Hi Bruce,
this sounds very good, and seems to fit my requirements quite well. :)
I will have a closer look to this concept.
I also had the idea to take the reference frequency, divide it, and mix
the division result again with the reference
to get an offset to the reference frequency which would give a higher
beat frequency which allows a reasonable measurement rate.
But it didn't turn out as i expected, because there was a lot of
'garbage' in the signals.
Best regards,
Wolfgang
Am 19.04.2012 21:51, schrieb Bruce Griffiths:
Using a dual mixer time difference system (either the digital dual
mixer time difference (DDMTD) or the analog variant (DMTD)) can easily
achieve the required resolution.
The DDMTD is relatively cheap to implement however it requires an
offset oscillator to beat against the 2 signals being compared.
However a DDMTD can use a 5MHz offset oscillator can be used with
5MHz, 10MHz, 15MHz ... input signals whereas a DMTD requires a 10MHz
offset oscillator to be used with 10Mhz input signals.
The DDMTD uses a pair of shift registers clocked by the offset source
where each of the 2 signals being compared is connected to the data
inputs of its shift register.
The time difference between beat outputs of the 2 shift registers is
then measured with relatively low resolution.
Some digital filtering of the shift register output transitions is
usually required.
A pair of 74HC164's will typically have a equivalent input jitter of
around 10ps or so, a 74AC164 will be about 4x quieter.
With a 5.000055MHz offset oscillator and 10MHz inputs the shift
register output beat frequency will be 110Hz.
It is usually advantageous to use an FPGA to implement the digital
filtering, timestamping and even the shift registers (although
external shift registers will have less crosstalk).
Bruce
_______________________________________________
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.