I have an AP System One Dual Domain. It has been used for spot measurements and THD sweeps and is fine. Passes all self tests (RED TICK BOX) very well indeed.
I just purchased the amazing USB interface from s1usb.com which makes the combination so good it prompted me to look at using the SYS322A for something important.
I needed a FFT sweep and started of with sweep of the internal gen.
I spotted some spurs on Channel B input at 6kHz, 12kHz, 18kHz which I dont recall.
But more worrying is that connecting a source, either AP internal Gen or expernal gen from say R&SUPV the presence of siganal lifts the noise floor way above that for the source or the analyser - by some 25 db or so.
More concering though is that there is a step increase in noise of almost 20dB afrom 6kHz to 18kHz - too much of a coincidence given the spurs maybe?
Dropping the input level by 10dB all the problems drop by 10dB as well - odd.
Similar behaviour on either CH1 or CH2 although the spurs are only really visible on CH2. And CH2 the step change in noise floor is slightly less pronounced.
Any thoughts?
Could easily be driver error - but if it is real hardware trouble I am really not sure where to start digging as the input amps etc must be all good to work OK for spot or swept THD measurements.
Where does 6kHz come into play??
This AP does not have the leaky brown electrolytics and all is clean and tidy inside so far as I can see.
Alan
I just purchased the amazing USB interface from s1usb.com which makes the combination so good it prompted me to look at using the SYS322A for something important.
I needed a FFT sweep and started of with sweep of the internal gen.
I spotted some spurs on Channel B input at 6kHz, 12kHz, 18kHz which I dont recall.
But more worrying is that connecting a source, either AP internal Gen or expernal gen from say R&SUPV the presence of siganal lifts the noise floor way above that for the source or the analyser - by some 25 db or so.
More concering though is that there is a step increase in noise of almost 20dB afrom 6kHz to 18kHz - too much of a coincidence given the spurs maybe?
Dropping the input level by 10dB all the problems drop by 10dB as well - odd.
Similar behaviour on either CH1 or CH2 although the spurs are only really visible on CH2. And CH2 the step change in noise floor is slightly less pronounced.
Any thoughts?
Could easily be driver error - but if it is real hardware trouble I am really not sure where to start digging as the input amps etc must be all good to work OK for spot or swept THD measurements.
Where does 6kHz come into play??
This AP does not have the leaky brown electrolytics and all is clean and tidy inside so far as I can see.
Alan
Attachments
Last edited by a moderator:
Well its clear something different is happening to each 6kHz chunk of spectrum somehow. The 12--18kHz region is very different from the 6kHz--12kHz if you look closely. That's go to be happening in the digital domain I'd have thought.
Thinking about it if the FFT were done in 6kHz chunks and some data corruption happening on a few input samples, this would show up as a uniform lifting of the floor maybe? So could be duff memory chip, or whatever.
I'd re-sit every socketed digital IC perhaps? And check all the connectors. The other possibility is data corruption down the line in the USB and beyond.
Thinking about it if the FFT were done in 6kHz chunks and some data corruption happening on a few input samples, this would show up as a uniform lifting of the floor maybe? So could be duff memory chip, or whatever.
I'd re-sit every socketed digital IC perhaps? And check all the connectors. The other possibility is data corruption down the line in the USB and beyond.
Looks to me like a relay or frequency setting component in a specific frequency range.
When you do an analog sweep, the generator and the analyzer are range-stepped with a fine stepping in each range, and if any of the (sub)range parts were off you'd get something like this.
But, if you use the digital generator and analyzer that would be a different story.
A very simple test would be to change the number of steps in the sweep to very low value or very high value and see if the step frequencies change.
Can you show the generator, analyzer and sweep panels for this test?
It would also be good to run the diagnostic routines, they're very good at identifying issues.
Jan
When you do an analog sweep, the generator and the analyzer are range-stepped with a fine stepping in each range, and if any of the (sub)range parts were off you'd get something like this.
But, if you use the digital generator and analyzer that would be a different story.
A very simple test would be to change the number of steps in the sweep to very low value or very high value and see if the step frequencies change.
Can you show the generator, analyzer and sweep panels for this test?
It would also be good to run the diagnostic routines, they're very good at identifying issues.
Jan