to organize a scope screen with a great readability and right amount of informations in a relative small size.   
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #151 on: February 03, 2016, 01:35:09 pm »
As I mentioned above, if you try to read more than 1420 bytes per transfer the TCP stack has to do resends and it slows it down quite a bit. With the TCP/IP frame overhead the ideal MTU size of 1514 bytes will speed things up making the LAN faster than USB transfers.

I read your post, but IIRC (I could be wrong, but I'm not around my scope to check at the moment) I get 3-4 second screen captures using USB. So your rate doesn't sound any faster to me.

EDIT: Actually, just double-checked and it takes ~2.5 seconds to get a screen capture to the PC and save it (using USB).

I' trying right now the RigolBildschirmkopie program, with my DS1074Z on USB interface i'm obtaining an average screen capture rate around 3.5/s with one channel and 2.5/s with four channels turned on.

It seems much faster than what you have seen, why ?
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #152 on: February 03, 2016, 02:22:15 pm »
I' trying right now the RigolBildschirmkopie program, with my DS1074Z on USB interface i'm obtaining an average screen capture rate around 3.5/s with one channel and 2.5/s with four channels turned on.

It seems much faster than what you have seen, why ?

You're talking about the SCPI command to send a screen capture (which always operated more or less the same on both DSOs), as opposed to the commands that transfer waveform memory data.

BTW, my tests were run when the DS1000Z series was relatively new. It's possible they've improved the speeds in later firmware - I have no way of checking.
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #153 on: February 03, 2016, 02:37:52 pm »
I' trying right now the RigolBildschirmkopie program, with my DS1074Z on USB interface i'm obtaining an average screen capture rate around 3.5/s with one channel and 2.5/s with four channels turned on.

It seems much faster than what you have seen, why ?

You're talking about the SCPI command to send a screen capture (which always operated more or less the same on both DSOs), as opposed to the commands that transfer waveform memory data.

BTW, my tests were run when the DS1000Z series was relatively new. It's possible they've improved the speeds in later firmware - I have no way of checking.

Hi marmad,

i was referring to your statement : "EDIT: Actually, just double-checked and it takes ~2.5 seconds to get a screen capture to the PC and save it (using USB).", where you talk about screen capture and not waveform memory data, but to be honest i do not know what scope you was benchmarking.

On the following web page :

http://peter.dreisiebner.at/rigol-bildschirmkopie-lan/

if you scroll down a little you can find a table with remote screen capture update rate  with RigolBildschirmkopie and it seems that under USB interface the DS1104Z is a little faster than MSO2302A, that i assume to be as faster as the DSO version.

I'm still considering the purchase of Rigol 2K DSO/MSO, so i would ask if you could gently check what update rate you can reach with your scope and the very same utility.

Thanks in advance,

Marco.
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #154 on: February 03, 2016, 03:00:32 pm »
...and it seems that under USB interface the DS1104Z is a little faster than MSO2302A, that i assume to be as faster as the DSO version.

I don't understand. The speed you mentioned is exactly the same speed I mentioned, for exactly the same screen size (800x480). What is faster?

"Actually, just double-checked and it takes ~2.5 seconds to get a screen capture to the PC and save it (using USB)"

I read 2.5 seconds in your statement, if it's the screen capture transfer time your rate is a slower (0.3 second).

Is it actually 2.5 seconds ?
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #155 on: February 03, 2016, 03:17:59 pm »
I read 2.5 seconds in your statement, if it's the screen capture transfer time your rate is a slower (0.3 second).

Sorry, I thought you meant 2.5 seconds when I read your post originally. And yes, 2.5 seconds was what it was when I last measured it (3 years ago). When I wrote I "just double-checked" above, I meant I checked this post that I wrote ~3 years ago:

https://www.eevblog.com/forum/projects/software-tips-and-tricks-for-rigol-ds200040006000-ultravision-dsos/msg171575/#msg171575

The RigolBildschirmkopie program fails to identify my DS2000 on the USB connection (even though Rigol UltraSigma sees it), so I can't test the speed using that software.

But I'm really not sure what getting a screen print 3 or 4 times a second has to do with the discussion we were having before. It doesn't even approach the 30FPS you claimed might be adequate before. Plus I could get all of the waveform data, plus all of the on-screen parameters - then recreate a DSO screen facsimile on a PC screen - much faster than 4 times a second.
« Last Edit: February 03, 2016, 03:20:28 pm by marmad »
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #156 on: February 03, 2016, 03:34:48 pm »
But I'm really not sure what getting a screen print 3 or 4 times a second has to do with the discussion we were having before. I could get all of the waveform data, plus all of the on-screen parameters - then recreate a DSO screen facsimile on a PC screen - much faster than 4 times a second.

Yes, you are right, has not a lot to do, i was only trying to understand if the DS2072A is slower or fater for screen capture transfer.

The main addition with screen capture over waveform data will be the intensity gradient (as you pointed out previously), if it would be possible to integrate it inside a remote control application like your and obtain something like 10-20 update per second would be great.

RigolBildschirmkopie uses WinUsb, in order to use it with your Rigol you have to use Zadig utility  (http://zadig.akeo.ie/downloads/zadig_2.2.exe) to install the support for it.
« Last Edit: February 03, 2016, 03:38:10 pm by markone »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #157 on: February 03, 2016, 03:43:46 pm »
The main addition with screen capture over waveform data will be the intensity gradient (as you pointed out previously), if it would be possible to integrate it inside a remote control application like your and obtain something like 10-20 update per second would be great.

I don't know how the screen capture functions on the DS1000Z, but, IIRC, on the DS2000 it either STOPs the DSO - or it seriously slows down the DSO's operation. In other words, it interferes/interrupts the waveform capture rate.
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #158 on: February 03, 2016, 03:56:12 pm »
With RigolBildschirmkopie program i get continuous screen capture transfer on pc screen and file write (if enabled) while the scope is running without any evident slow down, until i turn on the math, otherwise i get about3 FPS and i can operate freely the scope panel.

I could/should be the same for DS2000.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #159 on: February 03, 2016, 04:03:42 pm »
With RigolBildschirmkopie program i get continuous screen capture transfer on pc screen and file write (if enabled) while the scope is running without any evident slow down, until i turn on the math, otherwise i get about3 FPS and i can operate freely the scope panel.

I'm afraid just watching the DSO screen would not be enough to test this adequately. You'd have to have the DSO set to a timebase with a very fast update rate that taxes the DSOs resources (e.g. 20ns/div), while having a sufficiently fast changing signal or pulse stream that demonstrates the 30k (or whatever) wfrm/s.
« Last Edit: February 03, 2016, 04:05:45 pm by marmad »
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: Ridiculously small Rigol DS1000z series font size
« Reply #160 on: February 03, 2016, 04:24:15 pm »
I did try to measure waveforms per second on my scope.

My Rigol DS2000 has a spec of 50,000 wfrms/s, I can get up to 48,000 wfrms/s, If and only If I let the automatic memory depth to be 700 points, so to achieve 48,000 wfrms/s you have to be looking at a 5ns per division and the voltage and other things seem to matter as well.

If I manually select the minimal selection of 14K points I get 5,000 wfrms/s, and if I want anything with deeper memory then forget about it.

https://www.eevblog.com/forum/testgear/first-personal-'scope-purchase/msg681691/#msg681691

Edit: I didn't put on that post how I measured it, I did put a frequency counter on the trigger out BNC in the back.
« Last Edit: February 03, 2016, 04:27:01 pm by miguelvp »
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #161 on: February 03, 2016, 04:34:57 pm »
Ok, just to give an idea,  i made an animated PNG built with @ 5ns-1.00GSa/s - 12K mem captured screens, i set the frame rate in order to reproduce what i can see live on the screen.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #162 on: February 03, 2016, 04:43:29 pm »
My Rigol DS2000 has a spec of 50,000 wfrms/s, I can get up to 48,000 wfrms/s, If and only If I let the automatic memory depth to be 700 points, so to achieve 48,000 wfrms/s you have to be looking at a 5ns per division and the voltage and other things seem to matter as well.

I'm not sure this is the correct thread to discuss this, but virtually every DSO manufacturer always specifies maximum wfrm/s using the lowest memory depth possible with dots (not vectors) selected. Of course, it's the fastest speed - so they use the settings that would achieve the fastest speed.

 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: Ridiculously small Rigol DS1000z series font size
« Reply #163 on: February 03, 2016, 06:02:52 pm »
My Rigol DS2000 has a spec of 50,000 wfrms/s, I can get up to 48,000 wfrms/s, If and only If I let the automatic memory depth to be 700 points, so to achieve 48,000 wfrms/s you have to be looking at a 5ns per division and the voltage and other things seem to matter as well.

I'm not sure this is the correct thread to discuss this, but virtually every DSO manufacturer always specifies maximum wfrm/s using the lowest memory depth possible with dots (not vectors) selected. Of course, it's the fastest speed - so they use the settings that would achieve the fastest speed.

I'm just saying that achieving 30K wfrms/s is not going to be easy unless you can make the memory dept to be around 2.5Kpts in automatic mode.

Still 5,000 wfrms/s at 14Kpts is nothing to sneeze at since that is 70 million points per second, at max mem-depth it goes down to 35.5 wfrms/s but that's understandable for 56 Mega points, for 1.988 giga points per second!.

At  469,577 bytes per second maximum transfer time via USB/Network to the PC you must loose about 65.5 million points per second on the best scenario.

In other words, out of the 5,000 wfrms/s at 14Kpts you can only keep about 36 of those waveforms per second, being optimistic.

It wouldn't look that bad, but it's never going to be as good as what you see in the scope screen.


 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #164 on: February 03, 2016, 07:44:04 pm »
I made some check on waveform capture rate under continuos screen transfer load, what happens is that the "blind time" sometime grows (7.5ms to 25ms) while the capture burst maintain its frequency, around 17 Khz.

Useless to say, scope screen on PC monitor is waaay more clearer, not only for size but also due to pixel sharpness and you have visually lots of waveforms in it, way more than what is achievable with waveform data transfer.

A pity for  the slow rate (3FPS),  at 20-30FPS would great.

 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Ridiculously small Rigol DS1000z series font size
« Reply #165 on: February 03, 2016, 09:22:24 pm »
I made some check on waveform capture rate under continuos screen transfer load, what happens is that the "blind time" sometime grows (7.5ms to 25ms) while the capture burst maintain its frequency, around 17 Khz.

I guess you used quotes because you know that's not really the blind time, right?

Quote
Useless to say, scope screen on PC monitor is waaay more clearer, not only for size but also due to pixel sharpness and you have visually lots of waveforms in it, way more than what is achievable with waveform data transfer.

I can imagine it might be good for very slow moving waveforms - since all of the waveforms you are seeing are 250 - 400ms old.


EDIT: So, assuming a constant rate as shown in the screen shots, in the first example, your DSO is averaging approx. 14.6k wfrm/s - in the second example, the rate has dropped to approx. 10.6k wfrm/s (although with 25ms gaps at ~15Hz).
« Last Edit: February 03, 2016, 09:37:46 pm by marmad »
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #166 on: February 03, 2016, 11:27:39 pm »
I guess you used quotes because you know that's not really the blind time, right?

Exactly.
 
EDIT: So, assuming a constant rate as shown in the screen shots, in the first example, your DSO is averaging approx. 14.6k wfrm/s - in the second example, the rate has dropped to approx. 10.6k wfrm/s (although with 25ms gaps at ~15Hz).

Not all the gaps grow to 25ms under load, some remain almost at "unload" width, the capture rate loss is smaller  (see attached image).

I agree, 3 FPS are sadly not enough but now i'm starting to wonder which correlation exist between gap rate (about 20Hz) and scope lcd panel update rate (not waveform capture rate).

My bet is that is on par, during the gap time interval the FPGA writes the lcd controller frame buffer, so 20Hz screen update, at least for waveform graph section.
 

Offline markone

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: it
Re: Ridiculously small Rigol DS1000z series font size
« Reply #167 on: February 04, 2016, 03:25:38 am »
Well, i made a couple of tests in order to understand at which rate my DS1074Z refreshes the lcd screen panel waveform section content, what i found seems to confirm what i was thinking, i have around 20 FPS @ 5ns - 12K setting.

First one : i recorded some short movies of the screen with my panasonic SD700 @ 1920x1080 50p, so  i converted them in a png sequences then i visually checked every frame, there is one actual screen update about every 2 movie frames.

Second one (unscientific & subjective) : i built some animated GIFs with sequences of screen captures (made with RigolBildschirmkopie) at difference frame rate, the ones that more replicate the real screen scope "experience" are around 20-25 FPS.

It should be necessary a faster video recorder for an accurate measurement but i do not own one, anyway i feel that those number are quite correct.
In attachment an animeted GIF @ 25 frame per second, visually quite similar to what i see on the scope screen refresh rate wise, obviously its content is not correct due to skipped frames.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf