Hi If you need to know when the shifts occur in tens of ns, that significantly limits what you can do. Getting > 20 million readings a second is tough. If this is an off the air signal, it's noise is probably not good enough for you to estimate it accurately enough in that time frame.
If this is FSK, then modulation theory gets in the way a bit. 150KHz instantaneous shifts generate a lot of energy at the shift point. You also need to worry about dynamic range in addition to signal to noise. How about a simple FM discriminator with a bandwidth around 10 MHz and a center at 50 MHz?. Feed the output into an A/D and take it from there. It's probably as good as anything else you could do. Bob -----Original Message----- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Dan Kemppainen Sent: Thursday, August 11, 2011 4:11 PM To: time-nuts@febo.com Subject: Re: [time-nuts] Frequency Comparator Ideas needed Hi All, I could clairify things a little. My event of interest is basically a fast frequency shift of a signal that drifts between 1800 and 2600Mhz. There is slow drift of this signal of many hundreds of Mhz, with fast frequency shifts of approximatley +/- 150Khz. I believe that the 150Khz shifts are nearly instantaneous, but currently have no way to measure exactly how fast they occur. Slow drifts are corrected for by a loop in the down converter. Currently there is a first down conversion stage to ~900Mhz. This is then down converted again to an arbitrary frequency band (50Mhz in this example, but this could be moved from ~10Mhz to 200Mhz or greater). Obviously it would be advantageous to keep the low frequency band as high as possible, at least when trying to determine when the frequency shifts occur with any digital detectors. Probably for any analog detectors also. Bob, Obviously I'd like to get as close to the real zero crossing as possible, but I'm sure I don't need 0.6pS. If I could tell if the signal was within maybe +/- 15 or 20Khz, it may be acceptable. I could always double the frequency then down convert to increase the deviation. Basically, I have another timing device that will record every single event within 5nS, as long as they don't exceed ~1e6 events per second. That's the unit I'll feed the signal into, and what I need to keep happy. I did just learn about Gilbert cell mixer that works from DC to 500Mhz yesterday. I didn't realize they were available with that wide of a bandwidth. With this part as an option, it may be possible to do a quadrature detection around the several hundred Mhz range. That way subsequent low pass filters can have fairly high bandwidth. I'm sure the results here will probably be noisy, but maybe still acceptable??? However, I'd still like to it digitally if possible. Maybe even adding a second VCO with high modulation bandwidth to use a PLL with to track the input signal. Maybe then a phase detector comparing that PLL output and the reference signal and some high speed digital processing could prove useful. Not sure tho, still need to think about that. Probably would end up needing some multi Ghz flip flops to make that work... Ultimately this may not be possible digitally, but I thought if anyone here knew of anything it would be here... Thanks all! Dan _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.