MotoMeter cluster MFA repair [Fix found]

Discussion in 'Electrical' started by NateS2, May 3, 2019.

  1. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    As most people on here are probably aware, the MotoMeter MFA Gauges fitted to many MK2's are quite unreliable and so far the only true fix seems to be "swap for VDO". Hopefully, this thread might change that.

    DISCLAIMER: VERY electronics heavy, be warned :o

    The issue with my set of gauges, and most people its seems, is the rev counter, MFA, and certain warning lights don't work however the clock does work. No amount of playing about with fuses, removing the battery to reset the gauges or swapping the 10 volt regulator can solve it, meaning it is deeper within the gauges.

    The first step I've taken was ruling out the blue flexible PCB. This was done using a continuity tester to determine what pins on the CE2 connector correlate to which pins where the flex PCB connects to the gauges. Once this was done, I removed the flex PCB, took the gauges apart, and powered up just the rev counter section using a 12 volt supply connected to the pins found earlier:
    [​IMG]
    As you can see, it worked! Actually this was bad news as it means the fault is on the main PCB...

    So now the gauges are apart we can see the circuit board with all it chips, I've identified the key ones of the picture below:
    [​IMG]
    The NEC processor is what controls the gauge's. Its a fully integrated chip with in built-in RAM and ROM, LCD driver, AD converter and serial interface (Datasheet). As the clock is working, it suggests that the chip is working to some capacity as it's actually keeping accurate time and driving the LCD with it. This also means that chips power supply must be okay, which can be confirmed by measuring 5 volts at the regulator and at the test points marked in yellow.
    As the main power supply seems to be okay, the next step was checking that the chip is actually receiving the "run" single when the ignition is switched on, as if this was not present it would explain the symptoms. After tracing the tracks from the main connector I found the test point shown in yellow below:
    [​IMG]
    There is around 4.8 volts present at this point when the switched live wire is connected to 12 volts, and 0 volts present when not connected. Tracing further, the 4.8 volts eventually connects to pin P41 on the MPU. This 4.8 volts is within the HIGH logic threshold of the chip, meaning that is probably isn't the issue either.

    At this point, I'm at a bit of a loss. There is a relay, show in red on the first image, that doesn't seem to do anything, I'm going to try and trace it out but the tracks are very dense in that area so its a challenge. If anyone has any idea what it does feel free to let me know!

    My next step is going to be attempting to read the EEPROM, as this will contain all the configuration data for the gauges. Looking at the datasheet, this chip is only rated to retain data for 10 years!! This chip was programmed 30 years ago... So there is a very good chance that the configuration has simply been lost to time, and I'm assuming that if the MPU doesn't get any data it just doesn't enter its "run" mode and stays being a clock. If this turns out to be the case, I'll have to try and source the original configuration data, either from Bosch (owned MotoMeter at the time) or from a working set of clocks.
     
    Last edited: Apr 29, 2020
  2. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    Interesting reading :)

    Did you attempt to cycle through the MFA modes once the ignition feed was supplied? As the clock itself wont change ignition on or off until you try to change the MFA mode
     
  3. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    oh and the 'relay' is the oil pressure warning buzzer :)
     
    NateS2 likes this.
  4. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Cheers John, that makes sense, I was actually wondering where the buzzer was :)

    And yeah I have tried putting +12 volts and GND across all the pins associated with the MFA stalk, non of which generates a reaction from the cluster. When MFA's working, I think the LCD should display one of the little triangles that would point to the current mode, so that's how I'm testing it at the minute.

    I've connected a Logic Analyser to the Microwire (Similar to SPI) interface of the EEPROM chip so I can monitor the communication between the MPU and EEPROM:
    [​IMG]
    When I power the clock cluster up, connecting the constant and switched 12 volt lines simultaneously, the MPU send a command to read data from the EEPROM, and the EEPROM actually returns something as shown below:
    [​IMG]
    This shows that at least the EEPROM still works, however the data its returning may be corrupt, hard to test this without a working cluster. In total, the MPU reads 18, 16-bit words from the EEPROM during this boot stage. It starts reading 4 words in quick succession, then reads the remaining 14 in blocks of 2 with a pause between each block.

    Looking at the data, it all seems quite logical with several patterns. The next step will be using an external programmer to read the entire contents of the chip. This will let me see if there is actually only 18 words used, or if the MPU stops reading after 18 because it fails a checksum due to corrupted data.
     
    Last edited: May 3, 2019
    Antenor Morini Junior likes this.
  5. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    very interesting stuff. I actually have a selection of VDO and Motormeter 16v/8v boards sat about, but the strange thing is all of my 'motormeter' 8v clusters have the same boards as the VDO 16v units and look nothing like your board!

    back in the day you could have your motormeter clocks sent off to be fixed, though the fix made wasn't made clear. maybe the 'fix' was to take the motormeter board out, throw it in the bin, then fit a VDO board :lol:
     
    Antenor Morini Junior likes this.
  6. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Quite possibly yes :lol:
    The issue itself is very odd as none of the data being read from the chip seems to match what I'd expect (MPG lookup tables, speedo drive ratio etc.), so it is possible that could have something to do with it.
    However, I actually managed to get MFA working for a few seconds by quickly power cycling the dash several times. But then trying this again yields results ranging from no display at all, to the message "CErr" which, as far as I was aware, would only be shown on a Digifiz secret menu. [8-}]

    I'm also sure now that the signal from the ignition is working correctly as the current consumption doubles when this is connected, suggesting that the dash is actually entering a "run" mode.

    Do you have the contacts of anyone who used to fix them?
     
    Antenor Morini Junior likes this.
  7. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    Yeah I recon your board data is corrupt, I've seen it mentioned for MK3 clusters a couple times and solution is to just replace them.

    Another note with those the 4 circular pads on the rear of the board which are accessible when the clocks are fully assembled, I have a theory these are data ports to read and possibly write to the eeprom in the same way as the test port on the rear of the MK3 Golf clocks which you can use to adjust the mileage. But I've never got round to looking into it any further

    You find a few adverts on motormeter repair in the old rabit magazines plus some peopel have mentioned in the early days of the forum, but not sure if anyone is still actively repairing them?
     
    Antenor Morini Junior likes this.
  8. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Yeah I need to take a look at those, I can program the chip directly if needed but whether those pins offer additional functionality is something I'll have to take a look at.

    Unfortunately, I'd need a working cluster to compare the data too, or a dump of the chip from a working board, something which doesn't seem to exist.

    I emailed AwesomeGti about it, and it used to be a service they offered but the guy who did it has long since left, so they can't provide any information.
     
    Antenor Morini Junior likes this.
  9. Simon Peter Dodgson Forum Member

    Joined:
    Nov 13, 2018
    Likes Received:
    27
    Bugger I've just refitted my motometer clocks after refurbishment. As an electronics engineer myself I am watching this with interest. My analyser was dumped as no need for it anymore
     
  10. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Ahh shame, no worries
    What was the issue with your clocks Simon, was it the same as mine? And who refubished them?

    Update on trying to get schematics:
    MotoMeter - Were owned by Bosch at the time, so they don't actually have access to the documents
    VW - No Reply
    Bosch - Cannot provide information from head office so I need to contact a local service centre, So i'm feeling like this will be a dead end too
     
    Antenor Morini Junior likes this.
  11. Simon Peter Dodgson Forum Member

    Joined:
    Nov 13, 2018
    Likes Received:
    27
    Boring things like no temp gauge,speedo needle not put back on to the calibration line so reading high after needles painted red snapped mfa vacuum spigot and half illumination missing.
     
  12. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    random question, were these 120 or 140mph? as all of mine were 140mph, using the VDO type main board. As you can see, pretty different to the 'proper' motormeter board you have!
    IMG_20160828_110852.jpg

    on a side note, the blue foil and rest of the components are compatible between the 2 types, and indeed in ETKA there is no VDO or Motormeter specific part numbers listed for the main board or any of the sub-components unlike with CE1 where the 2 makes are very different and many of the parts are specific to either motormeter or VDO.

    I actually have a NOS blue foil which has motormeter printed on it, and if I ever used it would bother my VDO OCD pretty badly :lol:
     
    Last edited: May 9, 2019
    Antenor Morini Junior likes this.
  13. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Oh okay nothing too major then, luckily mine hadn't been tampered with, until I started [:D]

    Mine are 140 mph John. Looking at the VDO PCB, it looks far more simple and robust than the MotoMeter one, possibly explaining why they still work. There's far fewer passives, the voltage regulator is much bigger, and it's actually got what's looks like a solid state polymer capacitor towards the bottom left, which should last far longer and handle high temperature better. It also looks like they're using a processor with an integrated EEPROM (unless its on the other side of the PCB), which I think would have been quite novel for 80's so its possibly custom?

    I'm not sure what the chip in the bottom left does, as I can't see the part code in the picture, but I'd guess the one in the top left is the motor driver for the RPM gauge.

    That's good to know, if the worst come I'll just have to see if I can get a VDO PCB to fit to mine, and just try and ignore the OCD :lol:

    Note for John: When I click the picture I get a page with, even when I'm logged in:
    Club GTI - Error
    You do not have permission to view this page or perform this action.
     
    Antenor Morini Junior likes this.
  14. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    Bottom left is LM2902N, comes back as a 4 chanel amp? Top left is kefa 635 936510 and is the rev counter controller as you say.

    On the flip side theres a chip "924 lp29 51cm" which from googling is a voltage reg? and 93cs56em14 which I think is the eeprom?
    IMG_20160828_110956.jpg
     
    Antenor Morini Junior likes this.
  15. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    If you want an 8v board I think I have more of these than I have 'good' parts to assemble complete clusters so probably have a spare for you somewhere. They are all from motormeter clusters and dont say VDO on them anywhere that I can see so your OCD can remain intact :lol:
     
    Antenor Morini Junior likes this.
  16. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Yeah "93cs56em14" is a 2 KBit EEPROM, the "LM2902N" could be used to amplify or buffer some signal coming from the car, probably the RPM pulses or pressure sensor output.
    As for "924 lp29 51cm" this could either be the main Voltage regulator, although I think that is the TO220 package on the back of the board, or it could be the power supply for the LCD.

    And yeah that would be much appreciated, thank you![:D]
     
    Antenor Morini Junior likes this.
  17. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Found some more time to work on this project, finally having some success reading the EEPROM. I'm yet to look into what the programming pins do but by connecting a programmer directly to the chip I managed to read the following data:
    MotoMeter dump.png
    Assuming this data is correct (more on that later), it would seem to confirm that the chip is corrupt. It looks totally random, with no real patterns and no data that I'd expect to see, such as values displayed in the "secret" menu of the MFA.
    There is also a chance that there some issue with the software that I'm using to read the data, possibly that the Microwire protocol, being very old, has not quite been implemented correctly meaning that it works, but gives incorrect data. This is somewhat confirmed by comparing this output with the Logic analyser and seeing that they seem to be totally different even though they should match as they are reading the same data from the same chip. I need to do some more testing and possibly implement my own Microwire reader to confirm if this data is correct.
     
    Antenor Morini Junior likes this.
  18. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    Apologies for the delay, life and other car issues got in the way! Since last time, I'm now the proud owner of 3 more sets of clocks.

    VDO non MFA set (fully working)
    MotoMeter smashed lens and speedo (worked for a short time then stopped)
    MotoMeter broken blue flex PCB.
    All laid out it amounts to quite the collection...
    Broken clocks.jpeg
    The first set of MotoMeters that I bought with the cracked lens initially worked perfect, but after being plugged and unplugged a few times for testing they went intermittent and then totally stopped working.
    The second set of MotoMeters I bought had broken blue PCB, so I swapped one of the working PCB's onto the back of them. The MFA and rev counter both work on this set however the speed pickup isn't working. (Need to try the other flex PCB)

    As the first set broke while they were being handled, I'm going to try re-flowing the board to see if it is dry joints causing the issue. This would ideal as its a simple fix however I'm not very optimistic and flexing the board doesn't seem to have an affect.
    If this doesn't work, I'm going to try and compare EEPROM data from my working set, with two broken ones, and possibly start swapping chips to see if I can isolate the issue.
     
  19. NateS2

    NateS2 Paid Member Paid Member

    Joined:
    May 1, 2019
    Likes Received:
    257
    MFA WORKING.jpg
    FIXED! :)) A fully functional MFA!

    Huge credits to this forum post from 2012 and the German Forum Motor-Talk.

    The solution:

    There is a transistor on the board that controls part of the power supply, and this transistor can burn out, causing the clocks to fail.
    As a quick test, I removed the transistor and soldered on a generic through hole replacement and the clocks are now fully functional!

    I've just ordered the correct part so that should be here in a few days, and the in the meantime I'm going to continue to look for the original repair manual as this is apparently where this fix came from.

    If anyone has any broken or intermittent clocks drop me a PM and I'd be happy to take a look at them for you!
     
    Last edited: May 16, 2020
  20. rubjonny

    rubjonny Administrator Staff Member Admin

    Joined:
    Oct 31, 2003
    Likes Received:
    3,298
    Location:
    Bracknell
    Excellent work :thumbup:

    How is it all your motormeter clocks came with motormeter boards in but non of mine have :lol:

    I actually found a bit of time to swap my partially broken 16v clocks out for a set of fully working ones the other day and dug out a spare VDO 8v board for you, but I guess you have all the spares you need now ;)
     
    Antenor Morini Junior likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice