Building a phase noise measurement system for digital audio

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
On the software side I am using Linux, a dirty hack of JAAA from LinuxAudio-Downloads for the FFT analyser. Hacked to have a log display for frequency axis and to extend analysis bandwith down to 0.01Hz.
SISCO : LV2 Plugin Documentation to visualize the signal in time domain.
A PUREDATA patch to remove DC component and Jack the sound server to connect all this together.

If any interest I can give the dirty hacked sources of jaaa.

Joël
 
Mr. Riley is the author of the stable32 program that is the industry standard for analyzing
oscillator stability. Now that he is retired, stable32 is available for free btw.

He also has written some interesting papers on hands-on hardware that can be
duplicated by hobbyists.

< http://www.stable32.com/A 10 MHz OCVCXO and PLL Module.pdf >

There is also a homebrew phase noise test set by Bernd Kaa, DG4RBF.
It appeared in UKW-Berichte issue 4/2015, 1/2016, 2/2016 in German.

It probably also appeared in "VHF communications" in English, that is a
sister publication. I have a printout of the german version, that may
stem from a .pdf. Bernd Kaa's web site is
< Messgeräte im Eigenbau - Home-Page von DG4RBF >

Also from this posse:
< http://www.bymm.de/documents/39/PhaseNoiseMeasurement_V1_20.pdf >
< Miscellaneous Project documents > (Riley)
< Phasenrauschen von Oszillatoren messen > (German)

I have also done an oscillator board in the Riley style that can sloooow lock an
xtal oven to a, say, GPS reference, It can accomodate MTI 260 ,
Morion MV89A or HP 10811 oscillators.
Also a phase detector board that is not far away from Wenzel, or really HP3048.
Everything is switched with reed relays, that blows up the hardware somewhat,
it looks more difficult than it is.

But it In the last years I had no real interest to finish that since I had excessive
exposure to that stuff in the daytime job and easy access to an E5052B.

regards, Gerhard
 

Attachments

  • dsc_472.png
    dsc_472.png
    929.5 KB · Views: 538
  • dsc_0471.jpg
    dsc_0471.jpg
    106 KB · Views: 524
  • IMG_0009__1000_q50.jpg
    IMG_0009__1000_q50.jpg
    105.9 KB · Views: 529
Member
Joined 2007
Paid Member
@jdlvfr: Hi Joël ... This is just a brief message to say that I have seen your (interesting) posts and would like to give them & this subject more attention in due time. Personally I am, however, in these days looking into some amplifier circuits so I do not have the time to pick up on this right now. But I do look forward to doing so hopefully in not too long.

To this end: Do you have a link to the Morion oscillator (oscillator"s" - more than one I assume?) you used for calibration?

I reckon/assume you may also have posted your ADC design as a reply to my sentence:

I a.o.t. need a good-enough ADC to input the DC-Receiver's output to my PC

As it is, however, I have already made a PCM4202 ADC - only the ADC "was lost" due to a mishap - so I am waiting for a replacement. In any case - if you posted your design as a help to me then many thanks ;)

Cheers,

Jesper
 
Hi Jesper,

Morion MV89A comes from ebay, my choice was the cheapest seller. Sellers doesn't measure phase noise!

I think calibration should be done with the final oscillators, MV89A are 10MHz. But it's a good idea to first check the measurement system with known OCXO.

Joël
 
Thank you, now I can name the colours! I already knew the sky is yellow and the sun is blue. Or something like that, not really sure, maybe it depends on days...?

There was a French guy some 150 or 180 years ago who deduced that his
color vision was different from that of other people from the color of flowers
as seen by candle light vs. direct sun light.

Impressive achievement.

Gerhard
 
Hi Jesper,

Morion MV89A comes from ebay, my choice was the cheapest seller. Sellers doesn't measure phase noise!

The MV89As that I got from ebay are in a quite questionable condition.
Lots of scars, like they have been removed from their old home with
hammer & chisel.

BTW. I have cleaned up the control software for the Agilent 89441A.
It is a 1.9MB zip, including the 89441A GPIB manual. The program itself
is only a few pages of C. Email me if interested.

Having seen the creation date of the phase detector board above, I have
concluded that it's about time to complete my own phase noise tester.
Last weekend I have done most of the mechanical construction. It is quite
large because of the amount of Aircell coax cable needed & the relays
for the switchable delay line. It will be 2 channel so I can use the 89441A
for cross correlation.
The amplifier after the ring mixer is based on an earlier design of mine
using an IF3602 JFET. Ring mixers are SRA-3H since I had them.
The open box in the open box is one of the ring mixer/preamp assemblies.
 

Attachments

  • Auswahl_0021.jpg
    Auswahl_0021.jpg
    154.2 KB · Views: 459
  • DSC_0225.jpg
    DSC_0225.jpg
    525.1 KB · Views: 432
Last edited:
all of that is really good and inspirational. congrats !

jdlvfr, did you mention how you solved the VCXO ? are you using the one proposed by Rutgers or ?

No, I build two identical oscillators both with varicap for vfc. One vfc is connected to the pll output, the other one is grounded.

Ideally the reference oscillator should have better phase noise than the dut. If not, measurement gives the values of the worst one or if both are of same phase noise one should remove 3dB to measurement.
So by default I remove 3dB to accommodate worst case.

I have plan for cross correlation, maybe one day...

Joël
 
Has anybody used HP11848A phase noise interface in anger?

It is a bit clunky to control over GPIB but reasonably useable.
It is, basically, a mixer, LNA, PLL loop, set of of LPF and HPF filters, DACs and tonnes of relays that allows you to send measurement or test signals through almost any branch.
I like it so much I have two.
I have used it with HP3651A and custom 24bit 500ksps sampler.

Leo
 

Attachments

  • 11848A-block.jpg
    11848A-block.jpg
    599.5 KB · Views: 306
  • s-l1600.jpg
    s-l1600.jpg
    92.5 KB · Views: 316
Does anybody know where this mystery phase noise interface could have been made? It is a test set that take input from the mixer and has VCO driving voltage and sampler outputs.


I like it because it is fully analogue and very tactile.


Leo
 

Attachments

  • IMG_7325.jpg
    IMG_7325.jpg
    354.7 KB · Views: 272
  • IMG_7333.jpg
    IMG_7333.jpg
    350.7 KB · Views: 260
  • IMG_7331.jpg
    IMG_7331.jpg
    451.1 KB · Views: 270
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.