Test and Measurement Forum

Documentation

Forum for discussing PicoScope (version 5)

Documentation

Postby ed800uk » Wed Aug 09, 2006 2:02 pm

Hi,
The User Guide tells us that:

[in Scope mode]
Settings
Options...
Advanced>>>

lets us set a value for:

Maximum samples per scope trace

I can see the point in setting a maximum, but it does not say on what basis the software only saves a lower number.

Could you explain how it chooses the number to store?

Regards,
Ed
ed800uk
Active User
Active User
 
Posts: 10
Joined: Mon Jul 03, 2006 2:24 pm

Postby Sarah » Thu Aug 10, 2006 3:04 pm

Hi Ed

The number that actually occurs depends on a few things, including the timebase that is selected. The idea of specifying in the software means that the software can then "aim" at getting this number of samples. It may not be able to get exactly that number however. The maximum in all cases is the size of the buffer and the number of active channels also needs to be taken into account.

I hope this helps.

Best Regards
Sarah
 

Samples Calculation

Postby Autonerdz » Wed Aug 16, 2006 8:26 pm

Hi Ed,

You may find this sample rate calculator useful:

http://www.autonerdz.com/java/SampleRateCalculator.html

Divide the max samples selected and max ADC speed by the number of channels deployed for the 3000 series.

For example: the 3423 unit running 4 channels with 512,000 samples selected:

128,000 samples
5MHz max speed

Then enter capture time and calculate.
Tom Roberts
(The Picotologist)
http://www.autonerdz.com
skype: autonerdz
THE PicoScope Automotive Authority
In North America
User avatar
Autonerdz
Guru
Guru
 
Posts: 141
Joined: Wed Nov 20, 2002 4:19 pm
Location: Washington State USA

Postby ed800uk » Thu Aug 17, 2006 10:13 pm

Hi,
the calculator is neat for some of my results.

I did a set of trials using 4 channels with 125000 in
Settings Options... Advanced>>>

I also lowered the display rate in
File Setup Display (but I don't know whether that affects it)

I found the similar results as the calculator for screen times up to 1 s.

Then the values were much lower, but getting bigger.

Then at 500 s, when the calculator says 76,000, the scope told me 124,999

Something is wrong somewhere.

Best regards,
Ed
ed800uk
Active User
Active User
 
Posts: 10
Joined: Mon Jul 03, 2006 2:24 pm

Samples

Postby Autonerdz » Thu Aug 17, 2006 11:55 pm

Hi Ed,

AHH, you are experimenting...I like that 8)

The calculator assumes that you are using the buffer value that you entered.

PicoScope changes its data acquisition at 200ms/div to a streaming mode and no longer used the internal memory in the ADC unless you are set to Block Mode under Settings/Options/Advanced. At 100ms/div and faster it's using Block Mode regardless.

In streaming mode the data is not stored in the buffer but streamed to the PC during the screen time. This rate varies from PC to PC and can be affected by the PC performing other tasks at the same time. My PC streams at 500s/sec with PicoScope 5.

At 5s/div, Block Mode no longer functions and you will stream regardless of the settings. This makes sense because now you will likely collect more samples streaming than you will using the internal buffer.

So, the reason that the calculator doesn't match your results is that you are streaming rather than using the buffer. The calculator will be very close to actual if you are using the buffer in Block Mode. It will be a few samples off sometimes due to complex scope operational variances, but very close.

Hope this explains.
Tom Roberts
(The Picotologist)
http://www.autonerdz.com
skype: autonerdz
THE PicoScope Automotive Authority
In North America
User avatar
Autonerdz
Guru
Guru
 
Posts: 141
Joined: Wed Nov 20, 2002 4:19 pm
Location: Washington State USA

Postby Michael » Fri Aug 18, 2006 8:42 am

Hello Ed,

Another piece of information to the puzzle.

When using Block Modes regardless of time domain, the scope interprets the best sampling interval from the number of samples you require and the total time to collect samples in. The fastest sampling rate is 20Ms/Sec. Each sampling rate is half the previous one working backwards:
20Ms
10Ms
5Ms
Etc...

When you move into longer timebases the scope will have to choose between 2 very different sample rates. At the 2 sec per div and slower, you will notice a difference in the total number of samples collected. It may even appear that the buffer is not being totaly utilised.

To bridge this gap, we have written a fast streaming mode to stream data upto 3Ms/Sec straight down the USB pipeline. Currently, this mode is only available to programmers who write programs with our drivers. Application support will follow in due course.

I hope this helps you to get the most out of your scope.

Best regards,
Michael
Michael - Tech Support
Pico Technology
Web Support Forum
Michael
Zen Master
Zen Master
 
Posts: 656
Joined: Thu Jul 07, 2005 12:41 pm
Location: St Neots, Cambridgeshire

Postby ed800uk » Sun Aug 20, 2006 4:26 pm

Hi Autonerdz,

Impressed, eh? If you knew my obsession with trying to spot the pattern between requested versus achieved capture counts, you'd probably be sending in the Nice Young Men In Their Clean White Coats...

Luckily you don't know how many hours I've spent, repeating trial runs, trying to become competent, confident and "fluent" with this new tool. Not just with this question of capture size, but with most other aspects too. But the breakthrough evades me. I don't want to stay a simple-minded user, relying only on basic features.

Even trying to use only the basics, I go back to the broken car, patiently reconnect the test wires, and fail, fail, fail to execute the intended plan; fail, fail, fail to capture the set of traces that I need: flummoxed by the sophisticated tool that does something that I didn't expect, in response to a "trivial" change.

Back on topic: the Sample Rate Calculator is really neat and will help me to overcome this example of deficiency in the documentation.

My second thought: why isn't the Sample Rate Calculator built into the software?

My third thought: is something back-to-front here? Since capture size and sampling rate are (IMO) so important, why bury then in the background? -- Question for Michael ;-)

Best regards to all,
Ed
ed800uk
Active User
Active User
 
Posts: 10
Joined: Mon Jul 03, 2006 2:24 pm

Sample Rate Calculator

Postby Autonerdz » Sun Aug 20, 2006 5:04 pm

Hi Ed,

My second thought: why isn't the Sample Rate Calculator built into the software?


Under measurements you can select # of samples. Think you already discovered this. So we kinda already have a samples calculator built in.

Here at Autonerdz, we designed the calculator to help compare the raw sample rate performance of different instruments. By analyzing the performance, techs won't waste time comparing features of scopes that can't perform. Or worse, buy something that can't perform. The calculator is included on our training CD set "The Nerd Series"

http://www.picotech.com/auto/automotive ... g-cds.html

It is also on our CD of PicoScope training movies we provide to our customers. And also on the web site for all to access, as you know.

Have you downloaded the new PicoScope 6 Beta?

http://www.picotech.com/support/about2712.html

You might like this. Not only is it the quantum leap forward in the software we have been waiting for to make full use of the 3000 series automotive scopes, you can put the samples selection menu right in the tool bar. In addition, the streaming is mega faster. I have collected almost a million samples per channel with it with minutes of complex data on a single screen. This is new ground for scopes and even in this beta stage, it's already beyond that which has ever been done :shock:
Tom Roberts
(The Picotologist)
http://www.autonerdz.com
skype: autonerdz
THE PicoScope Automotive Authority
In North America
User avatar
Autonerdz
Guru
Guru
 
Posts: 141
Joined: Wed Nov 20, 2002 4:19 pm
Location: Washington State USA

Postby Michael » Mon Aug 21, 2006 8:50 am

Hello Ed,

Tom is right, perhaps you need to consider PicoScope R6.0. The performance of this version outstrips many of the features of our PS 5.0, and even some it will not be able to support.

http://www.picoauto.com/automotive-software.html

Active sampling rate calculator is on the wanted list of widgets for PS 6.0. We will get there. Until then, I recommend using Toms' calculator and working from the sample count in the measurements. As I mentioned before, the scope determines the best sample interval. At longer timebases this will vary greater than at shorter timebases. Hence the fast streaming modes in PS 6.0.

Also, the Nerd series offers an excellent training package and should be considered by anyone working in the automotive sector.

Best regards,
Michael
Michael - Tech Support
Pico Technology
Web Support Forum
Michael
Zen Master
Zen Master
 
Posts: 656
Joined: Thu Jul 07, 2005 12:41 pm
Location: St Neots, Cambridgeshire

Postby ed800uk » Mon Aug 21, 2006 3:55 pm

I have read the new manual, looks good. It covers a lot of improvements.

I'm having a bit of trouble with the installation because it keeps wanting me to go back on the internet, but I have to swap out the network card to put in the USB 2.0 card.

Should I just leave the network card in until the end and then substitute the USB card?

Best regards,
Ed
ed800uk
Active User
Active User
 
Posts: 10
Joined: Mon Jul 03, 2006 2:24 pm


Return to PicoScope 5

Who is online

Users browsing this forum: No registered users and 0 guests