[time-nuts] HP 5372A

Bob kb8tq kb8tq at n1k.org
Fri May 11 15:45:16 EDT 2018


HI

> On May 11, 2018, at 3:12 PM, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
> 
> Hi Don,
> 
> On 05/10/2018 02:59 PM, Don Murray via time-nuts wrote:
>> Hi Magnus...
>>  
>>  
>> Appreciate any help!
>>  
>> I participate in the ARRL Frequency Measuring Test,
>> and I would like to automate the measuring process.
>>  
>> I have two frequencies, A and B...
>>  
>> A is a GPS locked HP 3336B set to 455,000.000 Hz
>>  
>> I use the 3336B instead of the 455,000Hz BFO
>> signal from my GPS locked Racal receiver so
>> that I can measure AM stations without the
>> annoying beat note.  ;-)
>>  
>> B is an IF signal which should appear between
>> 455,000.000Hz and 455,001.000 Hz
>> 
>> I would like the 5372A to calculate the B minus A
>> delta, over a manually started and automatically
>> stopped measurement period of 110 seconds.
>>  
>> I would like a delta resolution of 0.001Hz to be
>> developed by the box.
> 
> There is no point in using two channels and a generator.
> You can do that straight out of the box just by locking it to the GPS,
> which you should do anyway.
> 
> Just measure the signal. Using the Math you can subtract or add whatever
> frequency you want. You can get A-B or B-A measures in Frequency mode,
> it's just not necessary.
> 
> Now, you want to use the single-shot run mode.
> 
> It should be relative trivial to setup a large number of samples with a
> suitable distance inbetween them.
> 
> Getting 1 mHz resolution isn't too hard that way, considering you have
> 110 s of time to do it.
> 
>> Now, there will be doppler.  I want the box
>> to analyze and give me a best statistical guess
>> for the B minus A delta.
> 
> Now, doppler is tricky, as the HP5372A in all its glory does not have
> built-in compensation for it, and besides, you need to figure out where
> the true zero is.
> 
> What the box will give you is a frequency estimation assuming doppler
> free conditions. If you want to do smarter things, you need to pull the
> data off the counter in binary form and post-process it yourself. The
> programmers manual for the HP5372A is a great teaching tool for how
> these beasts process stuff.
> 
>> BTW the B minus A delta will be added to the
>> frequency of my Racal receiver to give me the
>> best "guess" as to the correct frequency of the
>> target signal.
> 
> For sure. This is BTW something you can do in the Math part of the
> HP5372A, so you can have it crunch out the right value straight out.
> 
>> The Racal reads out to the Hz, and I am able to
>> accurately determine if the Racal is tuned above
>> or below the "target" frequency.
>>  
>> I will be tuned below the "target" and within
>> 1 Hz of the "target."
>>  
>> I assume I will be in the one frequency mode.
>>  
>> Can the box give me the desired resolution?
> 
> Sure, assuming the doppler limitation.
> 
>> What sampling period would be the best to use?
> 
> Use one short enough to get as many sample points over the 110 s you want.


The 110 seconds is the big gotcha here. What you really want is a unified record 
of the entire time they are transmitting. If it’s less than 2 minutes, you are fine. If
it runs for 5, 10 or 15 minutes that’s not quite so good. Stitching data together 
gets you back to binary dumps and a lot of post processing.  Once you do that, 
there really is no advantage over a SDR running on the same signal. 

Bob


> 
>> What statistical result would I use as my delta?
> 
> Mean value on the Statistics display. It's a linear regression result,
> which has good processing gain for frequency.
> 
>> Appreciate your help.  My head is spinning whenever I
>> get into the operating manual!
> 
> It's a beast, for sure. It took some time to master, but it's a lovely
> tool when you understand it and accept it's up-front complexity.
> 
> I could maybe take mine for a test-spin somewhat later.
> 
> Cheers,
> Magnus
> _______________________________________________
> 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