PicoScope 7 Software
Available on Windows, Mac and Linux
QWG run time vs frequency
QWG run time vs frequency
I'm using an AWG at low frequency (<100mHz) to apply a programmed waveform over a period of seconds. I'm finding two issues:
1) the program run time is slightly off - for example a 50mHz setting produces a run time of 19.7 sec rather than 20.0 sec.
2) the run time appears to occur in steps as a function of frequency, rather than continuous. For example if I use a 12,000 line program, run at 50mHz, the test time is 19.7 sec. If I run the same program at, say, 40mHz, the run time is NOT 1/0.040 = 25 sec, instead it is still runs at 19.7 sec. In fact the program does not run at any different time length other than 19.7 sec (e.g. at frequencies of 45mHz, 40mHz, 35mHz etc.) until I set at 30mHz at which setting the program runs for 39.4 sec (whereas it should be at 1/0.030 = 33.3 sec
3) the run time depends on both the frequency AND the number of lines in the AWG. For example, here's the run times at a setting of 50mHz as a function of lines in the AWG:
- 12,000 lines - 19.7 sec
- 13,500 lines - 22.1 sec
- 15,000 lines - 24.6 sec
- 16,000 lines - 17.5 sec
Can anyone explain what's going here? Is this a bug or a feature?
1) the program run time is slightly off - for example a 50mHz setting produces a run time of 19.7 sec rather than 20.0 sec.
2) the run time appears to occur in steps as a function of frequency, rather than continuous. For example if I use a 12,000 line program, run at 50mHz, the test time is 19.7 sec. If I run the same program at, say, 40mHz, the run time is NOT 1/0.040 = 25 sec, instead it is still runs at 19.7 sec. In fact the program does not run at any different time length other than 19.7 sec (e.g. at frequencies of 45mHz, 40mHz, 35mHz etc.) until I set at 30mHz at which setting the program runs for 39.4 sec (whereas it should be at 1/0.030 = 33.3 sec
3) the run time depends on both the frequency AND the number of lines in the AWG. For example, here's the run times at a setting of 50mHz as a function of lines in the AWG:
- 12,000 lines - 19.7 sec
- 13,500 lines - 22.1 sec
- 15,000 lines - 24.6 sec
- 16,000 lines - 17.5 sec
Can anyone explain what's going here? Is this a bug or a feature?
Re: QWG run time vs frequency
It will be easier to answer your questions in a single topic so we can use your other one topic41767.html
or you can email the support team directly support@picotech.com
or you can email the support team directly support@picotech.com
Martyn
Technical Support Manager
Technical Support Manager
Re: QWG run time vs frequency
it is quite simple
the sample time for the first 3 measurements is about 1.64 ms per sample
the last measurement is 1.09 ms per sample.
If you know this you can make exactly what you want.
The step in sample time is probably due clock dividers of the main clock.
So if you know the DAC conversion rate is 1.64 ms you can make any waveform you want at exatly the interval you want.
Benno
the sample time for the first 3 measurements is about 1.64 ms per sample
the last measurement is 1.09 ms per sample.
If you know this you can make exactly what you want.
The step in sample time is probably due clock dividers of the main clock.
So if you know the DAC conversion rate is 1.64 ms you can make any waveform you want at exatly the interval you want.
Benno
Re: QWG run time vs frequency
I have found the same issue in PicoScope 6.14.44.5870 controlling model 5444B. I have a 4096-point AW and want this block to have exactly 1.0 s duration. Using "1 Hz" frequency on the AWG control panel, the burst length for a single shot is approx 0.895 s, and setting fractional Hz value to try to compensate for this shorter-than-expected burst does not work. If I set a much faster repeat frequency (1 kHz, 100 Hz, 10 Hz) the burst length appears correct as expected. The issue seems to affect only <10 Hz values of repetition frequency. Any workaround or fix for this?