Blocking mode vs. streaming mode

Post your .Net discussions here
Post Reply
cmos
Newbie
Posts: 0
Joined: Thu Apr 28, 2016 6:10 am

Blocking mode vs. streaming mode

Post by cmos »

Hi
Is there a big difference between the block mode and streaming mode when collecting continuous data? For example, assuming that mi signal is always over my setup threshold, the system would always be triggering. In case that we want to collect data at a maximum rate (single channel, 10^5 samplepoints * 20 (reading to average), 16bitRes, timebase=5 (aka 32ns)
I understand that in blocking mode I have dead times related to the time the system takes to setup triggers. I believe a better performance could be acquired with FastBlockTransfer. If I want to change my program with my previous describe condition (always over thresh), should I devote my time to implement FastBlockTransfer or direct streaming?

Your advise would be very appreciated.

Regards,

Cristian

Hitesh

Re: Blocking mode vs. streaming mode

Post by Hitesh »

Hi Cristian,

From the Programmer's Guide for the PicoScope 5000A/B Oscilloscopes:
In block mode, the computer prompts a PicoScope 5000 Series oscilloscope to collect
a block of data into its internal memory. When the oscilloscope has collected the whole
block, it signals that it is ready and then transfers the whole block to the computer's
memory through the USB port.
Streaming mode can capture data without the gaps that occur between blocks when
using block mode. Streaming mode supports downsampling and triggering, while
providing fast streaming at up to 31.25 MS/s (32 ns per sample) when one channel is
active, depending on the computer's performance. This makes it suitable for high- speed data acquisition, allowing you to capture long data sets limited only by the computer's memory.
With streaming mode captures at 12-bit resolution or above, the driver will allow you to capture data with a 32 ns sampling interval but note that the transfer rate over USB 2.0 will be around 15 MS/s.

I presume you mean Rapid Block mode when referring to 'FastBlockTransfer'? If so, that might be a better approach as it will capture waveforms with minimum re-arm time between triggers.

When you say that the signal is always over the threshold, what sort of trigger are you using? Are you checking if the level is above the threshold rather than rising through the threshold level?

Regards,

cmos
Newbie
Posts: 0
Joined: Thu Apr 28, 2016 6:10 am

Re: Blocking mode vs. streaming mode

Post by cmos »

Hi
Here are the answers to your questions.
I presume you mean Rapid Block mode when referring to 'FastBlockTransfer'? If so, that might be a better approach as it will capture waveforms with minimum re-arm time between triggers.

Yes, I meant rapid block.
When you say that the signal is always over the threshold, what sort of trigger are you using? Are you checking if the level is above the threshold rather than rising through the threshold level?
We are doing magnetic field measurements. The software provided to us was using the blockMode. However, I foresee collecting data at a faster rate asthe project develops. Do we have a trigger? Not really at this point. The signal is always present or not. I believe having a trigger just means that when won't be able to capture weaker signals, for example, in case we want to do a one dimensional field profile of the source.

I believe the initial design of our acquisition software was following one of the examples provided in the SDK. I find I am learning the different features of the PS5000A. Based on these reflections, maybe streaming mode is the way to go for my application. In any case, from your experience, will streamming be always faster than rapid block? or would the USB transfer be the bottleneck in this operation?

Thanks for your comments!
Cristian

Hitesh

Re: Blocking mode vs. streaming mode

Post by Hitesh »

Hi Cristian,

When using the SDK, it is possible to use rapid block mode captures without setting a trigger - the device would collect the next waveform as soon as it is ready to capture data again.

Have you tried capturing data using the different modes in the PicoScope 6 software first to give you some idea of what is possible? If you enable the Properties display via the Views -> View Properties menu option, you can see the sampling rate being used.

The USB connection is the bottleneck when using streaming mode captures although as I mentioned before it should be possible to have the device collecting at a slightly faster sampling rate than the USB bandwidth e.g. for a 32 ns sampling interval at 16-bit resolution. The driver will indicate if a particular sampling interval is invalid when you call the ps5000aRunStreaming() function.

Regards,

Post Reply