[time-nuts] FE5680 GPS Disciplined Controller Update
EWKehren at aol.com
EWKehren at aol.com
Tue Jun 24 20:15:31 EDT 2014
Lets be clear the 1 E-13 is a totally different project and does not relate
to the FE 5680 A. Yes most likely we will use the same universal
controller but with different code, same board. GPS crosses the 1 E-13 line at
100000 seconds little more than a day. I whish it was just temperature as I
mentioned we are able to hold the back plate of a FRK at 0.01 C. But that is
the easy part.
Bert Kehren
In a message dated 6/24/2014 8:06:52 P.M. Eastern Daylight Time,
kb8tq at n1k.org writes:
Hi
If:
1) You are after better than 1.0 x 10^-13 accuracy
2) You are getting 1 to 9x10^-9 at one second ADEV out of your GPS
3) You have a telecom Rb with 1 to 5x10^-10 temp coef over a 70C delta
4) Your Rb self heats 20 to 30C in still air
Here’s some math:
You will need at least 10,000 seconds to get a single frequency estimate
and likely 100,000 seconds.
You are likely to go through temp cycles in a room at a 30 to 90 minute
rate. (1800 to 3600 sec).
Your room temp swings are *way* outside your likely loop. The Rb will have
to deal with them by it’s self.
——
Your Rb is “sort of” compensated in the temcom units. It’s more like a
TCXO than an OCXO. 1x10^-10 over 50C would give you 2x10^-12 per degree C.
That may be better or worse than the sample you have. The “worse” really
comes in when you have one that’s a parabola or third order temp curve.
At least around here a room swing of 2 to 4 C is pretty normal with the
heat or air-conditioning turned on. That gets you into the 4 to 8x10^-12
swing range. In a typical garage you are at 10C and 2x10^-11.
If you want that to be *below* your 1x10^-13 goal, you have to knock it
down by about 100X.
——
Is the goal rational? Well this is Time Nuts …. It is roughly the sort of
goal Bert has said they are after.
Most of the lightweight Rb’s have a major ADEV hump when the temperature
compensation cuts in. Without *good* temperature stabilization, you will not
disable this correction. There’s no real way to know what the ADEV is
without this hunting until you do it. Because of the self heating (and
gradients), it’s not a real easy thing to do.
If you are using an ensemble of parts to get the ADEV, temperature likely
will correlate between them. You will not get the “group vote” to suppress
it the way you some other sources of ADEV.
——
Yes, I’m sure Bert can cross the T’s and dot the i’s better than I can,
but that’s a pretty good outline of the problem.
Bob
On Jun 24, 2014, at 4:34 PM, Tom Van Baak <tvb at LeapSecond.com> wrote:
>> However it is disappointing that no one has stepped up to tackle the
>> temperature problem. how many have looked at the temperature attachment
and
>> clicked on the N5TNL link. Let me make it clear that yes the GPSDO will
work but
>> there will be one or two orders of magnitude degradation without active
>> fan temperature control unless the internal temperature compensation
is
>> disabled.
>
> Can you clarify the "two orders of magnitude" claim? That's hard for me
to believe, I think, without seeing the ADEV plots or actual lab report.
>
> I mean, even a cheap XO or TCXO or OCXO can be disciplined against GPS
and achieve superb results. Temperature (or rather, temperature rate of
change) has little effect short-term. Temperature also has little to zero
effect long-term. So it's only in the, what, tau 100 to 1000 or maybe 10000
second range that temperature even matters. As long as the LO is locked to GPS;
I assume you're not talking about holdover.
>
> Obviously you'd want a slightly shorter loop time constant for a
non-temperature-controlled Rb than a fancy temperature-controlled Rb. But does
this really make a one or two *orders of magnitude* difference?
>
> /tvb
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at 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 at febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
More information about the time-nuts
mailing list