Question on Tektronix 496P spectrum analyzer

[[email protected]]

Hello everyone. This is my first time posting to this group. I have
a 496P that I picked up on eBay and of course, it is not functioning.
I have the four manuals for it, however the two service manuals do not
match and both do not match the actual unit. This unit appears to be a
newer design as the layout appears cleaner, but serial number is low
(B020866).

Overall the unit appears in very good condition for the age. Hour
meter is at about 2000 now. Most of the features appear to be working
and the readings I get from it are very close. I ran the internal
computer tests for fun and this passes without any problems.

It would appear that the outside PLL loop of the 1st LO does not
function correctly, maybe. The sreen displays Error 58 anytime it
tries to lock. Looking at the schematic in the service manual I have
and tracing out the system, it does not appear that the FM coils
negative side is connected. All the hardware is in place to drive the
positive side of the coil. I assume that because both the service
manual and the unit are the same that all of the tuning was with the
main coil.

All of the frequencies appear correct in the innner loop. Looking at
the 25MHz and the 5MHz strobe signal, they apppear to be locked and the
IF on Q1040 of the offset mixer is at 50% and the frequency does appear
to match with what I see on jumpers K and N.

Removing the 5MHz Stobe enable jumper (P3057) and setting the PLLs loop
gain, I can make the adjustment to remove what I and the manual calls
the beat note from (TP2038, on my card but called TP2037 in the service
manual) to allow the Lock signal to set true. Of course there are no
waveforms showing what the beat note would look like, so I am not 100%
sure this is what I am seeing. I traced the Lock all the way to the
bus and indeed, when the processor reads bus (U4025 74240 pin 1/19
low), pin 9 (lock on instrument bus side) goes low. Up and down
also are low on pins 15 and 17.

So, it is not just Lock that the computer needs to see to make it happy
and I know it is not locked by the drift.

I suspect that the problem may be in the phase detector. What I am
seeing on its output is a 40mV p-p signal that corrisponds to when the
Strobe is active, riding on a -4.0868. The DC voltage depends on the
inputs to the phase detector and if I set the system to Max, I get a
signal that would appear to represent the error. What I am not sure
about is the 40mV. With the span set to 50KHz (or even 1MHz for that
matter) there does not appear to be enough signal to make the F ERROR
output change states. I can set the Error Count Breakpoint pot (R1061)
to get something out with a 5MHz span, but no lower. Because of this,
I never see the Valid Count (pin 8 of U4025) get set.

What really bothers me is that between the the two manuals and the
actual unit I have, Tektronix made a lot of changes to the PLL circuit.
I am guessing that this was not done for the fun of it and that they
had problems.

If you have any ideas, shy of telling me to get a different one, I
would like to hear them. Thanks in advance.

 

[mike]

Hello everyone. This is my first time posting to this group. I have
a 496P that I picked up on eBay and of course, it is not functioning.
I have the four manuals for it, however the two service manuals do not
match and both do not match the actual unit. This unit appears to be a
newer design as the layout appears cleaner, but serial number is low
(B020866).

Overall the unit appears in very good condition for the age. Hour
meter is at about 2000 now. Most of the features appear to be working
and the readings I get from it are very close. I ran the internal
computer tests for fun and this passes without any problems.

It would appear that the outside PLL loop of the 1st LO does not
function correctly, maybe. The sreen displays Error 58 anytime it
tries to lock. Looking at the schematic in the service manual I have
and tracing out the system, it does not appear that the FM coils
negative side is connected. All the hardware is in place to drive the
positive side of the coil. I assume that because both the service
manual and the unit are the same that all of the tuning was with the
main coil.

All of the frequencies appear correct in the innner loop. Looking at
the 25MHz and the 5MHz strobe signal, they apppear to be locked and the
IF on Q1040 of the offset mixer is at 50% and the frequency does appear
to match with what I see on jumpers K and N.

Removing the 5MHz Stobe enable jumper (P3057) and setting the PLLs loop
gain, I can make the adjustment to remove what I and the manual calls
the beat note from (TP2038, on my card but called TP2037 in the service
manual) to allow the Lock signal to set true. Of course there are no
waveforms showing what the beat note would look like, so I am not 100%
sure this is what I am seeing. I traced the Lock all the way to the
bus and indeed, when the processor reads bus (U4025 74240 pin 1/19
low), pin 9 (lock on instrument bus side) goes low. Up and down
also are low on pins 15 and 17.

So, it is not just Lock that the computer needs to see to make it happy
and I know it is not locked by the drift.

I suspect that the problem may be in the phase detector. What I am
seeing on its output is a 40mV p-p signal that corrisponds to when the
Strobe is active, riding on a -4.0868. The DC voltage depends on the
inputs to the phase detector and if I set the system to Max, I get a
signal that would appear to represent the error. What I am not sure
about is the 40mV. With the span set to 50KHz (or even 1MHz for that
matter) there does not appear to be enough signal to make the F ERROR
output change states. I can set the Error Count Breakpoint pot (R1061)
to get something out with a 5MHz span, but no lower. Because of this,
I never see the Valid Count (pin 8 of U4025) get set.

What really bothers me is that between the the two manuals and the
actual unit I have, Tektronix made a lot of changes to the PLL circuit.
I am guessing that this was not done for the fun of it and that they
had problems.

If you have any ideas, shy of telling me to get a different one, I
would like to hear them. Thanks in advance.

First suggestion is to go ask the question at:
http://groups.yahoo.com/group/TekScopes

Been a long time since I worked on one of these.
Built several from dumpster parts many years ago.
My job was fixing the bad boards.
Troubleshooting is a lot easier with a set of extender boards.
The things are packed in so tight you can't get a probe on 'em.

Tek made a bunch of improvements in the 49x series over the years.
They changed multiple boards at once. You have to have every board
matching the firmware version. It's not easy to tell what you need.
If you don't know the history, you may not have a complete/compatible
set of boards/firmware. Many of the changes were relatively simple,
but you'd need schematics for both to do the mods.


IIRC, the upgrade that provides the high accuracy timebase and
multi-digit frequency readout made things much more difficult to
troubleshoot, cause the processor complains about every little thing.
That's why I never updated mine.
mike

--
Wanted, Serial cable for Dell Axim X5 PDA.
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/