The way that the 8920 generates the low frequency is somewhat counter intuitive.

There are three paths in the Outpur Assembly

The synthesizer runs from 500-1ghz? ?so that range is just straight amplification for path 1 and the 500 -1G range
Path 2 uses a? /2 divider on the Synthesizer so you get 250 -500 Mhz range?

Path 3 which is what you are having problems with? mixes the synthesizer with 1Ghz? to get the .4 to 250 range
For a 1 Mhz? output you have 999 Mhz mixing with 1Ghz.
The higher the output frequency needed the lower the Synthesizer frequency needed.

This is the more complex of the three but at least at this point you know your 1Ghz and Synthesizer? signals are correct going into the assembly??

You can use a 9 Pin D extender which gets you everything but the RF signals? ...usually use some small coax and RS-232?Male and female pins to extend those as needed for the inputs.

You should have an external Spectrum Analyzer? ....without that I wouldn't bother.

It is better to fix the original module to maintain the calibration

The replacement Output Module arrived and the 8920A now passes all self tests.? Thanks to all for their assistance.

If i can figure how to crack open the Output Module, I may try to find the culprit.? Because it failed at 1 MHz and was on the lower edge of acceptance at 10 MHz, is it reasonable to first suspect a capacitive coupling problem???

Paul, W9AC

开云体育

Description: Connector body
Agilent Part No: 1251-4782
Tyco/AMP Part No:350777-1
Quantity: 1


Description: Contact Pins
Agilent Part No: 1252-0385
Tyco/AMP Part No: 350547-6
Quantity: 2


Copied from Agilent group ...should be on Digikey or Mouser ...I have ordered in the pass

开云体育

Thanks George.

开云体育

Good advice.? I did try re-seating the modules and re-tighten the bottom plane SMA connectors.? No go.? 1 MHz still fails.? I have ordered two spare output modules to try.? They should arrive later in the week.? The diagnostic result says the problem is localized to the Output Module but I realize that depending on switch paths, this could be misleading.?

Paul

just a quick note on my post.

The DUPLEX port is capable of higher level outputs ....like 7dbm IIRC?

The RF I/O can make it up to 0dbm? IIRC in some cases .

Please refer to the documentation

The short version is that you can try and set the output to whatever you want but if the 8920 cannot make the level on the port you selected that message will come up

looks like you have a bad output assembly but since it's on the low frequency side I would try to re-seat the output , referencne and synthesizer boards and also re-seat and? tighten the rf connections on the backplane? ?these are the coax connections that have a 5/16 nut (IIRC)? ?...you can loosen the nut and pull out the coax and re-seat.


Could be the synthesizer output is dropping at the higher end and due to the different path failing the diagnostic.



The second error message is kind of the same thing where it cannot get to the output level you specified for the port selected.

?if the DUPLEX port is selected you can get to higher outputs? like -12dbm? while the RF I/O port is like -20dbm max.

If you try to set 0dbm on the RF I/O port you will see that error for example.

This may be another? symptom of a bad output assembly as well as other issues with the RF level not being correct.

I ran a comprehensive RF-DIAGS error test on a recently acquired HP 8920A.? One failure occurred and it reports as follows:

"Output Assembly
Output Frequency Range Closed Loop
1 MHz Lower Vdc: 0.750V? Upper Vdc: 2.50V? Measured:? 0.588V **? "

So, it appears the voltage measurement at 1 MHz is below spec range.? All other frequencies up to 1 GHz test OK and are within range except for 1 MHz as shown above.

The error also shows:

"Change RF Generator Amplitude, Output Port, or Atten Hold (if on)."

The test routine doesn't allow for on-the fly changes that I'm aware of.? I did try going to the RF GEN menu to make those changes and the result is the same.

Any clues here?? Thanks in advance.

Paul


?

Hi Angelo!

Thanks for sharing your equipment calibration data, I'm sure it will
help me a lot! I'll let the group know when/if I manage to get something
usable out my "rainy winter" project (still some work to do)...

Not sure why 8920 spectrum analyzer cal. varies so much on your test
equipments compared to 8922.

From 8920/8922 block diagrams in the assembly level guides, both Spec
An have very similar architecture. I do not know how schematic are
close, but I'd be surprised seeing totally different designs. The
difference in the span between 8920 and 8922 comes (I think) from
something else: on 8920: below 2MHz span the Spec An local oscillator is
swept (that's why receiver continues to work in such case). Above 2MHz,
spec an LO is fixed and receiver LO is swept... Not sure how that works
on 8922...

Since the Spec An module self-calibrate itself from end to end on a
regular basis (spec an trace refresh briefly freeze from time to time),
the only reason the drift could be caused by spec an module is a drift
in cal tone generator... However I do not see cal source being
responsible for >2.5dB drift by itself... but who knows.

Other modules may also cause such drift: input, receiver and
measurements assemblies. I'd say the first two ones are the most likely
offenders (no self calibration there). Difficult to say for sure before
probing inside the test set...

Ciao,

Jeff

Hi Jeff,

I am sending to you calibration data files from two different HP8920 of mine.
One seems manufactured in USA, the other from UK.
The second developed the same problem of yours, with the notorious explosions on RIFA caps shooting as a machine gun.
They are both in pretty general and cal conditions, but both have major calibration problems on spectrum analyzer, that are very far from plus or minus 2.5 dB as stated on specs.? ?Love these sets and really would like to fix them !

Just for info, I have also a couple of 8922 an "M" and an "S" model. Calibration on this model seems very tight in every specs, and not so easy to loose as for 8920.? ?Perhaps due to s.a. reduced span...

Jeff, many, many Thanks for Your endeavour and great work !

Ciao,
Angelo? ?iK0-IBI

Hi Aretrac,

Thanks a lot for taking time downloading your test set cal data and sharing them. I just had a quick look through them so far, and that already gives me priceless information that will help me pursuing my spec an calibration project.

To avvidclif, sure uploading this to any other test set is definitely NOT something to do - my purpose is only to use this to validate a few assumptions I made on what spec an cal data should contain, as I'm currently trying to figure out how to calibrate spec an module I fitted on my test set - which never had one before.

About your RIFA cap issue, but ~2 years ago, one RIFA cap blew in the PSU of my 8920A (The one in the snubber circuit on the main switcher primary)... what a stinky mess! That's probably pure coincidence, but after repairing the PSU, I discovered that receiver was dead. After investigating, I found it was due to a faulty amplifier IC on the reference board 1GHz path. Symptoms were differents than yours, but ref. assembly is probably the faulty one. As? jyelmgren proposes, using an external 10MHz is a good starting point. The point is that reference board is not too complicated and should be not too hard to repair... Schematic is available in the CLIP, and this is a good debug starting point. Good luck with your repair!


Jeff

Why??? Each set would have different data because of component variations. IOW you couldn't plug in someone else's? data and expect it to work correctly. Where it would be handy is to have "your" cal data to reload into "your" set in case of a problem with memory.
--
Clif Holland KA5IPF
www.avvid.com

On Sun, May 31, 2020 at 11:59 PM, Aretrac wrote:

RIFA capacitor

you really need to start a new thread on the repair.

Regardless two things ..
1. use an external reference and see what that clears up.

2. verify the power supply your self? ...the selftest checks this but if the internal? DMM is not working who knows what you have ...also you may want to look at the reference voltages that are brought out to the top of the measurement board.


Have not seen too many RIFA caps fail on the 8920's but some of that depends on how long the unit has been on

Hi All,?
? ? ? ? I am new to the group. I am trying to get my HP8920A working again, it blew a RIFA capacitor when I powered it up. I have replaced the RIFA caps and it is now failing the DMM, Counter and Spectrum analyser self-test. When I run the self-test manually it fails the real time clock and the DMM times-out when trying measure the reference voltages. While looking at it the other day I noticed that the internal 10Mhz is not working so further investigation is required.

If anyone has any pointers it would be appreciated.

Anyway the CPU still works and I have downloaded the Calibration. It is running A.12.04

Regards
Ade.

Hi All!

Was anybody able to try the program and get cal data out of their test set?
If so, would that be possible to share the data? That would be highly appreciated!
Thanks!

Jean-Fred

Hello,

This email message is a notification to let you know that the following files have been uploaded to the Files area of the [email protected] group.

• /HP-8920a and HP8920b related/8920 CLIP.pdf

Uploaded By: Clif Holland <clif@...>

Cheers,
The Groups.io Team

That direct link will not work - you have to search for "", which will get you to a working download location.