Gossen Metratop 52 repair - schematics needed

[giosif]

Hi all,

Relatively recently, I managed to score the unicorn of unicorns: a Metratop 52 bench meter from Gossen.
As expected, this is a non-working unit and I'm trying to get it back to functional state.

Does anyone here happen to have the service manual and/or the schematics for this meter, please?
I am suspecting some traces being broken / not making proper contact and schematics would be most helpful here.

Thanks!

[coromonadalix]

Gossen to my knowledge never relased an service manual for any model they have,  even under an non disclosure agreement 

Lucky you,  i wanted to have this meter model


Send a mail at Gossen, and since its a "prehistoric" model,  i dont think the answer will be positive,  never did for my metrahit 28 and 29

good luck

[coromonadalix]

oh

You have this thread here  for some photos of a model 51,  no schematics tough

https://www.eevblog.com/forum/repair/teardown-and-repair-of-a-gossen-metrawatt-metratop-51-(model-m2510)-benchtop-dmm/

[retiredcaps]

As expected, this is a non-working unit and I'm trying to get it back to functional state.

As per usual, what exactly is not working?  Be specific.

[giosif]

Gossen to my knowledge never relased an service manual for any model they have,  even under an non disclosure agreement 

Lucky you,  i wanted to have this meter model


Send a mail at Gossen, and since its a "prehistoric" model,  i dont think the answer will be positive,  never did for my metrahit 28 and 29

good luck

Yeah, I had been looking for one such meter for years, until I finally got this one.
I just hope I can get it to be more useful than a door stop.
I already pinged Gossen, but they declined to provide the information, so far (was asking for schematics).

[giosif]

oh

You have this thread here  for some photos of a model 51,  no schematics tough

https://www.eevblog.com/forum/repair/teardown-and-repair-of-a-gossen-metrawatt-metratop-51-(model-m2510)-benchtop-dmm/

Thanks for that!
I'm aware of that thread (already downloaded all high-res pictures, for reference).

[giosif]

As expected, this is a non-working unit and I'm trying to get it back to functional state.

As per usual, what exactly is not working?  Be specific.

You are right.
I guess I was being lazy and was trying to avoid putting all the details here (there's quite a few things to mention).
Sorry for that!
I will be traveling in the coming days, but I promise that, on my return, I will be putting all the details I know, as well as photos.

In the meantime, the very short description of the issue is that, upon powering on the meter, the DCV readings are stuck at -0.128 mV, with no variations (not even the last digit).
If I switch to manual range, all upper scales (3V and above) display a reading of 0 V.
Switching to resistance measurements, the meter is displaying 0 Ohms and I measure no current generated between COM and the V/Ohms/etc. input terminal.
If I try to measure the input impedance with the meter in DCV, I get increasing values, well above the expected values (like a capacitor charging type of reading) for all ranges.
The PTC present for the input protection seems to be OK - about 800 Ohms value, if I recall correctly.

Again, I promise to come back with more info sometime next week.

[giosif]

Hi and a Happy New Year,

As promised last time, I am coming back with more details on this meter and the issues it's having:
1. Ever since receiving it, the meter has behaved the same way: no reading updates, with either a static value or 0 being displayed for pretty much all readings.
2. As seen in the pictures included, the meter is comprised of two boards: what I call "the main board" and "the display board".
    Each board has a microcontroller and associated EEPROM, all in PLCC package.
    The main board has an additional PLCC component - the ADC, which is a Gossen custom (or re-branded) IC - marking GMACD30.
    Initially, all these PLCC ICs were soldered directly to their corresponding boards.
3. Beside the main issue above, the meter would also sometimes fail to start properly (CPU seemed to be working, but display was garbled).
    This led me to think there might be a problem with the meter FW or access to it.
    So, I decided I would desolder the EEPROMs and check their contents.
    And, since I was at it and just in case one of the other PLCC ICs needed replacing, I decided to desolder also the microcontroller and the ADC on the main board.
    I did not desolder the microcontroller on the display board since, once started, the meter seemed to respond to controls just fine (e.g. changing measurement functions, going into setup, etc.).
    This is how the meter ended up looking like it is in the pictures - i.e. using PLCC sockets, except for the largest IC on the display board.
4. Going back to checking the EEPROMs and their contents, I was able to confirm they are working fine.
    Also, the meter intermittent startup issue seems to have gone away since then.
5. My next thought was to check the ADC IC and, for that, I took apart a working Metrahit 25S who has the same type of IC.
    So, I desoldered the ADC from the working 25S and installed a corresponding PLCC socket on that meter as well.
    Next, I installed the ADC from the 25S in our meter here, but that didn't make any difference.
    I then installed the ADC from the Metratop in the 25S and confirmed it was working (I only checked for correct multimeter functionality and rough measurement values).
6. At some point in time, I just happened to find some NOS for the microcontroller on the main board, so I replaced that, athough I didn't have any particular suspicions it was bad or something (i.e. I replaced it just because I could).
    As expected, this didn't make any difference in terms of meter behaviour.
7. Sometime during all the above, I also checked the power supply voltages.
    The unfortunate thing here is there is no silkscreen anywhere on the main board, so I didn't have any reference to check against.
    So, to the degree I was able to guesstimate the expected voltages, they looked ok, but take that with a not-so-small pinch of salt.
8. At this stage, with FW, main board microcontroller and ADC all confirmed OK, I turned my attention to the analog part, particularly the input.
    Here, as I mentioned in my previous post, I meassured the input impedance and here are the results:
    - DCV impedance:
        + Ever increasing (reached over 100 MOhms) for the 300 mV, 3 V, 30 V and 1,000 V.
           This was the case when measuring both with normal polarity (i.e. positive lead of measuring meter connected to V/Ohms input terminal of our meter here & negative lead of the measuring meter connected to COM terminal of our meter), as well as reverse polarity.
        + For the 300 V range, the impedance readings started at about 8.x MOhms, then gradually increased and settled to 9.95 MOhms.
            As with the other ranges, this was observed in both directions.
            However, unlike the other cases, when measuring the impedance with normal polarity, the display on the Metratop would actually change value and display -48.577 V (the value is, of course, wrong, but it was interesting that it displayed something different from 0 V).
    - ACV impedance:
        + Ever increasing (reached over 50 MOhms) for all ranges.
9. This particular meter model has 2 input channels, at least for voltage measurements.
    The observations at point 8 above apply only to CH1 of the meter.
    Input impedance for CH2 is within expected ranges.
    Yet, when activating CH2 readings, the readings would be wrong for it as well (see one of the pictures attached).

There are few more observations that come to mind, but I think this post is quite long already and I don't believe those additional points are that important (at this stage, at least).

Any suggestions on what to try next, please?

Thanks!

[Chris56000]

Hi!

It is normal for the input resistance of this type of lab quality meter to have gigaohm input resistances on the lower DC voltage ranges and typically 10M on the highest ones!

The inputs will be selected by fet switches on the analogue PCB by controller logic on the isolated analogue side, with communication across the isolated barrier to the main controller, but unfortunately trying to find it on a totally unmarked PCB is next to impossible!

PM Sent!

Chris Williams

[giosif]

Hi,

I am aware of bench meters normally having higher input impedance in the lower ranges, but the Metratop is not one of them - from the manual, for DCV:
300 mV____>20 MOhms
3 V________11 MOhms
30 V_______10 MOhms
300 V______10 MOhms
1000 V_____10 MOhms

For ACV, impedance is 1 MOhms across all ranges.

[Chris56000]

Hi!

Fair enough, you need to begin fault–finding round the circuitry associated with the input relays visible on your main board photos!

If you examine the PCB carefully you might be able to locate the input divider resistors, the relay drive/switching transistors or i.c., it's obvious from the readings you provided there's an open circuit in the input switching somewhere!

Check the relay coil(s) have a positive supply on one side, trace it back if missing. If the supply appears to be present, you can try connecting a d.c. multimeter on ohms to the positive V/ohms I/P terminal and then earthing the bottom side of the relay coils to chassis earth/0V to see if the expected 10M appears – if it does, trace back from the relay coil – if input resistance still very high/infinite with the relay activated, you'll have trace thro' from the input sockets thro' the contacts, etc., and seeing if you can pin down the Voltage Dividers!

Chris Williams

[giosif]

Fair enough, you need to begin fault–finding round the circuitry associated with the input relays visible on your main board photos!

I checked the one relay (I only see one, unless you mean something else?) and it is working fine.
In fact, part of the troubleshooting, I desoldered the relay and checked the relay coils are activated when selecting different functions (and they are).
And, it was with this occasion that I noticed something: before desoldering the relay, I could hear the relay clicking whenever changing meter functions; after I soldered the relay back, I couldn't hear it clicking anymore.
Upon investigation, I found out that, for the two negative pins from the two coils of the relay (they both go to ground), although it looked like there was a trace connecting them, I would measure about 4 Ohms between them.
I had to solder a wire between the two pins and now the relay is behaving as the first time, when changing functions.
The interesting part is that trace between pins is on one of the internal layers of the PCB, so we cannot speak of breaking the trace when soldering/de-soldering.
This makes me think the quality of the PCB might not be that great and it might also explain the current behaviour on the input impedance - i.e. another trace might be broken in a similar fashion.
Thus my original ask for the schematics of the meter.

If you examine the PCB carefully you might be able to locate the input divider resistors, the relay drive/switching transistors or i.c., it's obvious from the readings you provided there's an open circuit in the input switching somewhere!

Amm... Unless I'm misunderstanding you, the input divider resistors are the two longer blue components you see on the main board - one for each channel.
And I already checked them and they are fine.
Now, this leads me to one of the questions I have in mind: what components in any multimeter determine the meter's input impedance?
Because I don't think it's the divider resistor network, is it?

Check the relay coil(s) have a positive supply on one side, trace it back if missing. If the supply appears to be present, you can try connecting a d.c. multimeter on ohms to the positive V/ohms I/P terminal and then earthing the bottom side of the relay coils to chassis earth/0V to see if the expected 10M appears – if it does, trace back from the relay coil – if input resistance still very high/infinite with the relay activated, you'll have trace thro' from the input sockets thro' the contacts, etc., and seeing if you can pin down the Voltage Dividers!

As already mentioned, the relay coils do get driven.

[Chris56000]

Hi!

It's normally the input dividers themselves, sometimes with a series resistor of 1G for the lower voltage ranges!

Issues like you're getting might be due to open circuit "vias" on the PCB, which may or  may not be repairable depending on how the PCB was originally made, the only way of finding out is to meticulously examine and test them!

There doesn't look to be any issues with corrosion from yakky electrolytics, etc., or serious signs of liquid ingress so it could be down to problems with that lead free muck if the PCB was soldered with it!

Are you certain there's no obviously overheated or damaged SMD resistors on the PCB in the relavent area?

When I have a PCB or piece of equipment that is non–functional that needs drawing–out, I normally do this by means of a DMM with a fast continuity buzzer and my Circuit Diagram tool open, using Abacom's sPlan 7 for this, and make sure I have pin–outs for the ICs to hand!

The only DMMs of this class where schematics are available readily are the Keithley 2000 and Keithley 2001 and Agilent 34401A – it might repay you to study these as it will give you an insight as to how input circuits, voltage division, etc., is carried out on this type of meter – the design requirements for very low input currents and stability mean this circuitry is a lot more complex than the simple specifications in Gossen's book suggest!

Patience and perseverance should get you a resolution, but short of having one of these instruments myself I can only offer somewhat generalised advice based on my own experiences!

Chris Williams

[giosif]

Hi,

Sorry for the late reply (catching up whenever I get the chance)!

Issues like you're getting might be due to open circuit "vias" on the PCB, which may or  may not be repairable depending on how the PCB was originally made, the only way of finding out is to meticulously examine and test them!

This is what I am suspecting as well, but finding the problematic via(s) wthout the schematics is like looking for a needle in the haystack.

There doesn't look to be any issues with corrosion from yakky electrolytics, etc., or serious signs of liquid ingress so it could be down to problems with that lead free muck if the PCB was soldered with it!

I looked for any signs of leakage, but found nothing suspicious.
As a side note, I replaced almost all electrolytic caps, even if the original ones looked and tested good.

Are you certain there's no obviously overheated or damaged SMD resistors on the PCB in the relavent area?

I checked several times and no luck with such overheated/damaged SMD component.
That's not to say there isn't any present, as just recently found some perfectly looking SMD resistors, but which were way out of their intended specs (part of another repair, that one being successful).

When I have a PCB or piece of equipment that is non–functional that needs drawing–out, I normally do this by means of a DMM with a fast continuity buzzer and my Circuit Diagram tool open, using Abacom's sPlan 7 for this, and make sure I have pin–outs for the ICs to hand!

Hmmm... I can imagine this being a non-trivial task, so more of a reason to consider your offer. 

The only DMMs of this class where schematics are available readily are the Keithley 2000 and Keithley 2001 and Agilent 34401A – it might repay you to study these as it will give you an insight as to how input circuits, voltage division, etc., is carried out on this type of meter – the design requirements for very low input currents and stability mean this circuitry is a lot more complex than the simple specifications in Gossen's book suggest!

Good point. I will have a look at those and see if I can apply any of that info to the meter here.

Patience and perseverance should get you a resolution, but short of having one of these instruments myself I can only offer somewhat generalised advice based on my own experiences!

And this is much appreciated. Thank you!

[giosif]

Just an update to say that, after chasing my tail for some time, I was finally able to get the meter working again. 
In the process, I ended up buying a (working) Metratop 51 as well. 
But the two look good together (see picture), wouldn't you agree? 

The Metratop 51 cost a bit, but I was desperate to get it (I seem to have developed an obsession for Gossen meters) and it proved helpful in reviving the 52, although maybe not in the way one would expect.
That's because the fault with the 52 model was... corrupt firmware on the main board (or "multimeter" board, as Gossen call it when displaying firmware versions).
For no particular reason other than it was easy for me to try, I decided to copy the main board firmware from the working 51 on to the corresponding EEPROM of the 52.
And, when turning the 52 on, I could see the meter was now able to update readings and to autorange. 
Following that, I had to solder the ADC IC directly on to the board (as it was initially) and go through several IPA cleaning sessions (learned my lesson on that from previous meter repairs), before the meter could display readings close to expected values.
Even now, it's not spot on, in terms of accuracy, but I blame that to the fact it is not using its original firmware (which I suspect also stores calibration data).

Speaking of firmware, does anyone here with a Metratop 5x meter have any newer versions?
Mine are running:
 - Metratop 51
   * multimeter - AA
   * display - AI
 - Metratop 52
   * multimeter - AA (the original, corrupt, version was AC)
   * display - AG
I contacted Gossen and asked them for the latest firmware, but all they provided was excuses (that seems to be a trend with them lately). 
I also found another firmware dump on the Internet, for the main board only - that is version AF - and I could confirm it is working as well.

Finally, in case anyone in the future might be in need of such firmware, I am attaching the files I have dumped so far (yet to dump the AI display firmware).
I've also uploaded the same files on the KO4BB web site.

[Chris56000]

Hi!

You must have beat me by one euro on the 51 – I tried for that myself so I could get one to draw out the schematics and layouts for it!

Chris Williams

[giosif]

You must have beat me by one euro on the 51 – I tried for that myself so I could get one to draw out the schematics and layouts for it!

Oh.
I must have misread your last PM about that auction: I thought you were not going to bid for it and I wanted to make sure it didn't disappear, going to some "unknown person" (i.e. not on this forum).
I'll send you a PM, maybe we can sort something out so you can finally trace out the schematics.

Ah, and something I forgot to mention in my update yesterday: it turns out you were right about the imput impedance for the 300 mV range - it is much higher than the 20 MOhms mentioned in the official specs (above 500 MOhms for sure; I suspect something in the GOhms range).

[giosif]

And just a quick update: I dumped the display firmware from the Metratop 51 meter (i.e. version AI) and attached it here.

[Chris56000]

Hi!

I'm currently about 45% of the way thro' drawing the schematic for this instrument, but it's a load of disconnected bits and pieces at the moment, and I'm still dabbing thro' the board (meter on loan to me for this purpose) joining all the dangling lines on my drawing sections up!

It will be posted on K04BB when completed, and there'll also be physical layout drawings to follow at a later date!

Chris Williams

[coromonadalix]

Sad to see Gossen act like this,  even for my 28 and 29s,  no support nada niet, i think "they" want their old technology to simply dissapear from the earth the fastest way possible ?

I dont know for their newest models, prices are very high, acessories prices very high ...  and i got discouraged when joeqsmith has done a thread on a gossen stange proximity sensibility ... and enhanced the shielding.