[time-nuts] DMTD Ideas

[Magnus Danielson]

Brian,

Some quick comments...

Brian Kirby wrote:
> I have toyed with the idea of building a Dual Mixer Time Difference 
> setup for testing oscillators.  I have attached a drawing I made and I 
> have a few areas I need to clear up.
> 
> At Point #1 on the drawing (the output of the mixers) I expect to see 20 
> mhz and the 100 hertz beat note.  My plans were to put a 3 db attenuator 
> here and expecting it to provide a 20 mhz termination/match for the 20 
> mhz part of the signal.  Do you think this would be an adequate way to 
> terminate or does it need a better system ?

A passive termination/lowpass-filter is most probably best used. An 
inductor/capacitor network will essentially stop any 20 MHz and pass 
your 100 Hz beat-note.

If you look into the NIST papers you will find that they came up with 
that solution to improve the sensitivity of the mixer.

> The Op Amps at Point #2 would be a something like LT1000s or so and they 
> basically would only allow passing of the beat signal as they do not 
> have the response to handle the 20 mhz signal.  They would have to have 
> a lot of gain and also would simple RC low pass filter say for 120 hertz 
> be good enough ?  The DBMs have 7 db conversion loss, the RF port goes 
> into 1 db compression at +2 dbm.  The LO ports can take +13 dbm 
> absolute, and are recommended drive at +7 dbm.  What voltage range can I 
> expect out of the mixers (next question would be how much gain do the op 
> amps need....).

Since getting rid of essentially all of the 20 MHz sum signal can be 
done using trivial passive networks close to the mixer, you can 
concentrate on the beat note or 100 Hz difference frequency. Notice that 
the beat note output level actually depends on the loading network. 
Essentially the trick is to let the 20 MHz see a short (cap) while the 
beat note sees high impedance. The loss in level is essentially due to 
the traditional 50 Ohm impedance, which doesn't match the mixers 
effective output impedence and besides, for that frequency we don't 
really care about reflections as lumped parameter models may be used and 
we count highest voltage and not highest power.

So, let's say you end up with 50 V/s and you want the counter to see say 
5 V/us (really 5 MV/s or 5 MVHz) to avoid jitter-trigger then the 
slew-rate gain you need becomes 100.000. To achieve that the best 
approach is to use cascaded limited amplifiers with their bandwidths and 
gains adapted according to noise contribution.

If you look in the archived you will find that me and Bruce keeps 
nagging about these.

> At Point #3, I thought about using some type of Schmidt Trigger like a 
> 74AC14.  But is it necessary ?  The 5370B can be set to trigger from 
> zero to 2 volts or so.

Don't bother. Concentrate on the slew-rate amplifier chain and the 
needed slew-rate into the 5370B, which has the necessary final processing.

> If the 5370B can see one shot performance of 20 pS (2x10-11 at one 
> second) then a 100 hertz beat note should give 2x10-13 at 1 second, 
> 2x10-14 at 10 seconds, 2x10-15 at 100 seconds and so on.......correct ?

Yes. The 100 Hz beat note gives you a 10 M/100 = 100.000 gain in time 
resolution. However, you must consider the trigger jitter contribution 
and the very low slew-rate of the produced beat-note needs the 
gain-treatment such that the counter input does not produce any excess 
trigger jitter to spoil the performance.

Cheers,
Magnus