[time-nuts] Sub Pico Second Phase logger

[Bruce Griffiths]

Joe
Joseph M Gwinn wrote:
> Bruce,
>
>
> time-nuts-bounces at febo.com [Bruce] wrote on 12/15/2008 04:56:26 PM:
>
>   
>> Joe
>>
>> Joseph M Gwinn wrote:
>>     
>>> Bruce,
>>>
>>>
>>> time-nuts-bounces at febo.com wrote on 12/11/2008 07:44:01 PM:
>>>
>>>
>>>       
> [snip]
>   
>>>       
>>>> Using a passive splitter for the LO drives will gain at least another
>>>> 30dB in isolation between the 2 RF inputs if you use an appropriate
>>>> splitter.
>>>>
>>>>         
>>> True. 
>>>
>>> [Joe] I may have lost the thread here.  If we have one oscillator 
>>>       
> driving 
>   
>>> everything, one cannot have injection locking even if isolation isn't 
>>> perfect.  What isolation does gain us is a reduction in undesired 
>>>       
> phase 
>   
>>> shifts.
>>>
>>>
>>>
>>>       
>> You have 2 oscillators, the test source and the offset source, however
>> the >= 10Hz frequency offset between them means that the isolation
>> requirements are relaxed considerably.
>> If the offset oscillator is derived from the source then injection
>> locking doesnt occur.
>> I made the general comment to ensure that anyone following the thread,
>> who may be contemplating building a dual mixer setup with 2 sources very
>> close in frequency doesnt forget about the isolation requirements.
>>     
>
> I see.  We had two intertwined threads.
>
>  
>   
>>>>> [snip]
>>>>>
>>>>>
>>>>>           
>>>>>> The only configuration for which it makes any sense is an inverting
>>>>>> input amplifier with a finite input voltage offset.
>>>>>>
>>>>>>
>>>>>>             
>>>>> Why would non-inverting not work?  Both inputs source or sink bias 
>>>>> currents, and non-inverting presents a very high impedance.
>>>>>
>>>>>           
>>>> Non inverting amplifiers usually have lower noise and generally work 
>>>> very well.
>>>>         
>>>> I was only trying to come up with a preamp circuit for which the
>>>> comments in the Minicircuits application note on the effect of 
>>>>         
> amplifier
>   
>>>> input offset voltage made any sense.
>>>>
>>>>         
>>> Ah.  It may be hopeless.
>>>
>>> My reading was that they were worried about bias currents from the amp 
>>>       
>
>   
>>> flowing into the mixer and causing offsets, not amplifier offset 
>>>       
> voltages 
>   
>>> per se.  The amplifier offset voltage does not cause a mixer offset, 
>>>       
> and 
>   
>>> may be reduced by use of a chopper amp or very good balance.
>>>       
>
> By the way, I've noticed that Tek TDS3012B oscilloscope inputs can cause 
> offsets as well, again I assume from the bias currents.  The circuit has 
> the scope input in parallel with the Agilent 34410A 6.5-digit voltmeter. 
> With scope input set to DC, big effect.  Set to AC, small effect.  Set to 
> Gnd, no effect.  (Input is not grounded, so voltmeter is still happy.) 
> Didn't try changing the input volts/cm scale.  Anyway, I think that this 
> effect is what the mystery app note was trying to say.  A bias current 
> from the scope would cause a voltage offset that depended on the DC 
> resistance through which the bias current flowed, the DC load of the mixer 
> in this case.
>
>   
However the proposed remedy has little or no effect on the errors caused
by such bias currents (eg transistor base currents).
The series resistor could be reduced to zero without effect on the mixer
offset due to the bias current. However the preamp offset due to the
source resistance would be reduced.
>
>  
>   
>>>> If we design our own PCB then the AD7760 series ADCs are another
>>>> possible option.
>>>> These have a built in differential input differential 
>>>>         
> outputamplifier.
>   
>>> Yes.  But aren't we trying to use commonly available soundcards?
>>>
>>>
>>>       
>> Ideally yes, but they all seem to have built in performance limitations.
>> AFAIK the AP192 with its 4Vrms full scale balanced inputs with no
>> variable gain preamps or +48V phantom supplies seems to be one of the
>> best for this application.
>> Its major drawback is that its a PCI card located within a noisy PC.
>>     
>
> I think that there are many top-end firewire soundcards.  Whatever the 
> music folk like the sound of would be a good place to start - musicians' 
> well-trained hearing can be quite good.  At least above 20 Hz.
>
> Actually, the people that make the AP192 do have firewire and usb 
> offerings:
>
> <http://www.m-audio.com/index.php?do=products.family&ID=recording>
>
>   

I've looked at all of the M-Audio offerings.
The more expensive ones have built in preamps plus 48V phantom supplies,
which can be switched off, however the presence of the switched +48V
supply is perhaps an invitation to disaster.

I've also looked at the specs for several other high end sound cards.
Quite a few only have single ended inputs.
Maybe, I should document the various cards and highlight their
shortcomings etc for this application.

>> The 4V rms input allows the mixer preamp to use devices like the THAT
>> 1646 to drive the balanced sound card inputs without degrading the noise
>> floor too much.
>>     
>
> Or build an isolation amp with some gain, and kill two birds with one 
> stone?
>
>
>   
A low noise isolation amplifier with a frequency response down to 1Hz or
so without using a transformer may be difficult to do.
>> With a 1V rms full scale the noise floor degradation would be very
>> obvious when using a THAT 1646 (equivalent devices are even noisier).
>> It may be better to use a mixer preamp with a transformer coupled output
>> stage using hybrid feedback to achieve a low frequency cutoff below 1Hz
>> together with low noise.
>>     
>
> With a transformer, even if toroidal, keeping hum out may prove quite 
> difficult.
>
>   
High end (eg Lundahl LL1517) line output audio transformers come with mu
metal screens and metal foil interwinding shields.
>  
>   
>>>>>>>> Can alleviate [oddities at end of phase range} to some extent by 
>>>>>>>> driving a pair of such phase detectors so that their outputs are 
>>>>>>>>                 
> in 
>   
>>>>>>>> quadrature.
>>>>>>>>
>>>>>>>> One just selects the phase detector output that is in the linear 
>>>>>>>> range.
>>>>>>>>
>>>>>>>> The quadrature outputs also allow unambiguous assignment of the 
>>>>>>>> sign of any phase change.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>                 
>>>>>>> The Symmetricom 5120A does something very clever to alleviate this 
>>>>>>>               
>
>   
>>>>>>> problem.  Explained in US patent 7,227,346 and "Direct-Digital 
>>>>>>> Phase-Noise Measurement"; J. Grove, J. Hein, J. Retta, P. 
>>>>>>>               
> Schweiger, 
>   
>>>>>>> W.Solbrig, and S.R. Stein; 2004 IEEE International Ultrasonics, 
>>>>>>>               
> Ferroelectrics, 
>   
>>>>>>> and Frequency Control Joint 50th Anniversary Conference, pages 
>>>>>>>               
> 287-291.
>   
>>>>>>>               
>>>>>> I've read the patent.
>>>>>>
>>>>>>
>>>>>>             
>>>>> The paper is also worthwhile, and available on the web somewhere 
>>>>> (don't recall where, but google found the pdf).  I had to read the 
>>>>>           
> patent 
>   
>>>>> multiple times to figure out what's going on.  The correlation 
>>>>> approach is old as the hills, and only the digital phase detector 
>>>>>           
> was patentable.
>   
>>>>>           
>>>> It may be feasible to achieve the same effect by purely digital means 
>>>>         
> at
>   
>>>> least for low sample rates where FIR filters with tens of thousands 
>>>>         
> of
>   
>>>> taps are feasible.
>>>>
>>>>         
>>> It *is* feasible, and Sam Stein is doing it.  I've perhaps lost the 
>>>       
> thread 
>   
>>> here.
>>>
>>>
>>>       
>> No, I meant replace his 90 degree hybrids with a digital equivalent.
>>     
>
> I believe that his 90-degree hybrids are already digital.
I'm not convinced of that, if only because real time 10,000+ tap FIR
filters at 30+MSPS are probably still impractical.
>   I think that 
> there is a very short analog path, maybe just a buffer amplifier and a 
> bunch of fancy ADCs.  I had to reread the patent, to read between the 
> lines.  It is presented as if there were a number of physical components, 
> but one can also read it to mean that these are logical components 
> implemented in some kind of machine code. At 400 MHz, one would assume 
> that the signal processor must be a FPGA, probably one of the fastest ones 
> made.
>
> One example is the DDS core.  Somewhere in the patent text it in effect 
> says that one can also implement this by table lookup, and that if one 
> chooses the table length correctly, one can contain a full cycle (where 
> the "phase accumulator" returns to the exact same value) of full-precision 
> amplitude samples.  This allows one to eliminate all spectral spurs in the 
> DDS output, which is hard to do if there is phase truncation.
>
>
>   
>>> My understanding is that the 5120A is built upon a DSP or more likely 
>>>       
> FPGA 
>   
>>> (of unspecified make and model).  The 5125A will have a top frequency 
>>>       
> of 
>   
>>> 400 MHz, so the DSP and/or FPGA better be damn fast.  Little analog 
>>>       
> stuff 
>   
>>> remains.
>>>
>>>
>>>       
>> Although the 5125A appeared in the 2008 product catalog, it isn't on the
>> website yet.
>>     
>
> When I last talked to Symmetricom's sales folk about the 5120A and 5125A 
> (about a year ago), they said that Sam was still trying to get the 5125A 
> to work properly.  He may still be cleaning things up, as scaling up from 
> 30 MHz to 400 MHz is quite the jump, and may have forced a redesign.
>
> Joe
>
>
>   

Bruce