Nixie DVM restoration --- Rochar A.1335

[Vince]

Hello world !

I thought I would share this repair / restoration as it's quite a rare piece even here in its home land. I haven't found any other piece of Rochar TE on the forum. Well the search engine didn't at least.

It's really sick and driving me nuts, spent a lot of time and pulled a lot of hair already on that one... but vented all that in the TEA thread so far...the poor buggers there suffered for sure.
Here is my introductory post about it on TEA there a few months back, for more pictures and details.

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg4100953/#msg4100953

Will try here to summarize succinctly what happened so far, before I resume work on it and add some more.

I actually have two of these things, as you can see. Same basic model but different options.
The one on the left is the one I am trying to fix. It does only DC voltage and manual range.
The one on the right is the luxury model, can do both DC and AC, and even has autoranging, wow...
This one works fine and I am using it to help me fix the sick one, by taking measurements on the board, and swapping boards.


ABOUT IT

I mean, I don't expect anyone to know about this thing, it's no HP nor Tek, so a few words can't hurt I guess...
" Rochar " was a short lived and long gone French test equipment company. Lasted 20 years or so, between about the mid '40s up to mid '60s, at which point Schlumberger bought and killed them. No more Rochar.


CONSTRUCTION

Steel chassis that holds everything and slides out of the enclosure from the back. Held in place with just 4 little screws at the bottom, same screws that also hold the 4 feet in place.

It's all discrete components of course, given the era, with all germanium transistors. Lots of boards inside, that plug into a vertically mounted "backplane", via fancy top quality FRB "Hypertac" connectors... which sadly are not bolted to the backplane, so zero stress relief... the solder joints take all the abuse when you plug/unplug boards. Sure doesn't help with all the random problems I am having with this instrument  .


SPECS

- Accuracy : 0.2%, much better than my DMM daily driver, a Fluke 11 at 0.9%.... so the Rochar should not be ashamed at all, it's still a useful bit of kit ! 

- 4 Digits (Nixie tubes) + Polarity indicator tube + Decimal point

- 5,000 counts but, interesting quirk I have never seen before (have you ?) : it's designed to display only EVEN values. So you lose a bit of resolution but double the dynamic range, why not. So it can only display 0000, 0002, 0004 but not 0001, 0003, 0005 etc.

- 100 kHz (non RMS) bandwidth.

- 5 / 50 / 500V ranges

- Impedance 5 Mohms, but lowest 5V range can also be switched to super high-impedance if so you desire.

Rarity for any piece of old French TE, there is actually a manual for this thing ! Nowhere near as comprehensive as the big American players would produce at the time, but good enough to help me out. At the least, it's got the schematic for this thing, what a relief !  Schematic singular.... tehy put everything onto a single giant sheet...

You can find the manual there (it's too big to attach here) :

https://archives.doctsf.com/documents/feuilleter_document.php?num_doc=95099

For convenience, I extracted the schematic sheet and reworked it a bit so I can post it here, given the size constraints of the forum, as a JPG image.


CONDITION

Mechanical and cosmetics I mean.

EXTERIOR

Has a normal amount of "patina"/wear, but no major accident, like broken knob or big scratch or gouge or on teh face plate. It's very dirty of course but should clean up just fine I think.

Only point I would like to address are :

- Missing its 4 feet : not worried, looking at the other, complete meter, they look like standard shaped rubber feet, shouldn't be a problem finding something suitable...

- Chassis lose / rattling around in the enclosure ! Totally unacceptable of course...
The 4 screws that hold it in place, also hold the feet in place... so I need 4 feet AND 4 screws. Said screw don't look anything special so won't be a problem replacing them that looks period. Slotted pan cake head.

- Surface rust on the rear cover. Easy enough to fix.

- The orange colour filter for the Nixie tubes is lose, can't stay in place, it's dropped half way inside the enclosure. Typical problem of Rochar equipment : the filter is held in place only by strips of foam that push it agains the back side of the face plate. Obviously after many decade, the foam has turned to dust and is not exercising any pressure any more. Easy enough to fix.

INTERIOR

- Some surface rust, on the inside of the enclosure, mostly at the bottom.

- Some more surface rust, on the bracket that secures and aligns the Nixie tubes at the front.

- More annoying : slight corrosion on the steel chassis, and even some fungal looking "constructions" on some of the boards.
So I guess this puppy has had a bad life, stored in some damp garage for years, which is probably responsible at least in parts for the random, hard to troubleshoot issues that have been driving me mad for the past few months.


ELECTRICAL ISSUES

Getting to the core of the problem now... so what's wrong with this thing ?
Again it's hard to say because some of the problems are random and hard to chase, but some are easier to grasp. Some have even been fixed...

- Uses obsolete powers socket, one can buy a cord for this thing.

- Doesn't power up when you flick the power switch.

- Two exotic / special resistors (special 1,050ppm tempco) suffered corrosion damage, not easy to replace, they don't grow on trees. They are part of a board mounted underneath the chassis. It's job is to produce two reference voltages, from an aged and hand picked 6,2V Zener diode. One tap goes to the analog to digital conversion board, where it set the slope of the ramping voltage of the sawtooth. the other tap gives an accurate 4,000Voltage to let the operator calibrate the full scale deviation of the display, using a trimmer on the front panel. I have not yet got to the point where it makes sense to bother working on this, there are more basic/fundamental stuff to fix first !

- Nixie tubes, from cold, display garbage for a miute or two. Evnetually it clears up and they work fine. I am thinking the caps in the high-voltage section of the power supply, are on their way out.

- Power supply : other than the HV for the tubes (200V), the power supply produces +/- 6V and +/- 12V.  I can't remember numbers (search TEA for the details !  ), but voltage levels and ripples on some of the rails were not good at all.

- There was a problem with the display. The 4th / MSD tube would always stay at '0', and the tube before it would display both a '0' and a '1' at the same time...  So I swapped that board with the working meter, and that was it. So at some point I will have to fix that board...

- Massive zero offset problem : from cold it displays like, in the thousand counts ! Then after a few seconds it progressively drops exponentially and after a few minutes it settles to 50/100 counts. Still way too high. The trim pot on the front panel doesn't have enough range to compensate for that.
All the magic/brains of this instrument is located on the left most board, the one that carries the polarity tube. It takes the input voltage, from the attenuator, and does the conversion to time domain, then generates the clock to drive the counter/display. It also holds the trigger circuitry that let's you start a measurement either manually or automatically. I I swap that board for the one from the good meter, it's all fixed : offset is bang on zero with the trim pot centered, and does so from cold, and does not drift... it's bang on zero fro cold and stays there, it does not move. So the problem definitely comes from that board. Have started working on this issue but it's not easy. Have not yet found the problem. The manual calls this board the "Comparator" board, so I will call it that as well from now on. It's basically the brain of the instrument. I attached a schematic of its board, annotated to help identify its various sections.

- Trigger problem : the controls were acting weirdly, or not responding at all.

- Display "free-running" from cold. At power up, from cold only, the display would look like it's counting and never stopping, then after 15 minutes it stop doing  it the meter behaves normally.

- At some point, the leftmost tube, that indicates the polarity, started working intermittently, soon to fail solid.


WHAT'S BEEN DONE SO FAR

- Obsolete power socket

Replaced it with a modern IEC one. Old on was IEC too, just an older standard. So the new one had the same footprint, fitted right in !
Same shape, same pitch for the screws of the said tabs, a drop in replacement, I was pleased. Not just because it made my job easy, but also because it meant I didn't have to "rework" / mutilate the instrument in any way shape or form, and could even easily revert to the old socket if so I wished.

- No power issue
 was the power switch. Finding an exact replacement, to preserve the look of it so as not kkep the front panel consistent, took some effort. In the end they are made by a French company by the name of SECME, model "DJET", 10mm diameter. Company still exists today, though they are now calling themselves "EOZ". They still make these DJET switches a century later, but turns out they are MIL spec stuff so very expensive. Like in... I can't buy even just one, cost as much if not more than what I paid for the instrument shipped. But I was able to find a few used ones for sale on Ebay at a much more reasonable price, so I have now quite a lot of them in various configurations.


- Power rail issue

Was an open circuit filter cap. replaced it it's much better now...  However all these filter caps come in funny awkward packages, mounted vertically with the bottom end screwed onto the chassis, often with one lead goign underneath the chassis, but the other exiting at the top on the other side of the chassis. Anyway, never seen this type of package before a pain, so ended up leaving the old cap in place and hot snotting the replacement cap underneath the chassis, which it barely cleared, phew.


- Failed polarity Nixie tube

Turned out to be a bad solder joint on the rear/backplane PCB. However it's a failure mode I had never seen before : the solder joint itself, from the outside, looks just fine. You only discover the problem once you suck the solder up to reveal the underlying copper pad --> the pad is supposed to be TINNED, but what I saw there was bare copper instead !! How can you see bare copper when you suck up a solder joint... so basically what must have happened is the copper pads were not cleaned properly before tinning, so the tinning didn't stick to the copper. The solder joint would look nice because it would stick to the tinned pad just fiune, but since the tinning itself would not stick to the pad... then basically we end up with a solder joint that's kinda "floating" over the copper pad, hence crappy connection. So I fixed that joint, and that cured the problem.

However it cast a hge shadow of doubt over ALL THE OTHER joints on that board ! Problem is that as I said, you don't know if the joint is good, until you have sucked it up ! So  I had no choice but spend 2 hours with my manual solder sucker pump and solder wick, to remove dozens and dozens of joint, so I could inspect them. It was the effort though, as I found maybe a dozen bad joints. So it will help a lot in fixing random issues and preventing new issues from popping up ! Because this meter is really driving me nuts with faults that appear all the time and disappear, and new ones coming up etc.... it was a complete shit-show !

Now I still have issues, , but less random. Still haven't managed to get to the source of the problem but at least they are repeatable (enough) that I can probe stuff and try to trouble-shoot things. 


- Free running display and Trigger circuits

At first I thought they were separate issues, but I got nowhere. Eventually changed my mind and thought that maybe the two were linked. The trigger circuitrty generates the reset signal for the counters / Nixie boards. So I though OK, since I have problem in teh trigger area already, maybe there is more to it, and it also generates a crappy reset signal, which would cause havoc in the counters, and make them display garbage that "looks" like they are free running (displaying all digits at the same time", but are not really, and instead the counter outputs are just stuck in an illegal state. So I decided to concentrate first on the trigger area. It revolves around two PNP transistors, Germanium of course. One of them definitely tested bad, leaking heavily, as they do I gather, not even recognized as a transistor any more. I replaced both transistor to keep the circuitry balanced, just in case it might matter. Of course I had not these Germanium transistors i stock, so instead I made a bold move and replaced them with a general purpose Silicon transistors, good old 2N2905A just because they happened to use the same TO39 package as the old Germanium transistors, so it would not look out of place.

Well, Trigger was now working fine and it did cure the free-running problem from cold, yeah  !

TEASER : the above statement was true a few months ago at the time I did it ! But a few months later, it misbehaves again ! 
So, from cold only, free-running and trigger controls not obeying. Then afer a few minutes warming up, it starts working normally.

So am now thinking this "delay" before proper operation, could be due to some capacitive effect maybe. I had already replaced 3 electrolytic caps on that comparator board but it changed nothing.


WHAT'S UP NOW ?

OK so now I have recapped the main points of what happened, only skimming the surface of course sorry, for th sake of brevity...I can now resume work on the thing and add some fresh content here as I go.

This trigger / free running problem is my first nightmare. Been spending so much time on it.. it drives me completely nuts. However I am hoping I will prevail one day, armed with all the troubleshooting and attempts I made, and my knowledge of this board and how it works, which is now orders of magnitude better than when I first started. It's not that huge black box any longer. HOwever probing that board is still so time consuming since the components are not identified ! It's depressing... spending hours tracing the board every time I want to locate a pin I want to probe    Slows things down CONSIDERABLY ! 

I have been traumatized by this particular issue.. even months later I am still traumatized. So the plan now, is to work on lower hainging fruits so to speak, issues I think should be easier to trouble-shoot... so that it can boost my morale and then I might find the motivation to go back to the more challenging problems...

So what do I plan on working on then ? Well I am thinking of that faulty Nixie board. It's fault is solid, hot or cold, so it should be easy to trouble-shoot.

[Vince]

Pic of the schematic, reworked, attached.

Also close up pic of "comparator" board, the main board where all my troubles lie I think... spent so many hours studying it and probing it.
You can see the fungi that developed on it. I cleaned that of course, with some IPA. Looks presentable now.

Also the annotated schematic for it, to help understand how it's put together.

EDIT : added a pic showing that board being scoped. That's at least one good thing about this board.... as you can see it's very easy to probe it while the instrument is running. The chassis can slide out of the enclosure, and the board being the leftmost one, you get excellent access to the component side of it. So there is at least that going for it.... 

[Vince]

Bad solder joints.

Here showing the first one I discovered, that caused the polarity indicator to fail (that joint was carrying the 200Volts going to that particular tube).

You can see on the close up pic after I sucked the joint, that the pad sports bare copper, unbelievable....

[Vince]

Here you can see the two failed exotic special tempco resistors, part of the little voltage reference board mounted underneath the chassis.

The resistors are these two bright yellow big puppies. They are wire wound it turned out. The yellow jacket is plastic/rubbery stuff, and the green corrosion was coming out of the jacket where the leads exit. I think moisture got inside the resistor and corroded the insides, and eventually that "leaked" outside for us to see.

[trobbins]

Were you able to recover the existing WW portions of those two resistors by recovering or replacing the end caps and mounting leads?

[Vince]

Nope. Someone on TEA suggested attempting that, but I don't have the skills and tools and experience to tackle such a delicate job, sadly.

So these two resistors went to the trash, they are no more 

Someone on TEA had a look I remember, and said that this type of resistor (with that special 1,050 tempco I mean) somehow still exists today, he found some in the USA on Digikey or Mouser IIRC.
So I could go there, but it's expensive as you need 50 bucks minimum order to get free shipping and customs.
So I would try to find some locally / in Europe first, and resort to importing only as a last resort.

I have not researched the subject any further because as I said above, it makes no sense to waste time and money  on this issue, unless I have first managed to get the counter working properly and reliably.
If I can fix this counter, then it will be worth spending some money to buy these special resistors.

Interesting maybe, below is a pic of another Vref board, the one from my other meter, the one that works.

We can see that Rochar used different looking resistors. They are the two green ones. They are much skinnier than the big yellow ones from the other meter, and also look a different construction. Maybe they aren't even WW, who knows.  They look much more modern at any rate. The serial numbers denote that : the broken meter with the old looking yellow WW is serial number #581 , versus #1108 for the good meter with the more modern looking green resistors.

[Vince]

Just made this montage to compare the green and yellow resistors.

Tried my best to make apple to apple comparison : to scale as best I could, same viewing angle, and same resistor and spot on the board : sitting next the Zener diode (black bodied).

One can see just how much skinnier the green ones look in comparison. Different package as well, perfectly cylindrical, with square ends, no bulge on the body nor roundiness where the terminals exit.

[Vince]

Now I think of it !!!

Huuu....the meter that has the "tiny" green resistors... also shows obvious signs of previous rework/repair here and there (might be why the meter works fine... it's been maintained ! ).

So it's perfectly possible that these green resistors look more modern... because they ARE modern, non original. Maybe the meter originally came with the same big yellow resistors, which failed in the same way, so the previous owner replaced them with those more modern green ones.... and it works just fine.

Just an hypothesis, but I guess it holds water.

[pcprogrammer]

I noticed that the two boards are not the same, and also that the two devices are different even though the model number is the same. Just look at the front panels.

So maybe the green resistors are original and the design has changed during production.

[Vince]

The two meters are the same basic model, only different variations of it, as I explained in my long post !

The older, broken one with the yellow resistors, is DC only and manual range.
The later one with the green resistors is AC+DC and autorange.

90% of the instrument is the same. The parts that matter to me at least : all the Nixie boards, the main/comparator board, power supply, back plane, Vref board... it's all the same.


The AC  and autorange extra goodness are located / confined to two small boards that handle ranges / attenuation. But I don't have issues with these boards on my defective meter, so I don't care.

The schematic shows of course the most complete meter, the AC+DC autorange. But again, the changes are limited / isolated to the two relevant boards (top right corner of the schematic). All the rest of the schematic is common to both versions.

Yes you are right the two Vref boards look different. I posted at length about that on TEA at the time, when I looked into that.
Scared the shit out of me at first, thinking oh no I can't rely on the schematic but... to clear things up, I reversed-engineered both boards, only to conclude that :

1) The most recent one with the green resistors, 100% matches the schematic.
2) The older board with the yellow resistors, is basically the same circuit but with slight variations :

- The chain of the green 3K resistor in series with a "normal" 3.7K 1% resistor, has been replaced with a similar configuration but the values are split differently ( but the sum is still the same) : 1K for the yellow resistor and 5,73K 1% for the other.

- The other green one, 2.1k 1,050 tempco resistor has been replaced with a yellow resistor + a 1.33k "normal" resistor in series.

- The 1% 3.9k resistor has been replaced with a 3.75K one in  // with 56K. So we get a sliiightly lower resistance, but who cares because it's in series with a 200ohms trim pot anyway. It's there to adjust the 4,000 ref voltage to let you calibrate the scale deviation. So as long as I can adjust the pot to read 4,000V, it's fine...

I may convert the older board to the simpler (and proved to work !) later design of the green resistor board...

[Vince]

OK, 2 months later I am back working on this thing.

So as I said last time I decided to concentrate on trying to fix the defective Nixie board first, hoping it would be easier to diagnose than the other problems.
Unfortunately I was wrong, it's a pain.

So last time I did some mass DMM buzzing on the board, checking in-situ every diode, resistor and tranny on the board, comparing values with the working board form the other meter.
Was not fun buzzing the trannies because there leads are 100% inaccessible from the top side, and on the bottom side, every thing is coated with translucent yellow conformal coating, that my sharp SMD proves can't pierce  reliably, and it requires too much force anyway, it wears them out fast... not what I want. So instead I reflowed every solder joint of the trannies, so as to melt the coating and reveal some fresh solder the probes could work on. Was a pain.... 48 joints to do. All that for nothing : all trannies tested good.

Was a bit miffed, was hoping for some dead diode or tranny, because you know, these old Germanium transistors carry such a bad rep don't they ! Well it appears it's not all that deserved if you ask me !


Fast forward 2 months. Yesterday I spent my entire Sunday afternoon resuming work on this board.

Since buzzing the board didn't do it, next step is probing the thing live. Sadly unlike the main board which the outer most board hence accessible from the side... the Nixie boards are not. Need an extension... but need suitable connectors for that.

The connectors were not easy to identify, but people on TEA and old friend here helped out.
They are made by FRB, a French company, using tr "Hypertac" tech, " K " series, 4mm pitch, IIRC.
Fancy ultra expensive connectors... good luck finding them new today. Spoiler : don't even try.
Even used they are super rare. The best I found, well, I think someone else did find them for me, was an Ebay ad luckily in Frog land, a guy selling a whole lot of random vintage connectors... in the lot were a few of the ones I needed. So I had to buy the whole lot just for that.
There 5 pieces of the type I needed, pretty much all a different combination of PCB mount vs solder cups, right angle or straight, male or female... there is a bit of everything. Luckily there is both the male and female version in the configuration I need (solder cups, straight).

UN-fortunately... these connectors aren't the right size : 16 contacts versus 12 for the Rochar.  That's because I bough these connectors initially to make an extension to fix Nixie board in another instrument, a Ferisol counter, which do have 16 contacts.

That said 12 or 16 makes no difference in practice since as I said, this ad on Ebay was the only one around ! So 16 contacts it had to be anyway !

So...fitting a 16 way female connector onto the board... no worries, it works. It just sticks out but no harm.
However fitting a 16 way MALE into the female socket on the backplane inside the instrument... does NOT work : the ends of the 16 way connectors interfere with the plastic bae of the 12 way connector. I could make it work if I mutilated that connector to remove its plastic ends.. but of course I am not going to mutilate this poor little instrument.

So instead what I did,was to solder the wire of a ribbon cable on the OTHER side of the backplane connector, because it's easy to access of course, it's right there at the rear of the instrument. Plus the joints are big and spread far apart from each other, 4mm, so it was easy to solder my wires to that.
Long, painful, but easy.

Result in pictures.

So I could now start scoping around !


See schematic and board below to see ow it's put together.

It's all discrete components, all Germanium trannies and diodes.

The board is in 3 parts :

1) The counting
2) The decoding
3) The driving of the Nixie digits

Given the symptoms I witnessed, I figured the problem was most likely lying in the counting part : when this tube is counting, all it can do is display '0' and '1' superimposed, at the same time, instead of all 10 digits at the same time. Also, the next Nixie tube on its left, is stuck on '0'.... but I know this one board works fine (swapped it in the other meter). So it means that it's not being incremented, my defective board is not sending it pulses.

So all in all it looks like the issue lies in the counting part of the board, and that the problem is mostly likely that it fails to count above '1', hence it never overflows hence never sends pulses to the following board.

As you can see on the schematic, the connector on the board is nice : it gives access to the input of all 4 stages of the counter, even though of course only the last / fourth output is being used by the instrument.

so I scoped the input of the board / first stage of the counter : I get the expected 4,000Hz pulses coming from the previous board, good.
Then scopes the ouput of that first stage, expecting it to work : it does, it sends pulses.
Then I scopes the outputs of stages 2 3 and 4, and got nothing... as I assumed.

So I am now concentrating my efforts on troubleshooting that second stage.
So I am trying to scope things in there, but sadly it's rendered extremely differ in practice because the components on the board are not identifiied !
It's easy to look at the schematic and say OK let's probe here and there and this and that but... actually finding these point on the board... good luck.
As pictured, the general layout of the board is easy to grasp.." general" is the word... but going the next step and have an accurate, reliable representation of where each and every component is , and its orientation... good luck. Especially since th layout is artisanal so to speak, and not very consistent in places, so a given component is not always in the same place between stages.  I tried my best but it's giving me headaches...

So instead I mass tested all the trannies... I already had buzzed them in situ 2 months ago as I said, but this time, I took the time to pull them all (the x8 from the counting part, didn't bother with the 10 from the driving part of the board as it's irrelevant here....) off the board and stick them in my cheap Chinese.component tester... and they are all good 

Checked all diodes again... still good.

In desperation I pulled the ceramic caps, the yellow (all of them) and the red ones (only the two from stage #2).... all good.


What else did I do.... ah yeah. I though, hey maybe it's the FIRST stage that misbehaves... OK it does count, but may the voltage levels of the output aren't good and it fails to trigger the next stage. so I compared waveform and voltage levels with the good board, and they are identical.... so no, indeed the first stage is good and the second one is not.

Problem : after having done all that.... I just don't know what else to do ! It's supposed to be a simple board, should be easy to diagnose but it's proving not to be !!!!!!  So much for a low hanging fruit / easy job, this DVM has indeed decided to drive me NUTS from end to end ! I am cursed !!! 


So I though of the possible steps to go further :

1) since it's such a pain to locate components on the board, I thought maybe I could wire the good board in  // with the bad one, siting side by side.
This way I could just probe one point and immediately probe that same point whatever it is, on the bad board, and compare.

2) Components passed basic tests, but maybe there failure is less obvious. Since it's old, maybe one or more diode is leaking badly in reverse.
So I could pull them one by one and test their reverse current. Or I could just replace them all... there are 15 of them but it's cheap and easy to do.
I do have old Germanium diodes, or modern copies (Schottky), though in all fairness replacing them with bloody 1N4148 silicon diodes would probably work just as well...

3) Last resort : test all these little ceramic caps for leakage... but these aren't supposed to leak so... last resort.


I would like to finish on a positive, bright note : it's a SOLID failure, so I can probe around as much as I want, do a million tests and experiments, and eventually I will, mathematically... find the problem !!! It's just a matter of time and perseverance...


So I did that yesterday, but what I do TODAY ? NOTHING ! Left work late, then spent 2H30 writing this very message and editing the board and schematic pictures.

23H30 now, super tired now, going to bed, will correct all the typos tomorrow when I am back from work.

In the meantime put together your best ideas and share them... I need every bit of help I can get here ! 

[factory]

The mention of diodes reminded me of the Solartron LM1420 voltmeter, it uses a similar (but not identical) circuit for the counter/display boards, this was the thread in question: https://www.vintage-radio.net/forum/showthread.php?t=74904
Unfortunately that probably won't help as the failed diodes were found to be open circuit, I've got two but can't test them without the unobtainium input cable.

And yes ceramic capacitors can fail, it doesn't happen very often, not had one fail for a long time, then had one turn into a resistor in a Tek scope and another went short circuit in a item at work, within the same week.

David

[Vince]

Well that helped, thanks... "good" to know that it's not impossible for these old ceramic caps to fail.

That Solartron of yours looks a lot like this one :

https://www.leboncoin.fr/collection/2250922598.htm

Guy sells t for only 5 Euros.. if you are interested.. might be worth the shipping and import duties ?
Ad title says it's reserved to some guy, but it's been saying that for some time now.
Same seller that just sold me my Rochar DMM, which also was supposedly reserved... until it was not anymore.
So if you are interested maybe I can ask him if that guy it's reserved for has given signs of life or if he can sell it to you instead...
For only 5 Euros it's worth a try I guess...

[factory]

Yes that is almost identical to the two I have, except the markings are in French on that one.
The only thing I'm looking for is the test lead that plugs into the socket on the front, which as usual seems to be missing.

David

[Vince]

Ah, it uses a special input connector... that sucks no end 

[psykok]

Well that helped, thanks... "good" to know that it's not impossible for these old ceramic caps to fail.

That Solartron of yours looks a lot like this one :

https://www.leboncoin.fr/collection/2250922598.htm

Guy sells t for only 5 Euros.. if you are interested.. might be worth the shipping and import duties ?
Ad title says it's reserved to some guy, but it's been saying that for some time now.
Same seller that just sold me my Rochar DMM, which also was supposedly reserved... until it was not anymore.
So if you are interested maybe I can ask him if that guy it's reserved for has given signs of life or if he can sell it to you instead...
For only 5 Euros it's worth a try I guess...

It was reserved indeed ... to me

[Vince]

Oh cool, so it ends upon TEA then, lucky meter. Hopefully you will give it a good clean and get it working if it's not already... what about this special input connector, do you have the special cable that gos with it ??  ... I guess one can always replace the connector with a BNC socket or something...pardon the blasphemy, I apologize i advance ! 

[psykok]

I'm already working on it.

Was not working at all when I got it.
After replacing some very bad caps in the PSU section, I got it running again but it's just displaying random  numbers.


I also managed to clean the front plate


Regarding the special input connector, unfortunately  I haven’t  got the cable with the device.
I will try to replace it with something else 3 pins BNC or mini xlr 

[Andy Watson]

I got it running again but it's just displaying random  numbers.

It's been 40 years since I saw one of those working. If memory serves - the open-input readings are quite random - it has a very sensitive input. That input connector also has a guard/screen connection which helps considerably!

[Vince]

OK so as planned, I wired the other, god working Nixie board in // with the bad one so I can probe both boards on the fly and compare them easily.

So I soldered a piece of ribon cable at the connector of both boards.

The boards were therefore showing their solder / bottom side.

I powered up the instrument, noticed that the Nixie tube on oneof the boards was not lit... fast forward 60 seconds later, and I realize that I made wiring mistake, I swapped two wires at oen end of the ribbon cable... I swapped the 200 Volts supply for the Nixie, with an adjacent pin 

But before I got to see that, I went to flip both obards so I can see their top / component side instead, so I can probe component pins.
So I first grabbed one of the boards, by its nixie tube, rotated it and... as I was almost done flipping it, the solder side accidentalliy contacted the solder side of the other board... big sparks !!! 

... which means the 200V supply sent its juice to god knows what parts of the circuit.

So I basically destroyed my little Rochar. I don't deserve it. 

It's still "somewhat" alive, though. All Nixie tubes are still lit, but now it diusplays '0000' no matter what you do with the instrument controls.... it's dead jim.
Pllus, the good Nixie board now misbehaves, it display all digits at the same time...

This intrument was already probing to be a huge challenge, was extremely sick.... now it's not  just extremely sick, now it's kaput !

It's already 22H30 here, I am tired... I will do some basic trouble shooting for 30 minutes tops, then go to bed.....

I am cursed, this meter will never work..   

[factory]

Ouch, I had a similar accident with the display supply in a Philips counter while probing, I should have been more careful.
I hope you have another look at the Rochar, maybe work on the board for the working DVM first, looking at the diagram it looks like the +200V got swapped with the -12V, another thing you probably don't want the outputs connected together, i.e. the count to the next board and the four BCD printer outputs? If you have another look I would wire up the +200V last after verifying the other connections, maybe even use a different colour wire for the +200V and double check before power up.

Are there any obvious shorted transistors/diodes? if there are it might be possible to convert to silicon if you can't get the 2N1305 or similar, HP did this with their later revision boards, some resistor values also changed, I know this as I converted a later board back to older rev, for an earlier counter (that had a cracked PCB).

The all zeros, does that mean it's stuck in the reset state? This looks to be Z on the diagram.

David

[Vince]

I hope you have another look at the Rochar,

OH yes, I am totally in love with this cute little thing ! Plus it's extremely rare, so it's staying with me 'til the end of times, well the end of my time anyway... whenever that will be.

It's just that it's so sick and progress is being made so slowly yet requiring so many hours.... so it's a bit of a love - hate relationship as the brits would say.
So when things go wrong and/or I am going nowhere on a particular problem.... I feel the need to leave it alone and move on to something else to clear my head.... but only to come back to it later.
So, right now I am motivated to work on this thing for a few weeks in a row I guess... before I get exhausted again and put it aside for a few weeks/months again...

But ultimately I love this thing and I want it to work properly so I can actually use it on the bench.
So it might take me a year or two I don't know, but I will get it working, eventually ! 

[...]another thing you probably don't want the outputs connected together, i.e. the count to the next board and the four BCD printer outputs?

Oh no fuck me you are right, I TOTALLY forgot to take that into account, NOOOOO !!!! 
I am not worthy.... 

The all zeros, does that mean it's stuck in the reset state? This looks to be Z on the diagram.

I don't know, I need to start troubleshooting the thing to see about that....

Yes the 'Z' you see every where on the schematic is indeed the reset signal.

16H45 here. I am done with the QUAD stuff, so I will switch to the Rochar and probe around all evening to try to figure out more precisely what works and what does not

I will work on the thing a bit every evening during the week as well. I want to make progress, I am tired of seeing this thing (and it's brother that's working), both in bits all over the bench. It's depressing, and the longer it lasts, the higher the chances of losing or damaging parts. I need to bloody fix this thing !

[Vince]

OK I worked a bit on the thing.

YOu were right about diodes, I found at least a dozen bad ones, spread across many boards.
Replaced them with other Germanium diodes I have in stock, not the same ones by the looks of it, but similar enough I thought... transparent glass package, a hiar larger.. good enough.

Now works better, but still much to be done, basically all Nixie boards and the main board are now faulty even after replacing all the diodes.
I don't have many good boards left for sure... not enough to swap them all.

But overall I would say it('s probably fixable. The power supply is good enough, and there is activity going on every where... just not as much activity, and not the proper kind of activity... but activity none the less.

In short it's very sick, much more than it was before (which says a lot), but not blown/cooked... I am sure I can make it work.... gonna take time and effort, but despite having now 5 defective Nixie boards and two bad main boards, I don't see any clue that it's not recoverable.

I will keep at it and keep you informed as I make progress...

One really bad news though, and THAT is irrecoverable I think !
Just realized I have one dead polarity indicator tube ! 
Good luck finding a replacement !
The tubes that display digits, might be possible to find, with patience, but the one that displays the polarity and symbols..... erm.... not too hopeful 

I guess I will spend a year searching for one, failing then resorting to trying to adapt a symbol tube of a different brand or model, as long as it's the same form factor. The font and glow probably won't be 100% the same so might look a bit out of place but ..... 

[Dave Wise]

Keep the hope, Vince.  I blew up half the parts on the ADC board in my Monsanto 2000 DVM when I tested isolation between Input and Guard.  After replacing at least ten transistors and fixing the bad factory assembly that gave inadequate mechanical spacing, I eventually got the instrument working again.

[Vince]

Thanks Dave !

Yeah I will keep at it... it's become a challenge... if I failed to fix it, it would mean I am worthless at electronics, and I just can't accept that !