RF generator sweep settings for filter test

[cio74]

Hello,

Until I'll buy a proper SA I am playing with a combo made of an Anritsu MG3633A and a Picoscope 3206MSO.

I have made a simple crystal ladder filter with 3 poles, all xtal are 11.0592MHz and then set the generator to output a 0dBm signal

start at 11.00 MHz
stop at 11.10 MHz


The first graph has step size 10Hz and dwell time 10ms

The second step size 1Hz and dwell time 1ms

Apart from the later taking way longer to complete and a much cleaner plot the measured shape and the amplitude at top are very close.

Is there any rule to set dwell time, step size, span for better results?

Thanks.

[cio74]

Added 3 more crystals, 6 poles now, not exactly flat response

[Neganur]

Post a schematic with your tuning and coupling capacitor values, that graph looks like you're not tuned.

EDIT: and also, have you measured the xtal frequencies or are you just trusting the values printed onto their enclosure?

[uncle_bob]

Hello,

Until I'll buy a proper SA I am playing with a combo made of an Anritsu MG3633A and a Picoscope 3206MSO.

I have made a simple crystal ladder filter with 3 poles, all xtal are 11.0592MHz and then set the generator to output a 0dBm signal

start at 11.00 MHz
stop at 11.10 MHz


The first graph has step size 10Hz and dwell time 10ms

The second step size 1Hz and dwell time 1ms

Apart from the later taking way longer to complete and a much cleaner plot the measured shape and the amplitude at top are very close.

Is there any rule to set dwell time, step size, span for better results?

Thanks.

Hi

If you are going for a filter that is 1 KHz wide, a good guess is that 10 ms of dwell per point will work. The approach is to try 100 ms and then back off to 10. If the response does not change, you are ok. If it does then you need to slow down a bit. Bandwidth is not what really determines the number, so it's just a starting point.

Bob