[time-nuts] NIST isolation amplifiers

Bob Camp kb8tq at n1k.org
Tue Nov 25 21:31:26 EST 2014


Hi

The reverse isolation of a “typical” pc layout for this sort of thing is maybe 60 db. Getting to 120 is far from simple. Achieving the 160 (or whatever) numbers you see in some papers is “isolation nuts” territory. The circuit its self can do great numbers. Coming up with a box that has 17 100’ long coax cables into it that isolates well …. good luck if you haven’t done is before. Good luck even if you have and you can’t afford to tool a fancy enclosure. 

————————

So back to the “what do you need” rant.

If:

1) You are running signals into the reference inputs on the back of test gear.

2) You are using BNC connectors and using something like RG-58 or RG-59

3) Your gear is all on one bench or a bench plus a rack

4) The longest run of cable is < 20’

5) Nothing much ever gets unplugged from the distribution line (or if it does you don’t care about a 100 ps burp).

Then reverse isolation is not all that big a deal. I’ve seen people run this kind of setup with passive power splitters. If they had 30 db of isolation I’d be amazed. The power splitter might not even be the weak link isolation wise. I’ve seen some really rotten cables and connectors being used. 

Now, if you have “many hundreds of feet” type runs, you stop talking to Mars when a 100 ps bump hits, or you routinely measure phase noise on 20 day runs with this setup - yes that’s different. Hopefully you have a lot of money in your wallet.

Bob

> On Nov 25, 2014, at 8:57 PM, Bruce Griffiths <bruce.griffiths at xtra.co.nz> wrote:
> 
> Another issue is that if even one output needs high reverse isolation and 
> low crosstalk, then even those outputs that arent so critical will also need 
> high reverse isolation and low crosstalk to avoid degrading the crosstalk 
> to the critical output.
> 
> Bruce
> 
> On Tuesday, November 25, 2014 07:54:02 PM Bob Camp wrote:
>> Hi
>> 
>> Harmonics are (in general) the least of your issues on a distribution 
> amp.
>> There is very little difference in ADEV or instrument performance at -20
>> dbc versus -120 dbc.  Since filtering is relatively easy, adding another
>> inductor or two is about all it takes.
>> 
>> ———
>> 
>> If you are going with the NIST approach rather than gates, remember 
> that
>> there are a few issues:
>> 
>> 1) These circuits tend to “sing like a bird” at UHF if built from leaded
>> parts. Often it’s tough to spot due to the output filter.
>> 
> 
> The small resistors in series with each CB stage emitter are useful in 
> suppressing such parasitics as is a low inductance ground connection for 
> each base.
>> 2) Past a handful of outputs, the input impedance of the circuit will 
> become
>> an issue. You will need a more complex approach.
>> 
> A low noise input amp driving a splitter can be useful in resolving that 
> issue.
>> 3) The isolation you achieve is far more dependent on the layout than 
> on the
>> circuit. You need a *very* good layout to achieve the numbers commonly
>> tossed around for the circuit. That’s much easier to do with SMT parts.
>> 
> Shielding each individual amp from the others (SMT or not) may be 
> necessary. 
>> 4) Any (hopefully) low noise circuit needs a quiet supply. This one is no
>> different. That’s not just the regulator, the rest of the feed (ground
>> loops etc) matters as well.
>> 
>> 5) There is a tradeoff between filter bandwidth and temperature induced 
> ADEV
>> issues. Going crazy on filtering will likely degrade your ADEV.
>> 
>> 6) The amp(s) as shown are not matched either at the input or the 
> output.
>> That may or may not be an issue to you. If it is, you will need to do some
>> mods to the circuit. I’d suggest at least a 3 to 6 db pad on the input and
>> output.
>> '
> Input pads will increase the PN floor.
> With slight modifications up to 6 such isolation amp inputs can be driven 
> by a single 50 ohm source.  
>> Bob
>> 
>>> On Nov 25, 2014, at 6:45 PM, Dr. David Kirkby (Kirkby Microwave Ltd)
>>> <drkirkby at kirkbymicrowave.co.uk> wrote:> 
>>> On 25 Nov 2014 23:10, "Bob Camp" <kb8tq at n1k.org> wrote:
>>>> Hi
>>>> 
>>>> For a modern build, the PZT3904’s and PZT2222’s are a pretty good 
> way to
>>> 
>>> go with this amp.
>>> 
>>>> For normal distribution to instruments, there’s really no need to do
>>> 
>>> anything this complex.
>>> 
>>>> Bob
>>> 
>>> I am also thinking about the construction of a  distribution amplifier
>>> with
>>> 15 or so outputs.  One thing that came to my mind, is that there may 
> be
>>> some point in  having one or two outputs where more money is spent. 
> Then
>>> if
>>> one thinks an item might be particularly sensitive to some aspect of 
> the
>>> reference,  one can use that.
>>> 
>>> One could for example have one or two outputs which have harmonics
>>> suppressed 100 dB, without going to unnecessary expensive on all 
> outputs.
>>> 
>>> Dave
>>> _______________________________________________
>>> 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 mailing list