[time-nuts] 10 MHz DMTD

Bob Camp lists at rtty.us
Thu Jan 13 17:57:27 UTC 2011



We have headed off on the error multiplier thing and not addressed a
question that came up. Since that thread is a bit tangled, probably better
to split this off.


DMTD = dual mixer time difference. 


You have two mixers, a local offset reference, and get two audio beat notes.
Both are processed to get the difference between them and to get the
contribution of the offset reference  oscillator to drop out. The nice thing
here is that you don't need a super duper reference, one that's just "in the
same league" as the standards being compared will do fine. Boards for the
DMTD are floating around the list. I'd guess that if there was enough
interest, we could get a buy going


Another thing mentioned was a simple beat note system. All that's involved
here is half of a DMTD board. If you have a "known best" local reference,
you simply mix it with the device you are looking at. That's a nice system
if you can offset the test unit, and are looking at ADEV. Not so useful for
setting a part on frequency, or for testing stuff that can not be offset.
The limitations of this approach are what got us the DMTD.


While a DMTD is normally looked at as an ADEV tester, it's very possible to
use one to set a device on frequency. The "error multiplication" can be
huge. A 1 Hz beat note system will give you 1x10^7. You would need two
computing counters (or similar PIC TIC stuff) to get the data into something
you could understand. Weather 5334 counters or PIC TIC's are cheaper,
depends on your scrounging skills. 






More information about the time-nuts mailing list