Philips PM2534 - auction won - hard repair

[pixel63]

Hi all,
I had a PM2534 declared 'not working' for 42 euros and I'm trying to bring it back on.
The references are: TYPE: PM 25334/01, NC 9447 025 340 | 1, DY 0101663, 230 Vac 50Hz .
The ICs are marked '86 ! The board is perfect.
I found complete Operating Manual and Service Manual online.

The DMM is stuck on booting with the message '0000 nc', no button works on the panel.
The display has only this symbol on:
       tlk only
 + ^0000 nc V =
       speed 2

Inside, I found the litium battery with the negative link interrupted and some oxidation near the front panel circuit, over the RAM IC, so I was sure to have hit the center. The battery is at 3.2 V but RAM must be fatally empty, lost the calibration. FAIL.

The service manual states that if you have '0000 nc' or you have replaced the RAM, you must do: Hold CAL pressed, press and release RESET, then press CHECK button, then the S1701 button on PCB and again RESET.
I made it but had no response.
The RESET button seems to work (you can listen a beep on the button release) but not the factory reset sequence.

GT

[pixel63]

I made some checks:

• all voltage rails on the testpoints on PCB: +6V, -6V, +8V, -8V, +5V are all OK but for -6V and +8V that are slighty higer
• the biggest axial capacitors have the right ESR
• pressing RESET, CAL and S1701 you can see the right voltages applied to the relative locations and IC pins
• tried a new 6116 RAM IC and nothing changed
• redoed the soldering joints on the oxidized ones near the RAM chip

Nothing changed.
GT

[pixel63]

OK. I'm stuck.
Voltage Rails are here:  +6V is +6.42V, -6V is -5.99, -8V is -8.69V and +8V is +7.97V, +5V is +5.00V.
Also, you can ear a high pitch sound that seem to come from the big inductor on the analog part of the circuit, it goes away only when you press the reset pin.

I'm trying to understand the origin of the sound and why the DMM does not seems to reset as due.
GT

[Shock]

Probably just made your day, see 4.3.2 IEC IEEE messages, the manual says it has a remote talk only and local lockout, verify that first. There is instructions you need to read just below that section on returning it to local.

[pixel63]

Shock, thanks for the help.
I saw that instructions but my 2534 seems not to respond.

At power on the display is as in the photo1 below.

If I press and maintain pressed CAL, then press RESET, then release RESET, the CAL message lights up on the display, if now I press CHECK nothing appens, the same pressing S1701 on the PCB, then I press RESET and it return as after power on.

If I press CHECK after power on, nothing appens, I cannot see the 'rtl 0/1' or the 't.only 0/1' function, the display is stuck as in photo1.

I'm asking myself 'how it can maintain the talk only mode if the RAM is blank?'.
The 'nc' message doesn't mean that the RAM is erased?

The sequence to force the reset, stated at 'PART B, CALIBRATION ADJUSTMENT' (see photo2), seems not to work at all.

Bye
GT

[Shock]

To me it sounds like the keypad is still working if you can get it to reset. You might have to clean the check button or verify it's doing something functional. It expects a number to be pressed after that. Read 4.4.2.5.

So I think before you try doing the calibration reset disable the "tlk only" mode or whatever else it takes to get full control. Do this from a depowered and unplugged state instead of after the reset and see if that makes a difference.

Can you upload or link what manual your using and the section you are following, it makes it easier.




 

[pixel63]

Shock,
I made some measure on the keyboard to verify the functionality of the keys.

Each of the key is functional at the continuity check and the CHECK key press is visible at the relevant pins on the D2201 IC (PCF8574).
Really, only CAL and RESET button seems working well, but they are directly connected from the voltage rails to the PCB connector.
I want to check now  if the keypress is correctly reported ton the main PCB or eventually the D2201 is disabled or not funtional for some reason.

At the moment doesn't seems to be any way to exit from this locked state, the DMM is unplugged (and I don't have an IEEE equipment) and repowered any time. The reset brings you on the state you see in the photo1 of the early post.

I'm also wondering what is that noise that the device emits, it stops only when you press the RESET button. It could be possible that some fault is preventing the DMM to start correctly and let it stuck at this point. It is like a whistle. There is also an OVERLOAD indicator on (the arrow up on the left of the numbers) to investigate (see photo1).

The manuals I found are: 4822 872 30333 - 860117 (Op man), 4822 872 35313 - 900205 (Svc man), 4822 872 30389 - 880501 (RS232 intf)

GT

[Shock]

I've not fixed this specific problem before, so there may a simple fix out there which is why I would try to exhaust my options trying to get the keypad/display responsive.

But I'd next try the the IEEE interface, test voltages/ripple and any test points or waveforms from the SM, then caps and ESR. Plus go over the whole PCB section by section for joints, heat problems and reseating etc.

I'd go over your voltage and corrosion problem to see what it could be impacting and test surrounding components and traces.

After that take several reads of the ROM to check if it's ok, get a checksum off a known one if possible. Look for an illusive test ROM or an updated ROM. End up blowing away the ram/calibration config in case it's corrupted somehow.

That would leave troubleshooting logic.

[pixel63]

Shock,
Today I checked the keboard and I learned that the IC is an I2C expander, the same bus is used for the display, so I measured the SCL, SCA and INTK- signals at keypresses.
INTK- goes down when you press a key, OK.
SCL and SCA are too complex for my 20MHz analog oscilloscope (9012 Beckman Industrial, my first original scope), but I was able to see what probably is the display I2C traffic (see photo3). When I press a key the pattern slides slightly so I think the keyboard seems ok.

I don't have here a IEEE controller or adapter to view/test the interface and I don't have a programmer for verifyng the eprom and the RAM.

So I will do all the sistematic checks you suggested.
I'll post them.
GT

[pixel63]

I made a complete visual check on the soldering side of the PCB and redoed teo or three joints. No news.
Then I made a complete visual check on the component side of the PCB, no news, still a wonderfull make from Philips, old stile all trough hole, fantastic.

Then I went on the 'whistle', i tried each of the reed relais with the DMM on and .... I found only one in ON state but I found one (K1303) in continuos oscillating switching. That is the origin of the sound!

Now I have someting to investigate on... back to the schematics ....

GT

[senso]

At what Voltage per division was that scoped?
Seems a bit on the low side of the voltage.

[pixel63]

That photo was made with my poor scope tuned to sync something stable and wasn't simple. I used the settings that gave me a stable image so I don't know.

I will check the levels on the I2C buses.

From the SM it seems that the K1303 is part of the RMS Converter and should be close only during the AC voltage measure.
It is driven directly from one of the output of the ADC Control (OC) so it is strange.

The display is in V= so the K1303 should be open stable, but it's also showing the OVERLOAD sign (CF or CLIP?).

Investigate ....

GT