Fluke 8506A , need help with repair

[GigaJoe]

purchase for fun nearby ... (it had a cal seal on a screw case, and that important ... )

Voltmeter works, boards Ohm and AC do not detect when reset, and  do "error -9" when button option pressed  but installed, and cal. seal on screw ...


so I started from Ohm board ...
It appears that the board match with 5502A schematics, so ....  I'm wonder if OHM board from 5502A not compatible with 5506A ; can anyone clarify if possible ?

Below a digital part of board, If I correct, digital 1 on input IC1 IC2 IC3 (and IC0 , IC4 - ignored) command bus on left side generating ACK - "1" on board, by Q1
see: digit ohm parts.png

here below, after "reset" diagram; the ohm board wasn't installed, and device didnt reach  full board detection in my understanding , as ID 00011 didnt respond it stops, and when go to off routine detection, finally switching to voltmeter

"111" as well match with second column , consider C0, C4 ignored by ohm board, so ohm board would be never ever detected.
but -  fluke 5506a service manual page 214,  table 602a-8  defining the same situation : IC1,2,3 - HIGH

i really dont get it how it may detect ohm board ....
see: reset bus.png

[bdunham7]

It appears that the board match with 5502A schematics, so ....  I'm wonder if OHM board from 5502A not compatible with 5506A ; can anyone clarify if possible ?

I'm not 100% sure, but I can let you know in a few days.  I just happen to have examples of each torn to bits on my bench right now.  The 8506A is going back together and perhaps by the weekend I can get to the point where I can test that for you if nobody else has jumped in by then.

[precaud]

This is testing the memory cells, but IIRC the Ohms boards are the same for 8502 and 8506. AC boards are not.

[m k]

Boards have numbers, what are they?

Take it to the bare bones first, AC and ohms modules out.

First it should greet you with software version.
Then it shows the configuration.
Finally it starts measuring, blinking sample LED indicates that.

Next put AC module in, what it does?

Next replace AC module with Ohms module, use the same slot.
What it does?

[GigaJoe]

thank everyone,

Def. I did look at the wrong sequence,  here is correct one:
see attachment
3 ID (boards) responding .    3 ID expecting , no ACK signal

guess the reason, why AC cards are different for 8506 and 05 ,  The 06 unit looking for ID 01101 for thermal AC, but regular AC ID, 10101
I may assume the rest identical for AC board, so technically possible change ID by resoldering traces on the board.


so installed ohm board still do not respond on ID signal ....

[bdunham7]

So I had a few minutes to spend with my Frankenstein's-bench of meters.  The ohms module from an 8502A p/n 363968 is recognized by my 8506A with controller p/n 660563 and firmware 6.0.6.

I haven't yet had a compelling need to connect a logic analyzer to the data buses on these, but I am curious--how did you physically connect to the data bus and how are you triggering the LA?  Since I'm neck deep in two of these right now, I might be able to try and do the same thing and see if I can find the same sequence you are looking for.  If your AC converter is also not being recognized, have you tried removing the Ohms and AC modules one at a time?

[bdunham7]

guess the reason, why AC cards are different for 8506 and 05 ,  The 06 unit looking for ID 01101 for thermal AC, but regular AC ID, 10101
I may assume the rest identical for AC board, so technically possible change ID by resoldering traces on the board.

Wait, what?  Which AC module do you have?  Can you post a photo of your unit from the top showing the modules, the module p/ns and their location?

The 8506A uses the double-slot Thermal RMS converter only and its functions and ranges are completely different from the averaging and TRMS converters used in other 850xA meters.  I don't know what the 8506A firmware is supposed to do if it sees an ID from a non-thermal AC board, but I would assume it either ignores it or throws an error code or crashes. 

[GigaJoe]

I suspect that, 
It would be correct, as schematics of ohm  boards for 02 and 06 identical, but a bit different for 05 and 06.
but it doesn't matter, only ID on control bus are matter,  and it the same for all ohm board versions.

[bdunham7]

It would be correct, as schematics of ohm  boards for 02 and 06 identical, but a bit different for 05 and 06.

There are many different revisions for all or most of these modules, so what the schematics are will vary by which version of the manual you have.  Interchangeability is another issue that I'm not sure anyone has complete information on.  These meters are the test equipment version of low-volume Italian or British sports cars from the 70s and 80s.  

I believe that all the modules interchange between the 8505A and 8506A except the controllers and the AC boards.

[GigaJoe]

im wrong about AC module

my is double slot , and it not responding to Control ID , by ACK 

so ohm and AC dont recognizable , will deal with ohm board first ....

[m k]

I don't know what the 8506A firmware is supposed to do if it sees an ID from a non-thermal AC board, but I would assume it either ignores it or throws an error code or crashes.

I wouldn't be very surprised if 02AT TRMS modules are 06A acceptable.
06A schematics has a table for two variants, primary and alternate.
Software is also 502/603 and so on.

One more, primary set has 0 ohms for R18 and 0 pF for C19.
Obviously 02AT schematics is missing R18, but go figure, no value for C19.

What your device is reporting when all extra modules are off?
Mine is always 50x, but I don't have even a half of a TRMS module.
Must try that fake TRMS ACK thingy one day.

00A and 02A have different DC Sgn Conditioners and Ohms Converters compared to 05A and 06A.
For general exchange I can "guarantee" only R2 A/D and Active Filter, but I've not seen many different numbers.
I/O boards are most likely the same, but numbers are incomplete.
Isolator is probably also at least backwards compatible.
Since 06A is supposed to have TRMS modules those other versions are generally not necessary, but my educated guess is that they will work.
06A manual is also marking all front modules equally as add-ons.
What different 02A software versions do I have no idea but it's not far fetched that newer versions would support newer modules.

Those DC Sgn Cond and Ohms Conv differences then, and educated guess again.
Since new versions have those wider address selection gates, with inverted inputs, they just can't work if software tries to access them and some other using imperfect addressing.
But if those inverted inputs are grounded my guess is that they'll start working, software vice, measured values are a different thing, but when 05A DC Sgn Cond schematics is missing 7. extra info of R21-R23 it's only individual but equal resistors changed to resistor network.
(assuming here that my manual is missing one M for mega)

[bdunham7]

I wouldn't be very surprised if 02AT TRMS modules are 06A acceptable.
06A schematics has a table for two variants, primary and alternate.
Software is also 502/603 and so on.

Firmware space is at a premium here and the two TRMS converters have completely different operational methods, so why would they have code installed for something they explicitly say can't work?

[GigaJoe]

in my understanding

the controller determines to find a specific board by set a predefined ID in firmware, on IC bus, and then wait for ACK response from that board.
board ID that wasnt asking, will be ignored - became non active\working.

 if we are speaking about 2 variations of some board, and the same ID on control bus, it basically the same board from the controller's view. regardless of schematic

so ...  2 AC boards with diff. ID , only that will work who's ID are match with asking one

[bdunham7]

in my understanding

the controller determines to find a specific board by set a predefined ID in firmware, on IC bus, and then wait for ACK response from that board.
board ID that wasnt asking, will be ignored - became non active\working.

 if we are speaking about 2 variations of some board, and the same ID on control bus, it basically the same board from the controller's view. regardless of schematic

so ...  2 AC boards with diff. ID , only that will work who's ID are match with asking one

Something like that, but I'm not sure how that would detect illegal module installations, which the manual claims will set Error E.

9. Error E: Invalid module configuration:
A Calibration Memory module may be installed. The 8505A and 8506A do not use a
separate module for calibration memory entries (calibration memory is a standard
part of the Controller module). If a Calibration Memory module is installed, it must
be removed. This error may also identify an illegal ac converter configuration. For
the 8505A, only one ac converter (-01 Option or -09A Option) is allowed. For the
8506A, only the Thermal True-RMS Converter is allowed. Additional ac converters
must be removed.

[m k]

I wouldn't be very surprised if 02AT TRMS modules are 06A acceptable.
06A schematics has a table for two variants, primary and alternate.
Software is also 502/603 and so on.

Firmware space is at a premium here and the two TRMS converters have completely different operational methods, so why would they have code installed for something they explicitly say can't work?

You can read it differently and I'm wondering what makes the model.
If Thermal TRMS is mandatory for 8506A it obviously can't have other versions.
But if TRMS makes the model then without it the model is different and other variants can be used.

So what your machine says if you leave AC modules off?

The software also includes both greets and value of ram 0x400d decides what is outputted, after some tests.

8506A March 1983 manual has a motherboard named 8505A-4002 and front panel named 8505A-1601.
Front panel schematics top right corner has 8506A and bottom right 8505A-1001.

One ebay 8502A picture has after market sticker saying
OLD TYPE DC SIG
NEW TYPE OHMS

There DC Sig 383901 is regular 00/02 thing.
But Ohms 881842 is highest I've seen, 05/06 model is 881714.
Not very clear, overall.

I also did a module ID test by connecting IC0 to 3 with a diode, from unguarded side so ACK timing was most likely correct.
Idea was that Active Filter will respond when Thermal AC module's presence is tested.
Configuration was C 8, so no extra modules, but no errors either.
IC1 or 2 to 3 didn't give extra modules either, but A/D was disturbed and Error 5 appeared.
Error also appeared very late, Zero and VDC LEDS blinked, so testing was over and it started measuring.

[GigaJoe]

look at my message with attachment : IC BUS ACK.png

this is a sequence from the controller  (8506A) on control bus to detect all modules,  only 6 sequences (and it hardcoded) detected on analog side , so minimum 3 modules installed, as bare minimum, the rest 3 non exist. so it working as Voltmeter.
 

[m k]

What texts it is showing before it starts metering?

First text is normally a software version.
Mine is HI-5.0.6 and bdunham7's #5 is HI-6.0.6, but our software versions are different, sets are 505/606 and 506/607.

Next text is module configuration.
C is A/D and 8 is Isolator, that is minimum text with Isolator, other mandatories are not reported.
If you add GPIB the text changes to C5 8.
It's possible that some early version is also showing D and F, but I can't remember seeing any.

For your IC BUS ACK picture, there must be more.
First 3 are as they should, but 4th is AC/DC Avg (add on 1) and because it's not present the software is waiting a timeout.
Next 2 are ok again, Ohms (2) and Current Shunts (3), but interesting parts are not there, what it does after (3), and after AC RMS (9).
External I/O module tests are also missing and they obviously must be somewhere, despite the meter model.
If testing really stops after Current Shunts it may have something to do with Cal Memory, it shouldn't be present with newer controller.

Three last slots are identical, so are four first.
After removing any external I/O module your meter should have only one module after Isolator, a single module controller.
8506A manual is putting DC Sgn Cond to first slot but that is not mandatory, in theory you could expand the front and get Currents and Ohms present simultaneously.
Again and in theory, you could expand one slot to the back and get room for external I/O after double module Controller and Cal Memory.

(IC values in software are inverted)
506/607 software has at least 2 sets of module lists.
0x0746 is a list of modules and their configuration letters, DFC123-56789A.
(D DC Sgn Cond)
(F Active Filter)
(C Fast RR A/D)
(4 Cal Memory)
(A Therm TRMS AC)
0x3a04 is add ons, 2nd part has Isolator and Thermal AC only, there Cal Memory is still present, IC 100 0110.
So the list is 12345679 8A in inverted IC values, including also MSB of course.

[GigaJoe]

A small step forward for OHM card,

after repair logic now it does recognize by unit, I have option -2 and it of course it not working
switching to "OHM" , input shorted, any manual range - result "HHHHHHHH"  or "error H" on display when switch to "OHM"

current state :
the board has gas discharge tube - it has some signs of discharge inside,   transistors Q6 , Q7 removed as suspicious
maybe someone experienced same\similar , some clue ?

schematic attached

PS:  im suspicious on Q9 (?) input diff dual N transistor, if it fail ?  Vdiff between PIN 2,3 would it be large ?

[bdunham7]

after repair logic now it does recognize by unit, I have option -2 and it of course it not working
switching to "OHM" , input shorted, any manual range - result "HHHHHHHH"  or "error H" on display when switch to "OHM"

current state :
the board has gas discharge tube - it has some signs of discharge inside,   transistors Q6 , Q7 removed as suspicious
maybe someone experienced same\similar , some clue ?

Try strapping the sense/source terminals together so that you have a 4-wire short.  You can just do it with regular leads, no need for precision.  If that doesn't help, you might pull the input switch assembly out and see if there are any blown fuses on it.

[GigaJoe]

did try with 4 terminals shorted - same "HHHH"
I'm really suspicious it the board itself;  as voltmeter's part of unit are working

[bdunham7]

OK, but the voltmeter will work with blown fuses.  I haven't worked out a complete table of what symptoms might accompany each blown fuse, but on my 8505A which appeared to have been overvoltaged there were blown fuses on the input switch board and IIRC the ohms also showed the Error H. 

Also, what voltage are you seeing at TP4 relative to reference ground (TP3)?

[Kleinstein]

Q6/Q7 are there for protection. They are candidates for failure, but one should be careful with just removing them, as this makes the instrument sensitive to damage, e.g. from ESD. Q6/Q7 should than be replaced with new transistors - rather normal ones (no special RF types) should be OK.

A failure of the double JFET would be one of the more difficult parts, but one can get around  with 2 selected single JFETs for a start. OPeration without Q6/Q7 may lead to such a failure.
Chances are a broken FET would cause the U5 output to go to one if it's rails and TP4 no longer be a good virtual ground. There are still other errors that can cause this problem.  Q9/U5 would be suspect of TP4 is positive and Pin1 of U5 also positive  or both negative.  If they have opposite signs, chances are the problem is somewhere else.

[m k]

did try with 4 terminals shorted - same "HHHH"
I'm really suspicious it the board itself;  as voltmeter's part of unit are working

Can you hear relays?

With Ohms Conv the signal route must change to RT1.
That includes also some DC Sgn Cond relay connections.

[GigaJoe]

HURRAY !
Ohm board working !!
as everything was correct (in my assumption )  opamp output (+10V) , relay clicking,  voltage TP4 (+20mV) ,  +18V on current output , and finally ...
R12 - 100K resistor in open state, no any visible damage, but infinite resistance .... so i get it  "HHHH" basically open resistance measure ; 
so short input  gives 1.5 ohm, and it seems show all decades using  10M decade resistors box  ( don't have 100M)
guard transistors actually faulty

I thank everyone for help and mental support (!).  but thing not over yet, due to:

next chapter:  true RMS AC board , not detected at all,  I will start diggin soon


PS:  "blown fuses"  - is my understanding correct that board on the side that holding switches, front\back, 4W ohms, also has fuses ?

[bdunham7]

PS:  "blown fuses"  - is my understanding correct that board on the side that holding switches, front\back, 4W ohms, also has fuses ?

Congratulations!  Yes, the board has 4 HRC 3A fuses soldered to it in addition to the ones in holders at the front and rear of the unit.  The board is easier to remove than it might appear at first:

Pop the extended buttons for the switches off by pulling with one hand and carefully prying on the visible groove with a medium-smallish flat blade screwdriver.  A sharp/new on helps.

Remove the TRMS converter connector from the top.

Remove the three screws holding it from the outside of the unit.  Remove all the screws holding the small back panel on and pop it loose.

Grab the board with a needlenose pliers somewhere near the TRMS connector and pull it upwards out of the edge connector on the motherboard.

It should be loose if I haven't forgotten anything and you can now carefully work it out the back.

Edit:  I did forget something--there's a tab at the top front of the circuit board that protrudes out of the inner case just enough so that you can push it upwards with a screwdriver and pop the board out of the front edge connector.