[time-nuts] Synthesized Signal Generator query

[John Ackermann N8UR]

A couple of comments about the PTS units...

While the potential noise floor is very low (depending on the quality of 
the reference oscillator), there can be significant spurs depending on 
how many non-zero digits you have dialed in.  With the analog multiply 
and mix architecture, the more digit stages are in use, the more spurs 
you get above the low noise floor.

The lower frequency units tend to have better noise than the ones with 
higher frequency ranges.  The PTS X10, 040, and 160 models are the 
quietest if you can live with their upper frequency limits (10, 40, and 
160 MHz respectively).

There's also one very tricksie version out there that is particularly 
interesting for time-nuts -- the PTS250/SX51 where the SX51 option is a 
switchable divider that turns a 250 MHz synthesizer with 1 Hz resolution 
into a 25 MHZ synthesizer with 0.1 Hz resolution, and a significant 
phase noise improvement.  Its noise floor is around -145 dBc/Hz, while 
the other units are in the -125 to -135 range.

I have a bunch of measurements of PTS boxen at
https://febo.com/pages/pts_synth/

(And, I'm getting ready to sell a few of the too-many PTS boxes in my 
lab, including an SX51... look out for a notice in the next week or so.)

John
----

On 3/7/19 7:15 PM, Alex Pummer wrote:
> that is a relative low noise synthesizer
> 
> 
>   Programmed Test Sources PTS 160 Frequency Synthesizer
> 
> 73
> 
> Alex
> 
> On 3/7/2019 11:04 AM, Richard (Rick) Karlquist wrote:
>>
>>
>> On 3/7/2019 8:56 AM, Didier Juges wrote:
>>
>>>
>>> The king of low close-in noise are the HP8662 (990MHz) and HP8663 
>>> (2GHz?)
>>> by about 20dB at 1kHz compared to the 8644A
>>>
>>> Didier KO4BB
>>>
>>
>> When I worked at Agilent, my lab had a half dozen HP8662's and some 
>> 63's.  They were not useful for any kind of high performance clock 
>> requirement, unless we divided them down with, for example, a 
>> Centellax (now Microchip) divide by 8 to 511 eval board.  That was SOP 
>> in those days.
>>
>> If you just want 100 MHz, there are inexpensive 100 MHz VCXO's
>> (about $25 at D/K or Mouser) that will blow the doors off of
>> an HP signal generator.  You can phase lock them to a 10 MHz OCXO
>> if necessary.  Besides 100 MHz, there are various other frequencies
>> available.  See Crystek and Abracon, etc.  These VCXO's put out
>> CMOS logic signals, so you have to convert sine to square.
>>
>> Rick N6RK
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to 
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>>
>> ---
>> This email has been checked for viruses by AVG.
>> https://www.avg.com
>>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to 
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.