HP 8902A Troubleshooting - AM Modulation Problem

[Hawaii596]

I have a problem with my HP 8902A in case any of you have seen this problem, might save some truobleshooting time....

THE SETUP:  Testing an HP 8657A at 100 MHz, 0 dBm.  I can do Freq, Tuned RF and FM with no problem.  If I do an INSTR PRESET with 100 MHz 0 dBm CW (no mod) in, Freq reads fine, RF Power reads fine, Tuned RF reads fine.  I do the CALIBRATE at 0 dBm, run down to -40 and -80 and cal does fine there, reads fine and stable all the way down to -130 dBm.  Then if I turn on FM 1 KHz mod freq, any Pk deviation measures fine (I tested 5 to 50 KHz).  All measured fine.  Then I turn off mod, go into AM mesaurement and turn on AM (int 1 KHz) and mod level (I tried f2%, 5%, 10%, 30%, 50%, 70%, 90%).  I just get the two dashes, no reading.  If I press MHZ key to lock the frequency on the 8902A, the AM tries to read, I get four or six dashes, and hear clicks while it tries to find the AM, but it never comes in.  I tried at various frequencies (1, 3, 10, 100, 500, 1000 MHz), and same symptoms.

We have a spare 8902A in the storage area (someone had begun retrofitting rear to front input on a working unit, and partially disassembled what we believe is a working unit).  So I swapped out the A6 AM DEMOD board, and no change in symptoms.

After extensive tinkering last night, I found that if I drop carrier level down to about -10 dBm, the AM starts working.  I started incrementing power up and found it stops working at about -7 dBm.  I did this at 10, 100 and 1000 MHz and about the same amplitude worked at each frequency.

Also, when I connect the sensor module (11722A) to the AM/FM CAL output, I get the normal ~100% (reading 100.10%).

Anyone seen this problem?  If so, does it sound like an alignment problem or a failure?

Thanks for any inputs anyone could give.

[Broken_Wings]

In my experience most of the time with the 8902A its not adjustments. Its broken somewhere. Many areas overlap.

Not familiar with what you are testing though.

A test I do on some 8902's if you have opt 002 is pop the top.
Remove the input to A16. Put the rear time base in tell it 10 MHz. and if you can feel the programmable attenuator repeatedly clicking that is the problem. It can't put the pad in it wants for the level.

If the problem you see on an 8902A occurs mostly at lower frequencies try replacing the two RF switches (if not replaced by an IC switch).

But if you aren't specifically testing your 8902A go to the setup with the clicks pop the top and place a finger or two solidly on the programmable attenuator At1 or At2 its the one not at the front panel.

If you feel those attenuators being toggled either the level of your source is unstable causing the 8902A to be unable to keep it at the desired level in order for the frequency dividers to work and track it or those attenuators or the 10 dB on the front could be bad.

A spec an set to 1dB/div may help you visually see if your TI is being flaky in AM.

There is also a special function to turn on the error readout. Its either in the 40s or 70s but troubleshooting mentions it.

[Hawaii596]

Don't know how much this means....

But I tried going in with straight CW (100 MHz, 0 dBm).  I tried manually setting the attenuator (1.X codes) and manually setting the input IF gain (9.X codes).  I was sometimes able to "trick" it into reading %AM at 0 dBm, but not always.  This is making me lean toward possibly the attenuators.  I'm fortunate in having a spare to board swap with.  But those very tight T10 torx's are giving me carpal tunnel.

I'd really appreciate any help on this.  I've been board swapping all day - so far A2 3, 4, 5, 6, A54, A55, A16.  I'll try the big attenuator (next to A16 next).  I'll be watching this post while I work.

The flate rate repair at Agilent is $3802.  Way too much.

[Hawaii596]

BINGO!!!  I just swapped out the 10 dB/step attenuator.  That seems to be it.  Thanks for the input.