I'm a 25+ year metrology veteran (USN 10 yrs, Lowry AFB) with primarily RF background. I recently have had to look into phase noise cals (such as on HP 8662/3/5 - mostly 8665). We have an old HP 11848A phase noise interface, the adapter that connects to the front, and a floppy disk. We don't have the complete phase noise measurement system (don't have HP computer, don't have HP 3561A).
After downloading various manuals from Agilent and other places, I believe I understand you can use an HP 3585A/B LF Spec An. I feel really ignorant with this question, but I think if any group may understand this, you all will.
We have an HP 3585A. I'm trying to get up and running on this. Anyone have any info on how to do phase noise with what I have for generators such as HP 8665B, 8662, 8663? Again, I have an HP 3585A Spec An, HP 11848A Phase Noise Interface, the downconverter and a floppy disk (that I don't know what to do with). Is there a method to make this work, or is the HP 3561A a MUST? Also, if the HP 3561A is a MUST, what are some other alternate models that can be used?
A second half of this question, is there an alternate method for doing phase noise tests on the above generators. I have a very good overall RF background, just haven't done the high end ultra clean generators, and want some inputs on how to do so. My GIDEP membership is temporarily in limbo (working on that), so can't download from there. But I do have a MIDAS CD ROM with a lot of DOD procedures. So if you ref a 33K, Navair or TB procedure, I will likely have it.
And finally, in my lengthy question, are there some papers out there that could help me get up to speed on this. I do quite well and developing my own methods once I get the theory down.
Thank VERY MUCH for any help.
Here is what I know.
That 11848A is an anchor without a controller & the software that came with the thing. It set various switches that make it usable.
The 3561A was used for offsets up to 100 KHz, the 3585A took it to 40 MHz offsets. I don't know of a sub for the 3561A ( at least that the software could deal with) I do know the software was supposed to allow and 8566 to work in palce of the 3585A. OEM testing for the 8665B is 20 KHz offsets as I recall.
If you have an 8663A Ops & Cal manual (you can get a pdf from Agilent's website-8.98 Meg in size) check out the method they used in the performance test pgs 4-28 thru 4-32, you might be able use that as a basis for an alternative measurement technique. Also take a look at a 3335A test procedure (OEM), it has a testing method but the performance level for the thing isn't nearly as stringent as the 866X family. The 866X also is normally beyond the capability of spectrum analyzers with phase noise personalities that I have seen.
Hopefully someone can come along and tell me I'm full of it and give you a better answer but that's the best I know at present.
Good Luck
I ran into the same problem, but found a very useful tool/sw that works. I run this using a 8566B and even a 8565E up to 50Ghz. The only limitation is of the noise floor of the specan.
Check it out, in my opinion, KE5FX (the writer) did an excellent job.
www.thegleam.com/ke5fx/gpib/pn.htm
Good post Bryan, I was going to suggest using the 8663 manual also, but you beat me to it.
We're using the latest Agilent has (E440 something, my brain is shot this early) with the new E series generator and Z divider..sweet. Runs on XP and does an EFC reading very fast.
I'd advise any civilian lab that needs phase noise to invest in it..the 8566 method can be used, but it is tedious.
Mike
Quote from: mdbuike on 01-05-2008 -- 06:10:03
Good post Bryan, I was going to suggest using the 8663 manual also, but you beat me to it.
We're using the latest Agilent has (E440 something, my brain is shot this early) with the new E series generator and Z divider..sweet. Runs on XP and does an EFC reading very fast.
I'd advise any civilian lab that needs phase noise to invest in it..the 8566 method can be used, but it is tedious.
Mike
It is the E5504B.
If you do a lot of phase noise calibration. And you have the budget.
Take a quick look see at the Agilent E5052B.. This little toy is expensive.. But it makes the phase noise measurements in about 30 seconds.. It finds the signal.. Tunes it, gives you the results. (So get your coffee before you start.)
I was inpressed. No monkeing around with the settings and configurations through the software. Just select the screen (Phase Noise) and connect the signal, wait 30 seconds and you are done.. It will take you longer to print the data.
We wrote some automated software to calibrate these about 2 months back. Agilent calls it a Signal Source Anlayzer so it is a little hard to put 2 and 2 together.