[time-nuts] DMTD was: high rev isolation amps

Bob Camp kb8tq at n1k.org
Thu May 12 10:59:14 UTC 2016


Hi

The offset oscillator in a DMTD needs to be uncorrelated with the devices under test. If
it correlates with one or the other input, you have a single mixer system. That’s fine, just 
don’t bother with all the extra parts. A better way is to just grab a random OCXO and ground
the EFC. That gives you an offset and it’s uncorrelated with the inputs.

Isolation wise, the offset is the main issue. You have a path between the two mixers via the offset 
input. You might get 40 db of isolation with ideal terminations on all ports on the mixer. You might get
10 db with real terminations. Analog isolation amps will always be a bit better than digital parts. In the 
case of a DMTD, you want the noise on both channels to be identical. That’s a bit different than hitting
this or that specific number. In many cases, a ‘125 gate based amp chain will do fine. For the ultimate 
circuit, something around two dozen transistors (plus a lot of detailed RF testing)  will do a bit better :)

Bob



> On May 11, 2016, at 11:44 PM, Bob Stewart <bob at evoria.net> wrote:
> 
> Hi Bruce,
> 
> I was thinking of using the 125s as the zero crossing detectors.
> 
> Won't the DDS spurs cancel out across the 2 DBMS?
> 
> Bob
> --------------------------------------------
> On Wed, 5/11/16, Bruce Griffiths <bruce.griffiths at xtra.co.nz> wrote:
> 
> Subject: Re: [time-nuts] DMTD was: high rev isolation amps
> To: "Bob Stewart" <bob at evoria.net>, "Discussion of precise time and frequency measurement" <time-nuts at febo.com>
> Date: Wednesday, May 11, 2016, 10:33 PM
> 
> For
> DMTD service with an offset frequency of 100Hz or less the
> flicker phase noise of the mixer is more important than its
> wideband PN floor.The
> RDP series phase detectors are amongst the lowest PN noise
> mixer/phase detectors commercially available. They are
> considerably quieter than the TUF series mixers.To
> minimise noise both the RF and LO ports should be
> saturated.The
> specified IF port termination at the sum frequenncy should
> be used  to avoid degrading the isolation between the RF
> and LO ports.
>  For
> even lower flicker noise using a mixer assembled using
> 2N2222A's as diodes is even quieter as NIST have
> shown.
> 125
> gates have large flicker noise and using these in the front
> end of the zero crossing detector will increase the noise
> substantially over that obtainable with a well designed zero
> crossing amplifier chain using opamps.Using
> a DDS to generate the offset frequency raises the ugly
> sepctre of unwanted close in
> spurs, unless one chooses one of the magic frequency outputs
> that is free of such spurs.
> Bruce
> 
> 
> 
>    On Thursday, 12 May
> 2016 3:00 PM, Bob Stewart <bob at evoria.net> wrote:
> 
> 
>  Hi
> Bob,
> 
> OK, I can see that. 
> I had been wondering about how to build it out, but it seems
> like each module should be on its own PCB for isolation. 
> That keeps the costs down, except for the numerous SMA
> jumpers - and the eventual big box.
> 
> Bob
> --------------------------------------------
> On Wed, 5/11/16, Bob Camp <kb8tq at n1k.org>
> wrote:
> 
>  Subject: Re:
> [time-nuts] DMTD was: high rev isolation amps
>  To: "Bob Stewart" <bob at evoria.net>,
> "Discussion of precise time and frequency
> measurement" <time-nuts at febo.com>
>  Date: Wednesday, May 11, 2016, 8:44 PM
> 
>  Hi
> 
>  For extra added “Time Nut” grade fun:
> 
>  Clip the collector lead off
> of
>  the SOT-223 packages (between emitter
> and base leads). You
>  probably will gain a
> couple of db of isolation. (running
>  common
> base).
> 
>  Bob
> 
>> On May 11, 2016, at
> 8:24 PM, Bob Stewart
>  <bob at evoria.net>
>  wrote:
>> 
> 
>> Hi Bob et
>  al,
>> 
> 
>> I'm
>  finally
> drawing up a schematic for a DMTD and I thought
>  I'd use the NBS circuits for a lot of
> it.  So, how
>  about PZT3904 transistors in
> the isolation amps? 
>  They're in
> SOT-223 with the heat sink tab.
>> 
>> I figure to use some
> 
> version of the TUF-1 for the DBMs.  The first version
> will
>  be a bit breadboardish.  I'll use
> the PRS-45A as the
>  reference oscillator and
> the 8640B as the offset
>  oscillator.  If
> it's workable, then I thought I'd
> 
> put an Analog Devices DDS onboard with a small PIC to set
> it
>  up.  I haven't gotten as far as the
> zero-crossing
>  detectors yet, but I was
> thinking of using 125 gates.
>> 
>> Bob
>> 
> 
> --------------------------------------------
>> On Tue, 3/29/16, Bob Camp <kb8tq at n1k.org>
>  wrote:
>> 
> 
>> Subject:
>  Re: [time-nuts] high rev
> isolation amps
>> 
>  To:
> "Discussion of precise time and frequency
>  measurement" <time-nuts at febo.com>
>> Date: Tuesday, March 29, 2016, 6:08 PM
>> 
>> Hi
>> 
>> There were (and
>  maybe still are) SOT-89 versions of the
>> 
>  2N3804 and 3906. They
> will handle more
>> 
> 
> power than most of the other versions. That gives you
>  better
>> Vce on the
> string. They also
>  have less
>> package inductance which
> 
> helps tie the base to ground. If
>> 
> you
>  are building some of those circuits,
> they
>> are worth looking for.
>> 
>> Bob
>> 
>>> On Mar 29,
> 2016,
>  at 6:47 AM, Charles Steinmetz <csteinmetz at yandex.com>
>> wrote:
>>> 
>>> See below for schematics of the
> NIST
>  isolation
>> 
> amplifiers from 1990 and
>  1997.  NIST
> reported the
>> isolation as
>> 120dB.
>>> 
>>> I built isolation amplifiers
> similar
>  to these (with
> 
>> lower-noise power
>  supplies and biasing
> tinkered slightly for
>> better dynamic
> range), and with careful
>  construction
>> achieved isolation
> 
> substantially better than 120dB (see my
> 
>> 
>  post of 11/25/14 for more details).
>>> 
> 
> 
>>> Best regards,
>>> 
>>> Charles
>>> 
>> 
> 
> <NIST_120dB_isolation_amplifier_schematic_email.gif><Iso_amp_NIST_schematic_with_values_5MHz_1997_email.gif>_______________________________________________
>>> 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.
>> 
> 
> _______________________________________________
>> 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.
> 
> 
> 
> 
> _______________________________________________
> 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_lists.febo.com mailing list