Test and Measurement Forum

Slow waveform update rate in scope mode

Post any questions you may have about our current range of oscilloscopes

Slow waveform update rate in scope mode

Postby hg_from_g » Fri Jun 02, 2017 8:45 pm

Dear Forum,

on my PicoScope 3403D (usb3.0) the waveform update rate is only about 13 waveforms/s in scope mode. Because I want to view an audiosignal with analog feeling, I need at least 30 updates/s. Is there any way to increase it, on what it depends, is it a willful limit?

My measurements are:
1ms/total, 1Gsa/s --> 13 wf/s
1ms/total, 125MSa/s --> 13 wf/s
1ms/total, 10MSa/s --> 14 wf/s
1ms/total, 1MSa/s --> 18 wf/s
1ms/total, 100kSa/s --> 20 wf/s

Why are there setting up to 60 wf/s in the preferences when never reach this?

Thank You so much!

hg
hg_from_g
Newbie
 
Posts: 0
Joined: Thu May 25, 2017 10:27 am

Re: Slow waveform update rate in scope mode

Postby Martyn » Fri Jun 02, 2017 9:18 pm

If you post a psdata file with your settings we can run it against a device here and see what is happening.
Martyn
Technical Support Manager
Martyn
Site Admin
Site Admin
 
Posts: 2716
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Slow waveform update rate in scope mode

Postby hg_from_g » Fri Jun 02, 2017 9:34 pm

Here is the file.
Attachments
20170602-0001.psdata
Audio Waveform Settings
(445.94 KiB) Downloaded 14 times
hg_from_g
Newbie
 
Posts: 0
Joined: Thu May 25, 2017 10:27 am

Re: Slow waveform update rate in scope mode

Postby Martyn » Fri Jun 09, 2017 11:10 am

Can you repeat the test for me, but this time use Views->Channels and disable the display of all channels. Run the test and then check the update rate. This will prevent the software from requesting downsampled data from the device to display on the screen after each waveform has been collected before starting the next. Once the collection is complete you can re enable the display of the data.

I have been getting 40+ wf/s at 1GS/s, although the speeds do vary a bit depending upon what other applications on the PC are doing, and also the specification of the PC being used.
Martyn
Technical Support Manager
Martyn
Site Admin
Site Admin
 
Posts: 2716
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Slow waveform update rate in scope mode

Postby Martyn » Fri Jun 09, 2017 11:11 am

Can you repeat the test for me, but this time use Views->Channels and disable the display of all channels. Run the test and then check the update rate. This will prevent the software from requesting down sampled data from the device to display on the screen after each waveform has been collected before starting the next. Once the collection is complete you can re enable the display of the data.

I have been getting 40+ wf/s at 1GS/s, although the speeds do vary a bit depending upon what other applications on the PC are doing, and also the specification of the PC being used.
Martyn
Technical Support Manager
Martyn
Site Admin
Site Admin
 
Posts: 2716
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Slow waveform update rate in scope mode

Postby hg_from_g » Sat Jun 10, 2017 9:36 pm

No problem, here are the tests with Views/Channels/all disabled.

My pc-system is:
AMD Athlon 5350 (4 x 2.1GHz, 2578 cpu passmarks) + Radeon R3 onboard graphics
Windows 7 64Bit, 4GB RAM

Results on onboard AMD usb3.0:
1ms/total, 1Gsa/s --> 17 wf/s
1ms/total, 125MSa/s --> 17 wf/s
1ms/total, 10MSa/s --> 17 wf/s
1ms/total, 1MSa/s --> 17 wf/s
1ms/total, 100kSa/s --> 17 wf/s

Results on onboard AMD usb3.0 (usb2.0 mode):
1ms/total, 1Gsa/s --> 22 wf/s
1ms/total, 125MSa/s --> 20 wf/s
1ms/total, 10MSa/s --> 20 wf/s
1ms/total, 1MSa/s --> 20 wf/s
1ms/total, 100kSa/s --> 20 wf/s

Results on pci-e Fresco Logic usb3.0:
1ms/total, 1Gsa/s --> 22 wf/s
1ms/total, 125MSa/s --> 25 wf/s
1ms/total, 10MSa/s --> 25 wf/s
1ms/total, 1MSa/s --> 25 wf/s
1ms/total, 100kSa/s --> 25 wf/s

Results on pci-e Fresco Logic usb3.0 (usb2.0 mode):
1ms/total, 1Gsa/s --> 22 wf/s
1ms/total, 125MSa/s --> 20 wf/s
1ms/total, 10MSa/s --> 20 wf/s
1ms/total, 1MSa/s --> 20 wf/s
1ms/total, 100kSa/s --> 20 wf/s

Another test on my good old Intel Atom N450 (1,66GHz, 297 cpu passmarks) on WinXP 32bit sp3 with usb2.0:
1ms/total, 1Gsa/s --> 10 wf/s
1ms/total, 125MSa/s --> 10 wf/s
1ms/total, 10MSa/s --> 10 wf/s
1ms/total, 1MSa/s --> 10 wf/s
1ms/total, 100kSa/s --> 10 wf/s

Enclosed screenshots of the cpu load of Athlon 5350 at 1ms/total, 1Gsa/s. With or without viewing of channel A the cpu load is about the same.
cpu_load_view_channel_a.png
cpu load of PicoScope with viewing channel A is lower than 30%

cpu_load_view_no_channel.png
cpu load of PicoScope with viewing no channel is lower than 30%


What is Your pc-System? Why the waveform update rate is not faster when using usb3.0? Why the Athlon 5350 is only twice as fast as the Intel Atom?

Thanks a lot!
hg_from_g
Newbie
 
Posts: 0
Joined: Thu May 25, 2017 10:27 am

Re: Slow waveform update rate in scope mode

Postby bennog » Sun Jun 11, 2017 8:22 am

Because you system has quad cores the value you see needs to be multiplied by 4.
So 1 core (the core picoscope is running on) is utilized by 96%, so you are at the max of your PC.
Some processes can be split on multiple cores / threads. others are not.
But your assumption of 24% is wrong !

Benno
bennog
Advanced User
Advanced User
 
Posts: 89
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: Slow waveform update rate in scope mode

Postby hg_from_g » Sun Jun 11, 2017 12:21 pm

@Benno
Why my assumption is wrong? The max of my PC would be 4 x 100%.

Can it be true that PicoScope doesn't really supports multicore? But the manual says "...will greatly benefit from a multi-core processor". At the moment it doesn't! One core is at the end and the others are bored.
minimum system requirements.png


@Martyn
What CPU do You working on to get > 40 wf / s? USB3.0 doesn't speed up the waveform update rate in scope mode (short timebase)? Isn't it?

hg
hg_from_g
Newbie
 
Posts: 0
Joined: Thu May 25, 2017 10:27 am

Re: Slow waveform update rate in scope mode

Postby Martyn » Mon Jun 12, 2017 6:07 am

My PC is an Asus Laptop with i5-4210u 8.00GB WIndows 10 , so nothing particularly special, it is kept up to date, including the hardware drivers to ensure the USB ports are working well.

We used to specify the PC requirements, but it is easier to say that if the PC can run WIndows 7 or above then it's specifications will be sufficient to run PicoScope 6.
Martyn
Technical Support Manager
Martyn
Site Admin
Site Admin
 
Posts: 2716
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: Slow waveform update rate in scope mode

Postby hg_from_g » Mon Jun 12, 2017 7:44 pm

The simple inexact comparison shows:
Single core performance Intel i5 4210u = 3400 Passmarks / 2 cores = 1700 Passmarks -> 40 wf / s
Single core performance AMD Athlon 5350 = 2600 Passmarks / 4 cores = 650 Passmarks -> 13 wf /s

...that only single core performance speed up PicoScope and I think you know. Until Picotech doesn't really wangle multicore and usb3 it would not be a tool i can work with. The note in the manual "will greatly benefit from a multi-core processor" isn't the truth. Sorry, but I think You should fair play...

hg
hg_from_g
Newbie
 
Posts: 0
Joined: Thu May 25, 2017 10:27 am

Re: Slow waveform update rate in scope mode

Postby Martyn » Tue Jun 13, 2017 7:00 am

There are many things that can affect performance, your test showed a difference between different USB 3 cards and how they operate in USB2 mode.

Was this just using a USB2 cable, and how did the USB2 mode on the onboard AMD port perform quicker than native USB3.

Performance of the O/S, and any other services that are running will also have an impact. I made sure that I turned off or disabled anything that was unnecessary, and made sure that my system was fully updated, and with the latest hardware drivers installed.

I will refer your comments about the multi core to the development and documentation teams, and reply back when I have an answer.

If you would like to discuss this further please email me at support@picotech.com
Martyn
Technical Support Manager
Martyn
Site Admin
Site Admin
 
Posts: 2716
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots


Return to USB PC Oscilloscopes

Who is online

Users browsing this forum: No registered users and 0 guests