Re: [time-nuts] TICC / TimeLab fun
I can concur with this. When I have to measure 2 inputs using TICC, I always use lady heather for this reason. Either that dump TICC output into a text file, check it first, then feed it into other tools. --- (Mr.) Taka Kamiya KB4EMF / ex JF2DKG On Wednesday, April 29, 2020, 8:33:59 PM EDT, Mark Sims wrote: Lady Heather's TICC parser does not expect alternating ABAB readings and handles the ABAB vs ABBA situation properly. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
[time-nuts] TICC / TimeLab fun
Lady Heather's TICC parser does not expect alternating ABAB readings and handles the ABAB vs ABBA situation properly. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] TICC / TimeLab fun
Hi You have three inputs to the TICC: 1) The “frequency standard” input that delivers 10 MHz to the device. 2) The (likely pps) to channel A 3) the (likely also pps) to channel B The device runs a continuous time count based on the frequency standard input. Each edge it sees on channel A gets a time stamp indicating when that positive edge happened. It will keep on counting time for quite a while ( = I forget how long :) ). Yes, this *assumes* you have the TICC set to time tag. If it is set to something else, you will get whatever you asked for. If all you have is a pair of standards, put one into the 10 MHz and the other into channel A or channel B. That gives you the whole picture of what’s happening with the two standards. You have two devices and thus only one difference. It will show you that difference. Unless you have a third accurate standard, that’s as far as you can go. TimeLab needs to understand what it is being fed. If it does, you will get a plot of some sort. It may be the noise floor of the TICC. With a third standard, it depends a bit on just *where* you decide to feed it in. Does it go into the reference input at 10 MHz? Is it one of the “likely a pps” signals? In most (but not all) cases it does go into the 10 MHz input. First step in doing three is to do a noise floor check. Grab a piece of coax to offset the pps to channel A vs channel B. With a single source feeding both inputs (and the device set up properly) the data output (chan A vs chan B) will be the noise floor of the TICC. If it’s not, either the TICC has a setup gotcha or TimeLab does not understand what it needs to do. To keep everything happy coax in the range of 20 or 30 feet should work fine. The same process (sort of) could be done with a device that delivers both a pps and 10 MHz. You could feed it into the TICC the same way as a two standard test. The result would be the noise floor of the TICC plus the noise in the 1 pps divider chain. Is the divider in this or that device more or less noisy than the TICC? Who knows …... Bob > On Apr 29, 2020, at 1:12 PM, Chris Burford wrote: > > I'm trying to characterize a pair of GPSDO units using the TICC / TimeLab. > The issue I'm having is when both GPSDO units are wired simultaneously to > the TICC the resulting ADEV plot looks a little strange, beginning at > 3.74E-1. Both GPSDO units are referenced against my PRS-10 Rb for the clock > source. > > If the GPSDO units are wired individually to chA then all seems fine and the > plot begins with a more anticipated value of about 1.4E-8. I'm at a loss as > to where to proceed from here to get both units characterized > simultaneously. > > Thanks for the assistance. > > Chris > > > Channel.png>___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] TICC / TimeLab fun
Chris I don't have any of that kit yet but it seems to me that the two units have locked to each other somehow, a bit like the question I raised a while ago about crystals locking to each other. Just a pure guess though. John ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] TICC / TimeLab fun
Hi Chris -- Two possibilities: 1. Make sure you have TimeLab set for the correct mode (timestamp vs time interval), for two channels, and for the right scaling factor (for timestamps, scale by 1). Any of those can cause 1e0-type results. 2. If the two channels have timestamps that are within a few nanoseconds, there can be a crossover where the channel sequence doesn't maintain strict "ABABABABAB" order. TimeLab expects that order and if you get something like "ABBA" apart from bad '80s music you also get bad readings. The easiest way to avoid the problem is to split the TICC output stream into the two channels and then feed them into TimeLab separately. It is possible to do that and still get real time readings by using the "Read Streaming ASCII file" TimeLab acquire option, but it takes a little finagling to do it. [ BTW -- this is not really a "bug" in either the TICC or in TimeLab, it's just the way things work. I've spent much time thinking about addressing it in the TICC code, but there's no sure-fire fix that works across all possible operating scenarios. ] Hope this helps. John On 4/29/20 1:12 PM, Chris Burford wrote: > I'm trying to characterize a pair of GPSDO units using the TICC / TimeLab. > The issue I'm having is when both GPSDO units are wired simultaneously to > the TICC the resulting ADEV plot looks a little strange, beginning at > 3.74E-1. Both GPSDO units are referenced against my PRS-10 Rb for the clock > source. > > If the GPSDO units are wired individually to chA then all seems fine and the > plot begins with a more anticipated value of about 1.4E-8. I'm at a loss as > to where to proceed from here to get both units characterized > simultaneously. > > Thanks for the assistance. > > Chris > > > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.