Nixie DMM repair updateBeen working all day long on the thing. I think I am making progress... I swear I do !!!
SwitchesI received the two lots of SECME DJET 10mm switches I bought on Ebay France from these two vendors.
Then I realized I HAD some of these in stock already !!
I had/have 9 toggle switches no less ! All SPDT. 3 are the 10mm ones I need for these old Rochar instruments, and 6 are the skinny mini 6mm ones.
So I have lots of these switches now....
So I could at long last replace the faulty power switch !
It works fine now, yeah... so much easier and more convenient to have a little switch on the front panel rather than having to unplug the thing all the time.
Once I had pulled the old broken switch, tested it and looked at it closer and... hmmm.... some fuckery has been going on here ! Someone messed with that switch ! Turns out... it DOES work ! And it's not a SPST it's a SPDT ! But... the middle pin, the common... see picture, broke off somehow. You can still see a tiny bit of it sticking out of the plastic body. Sticks out enough for me to put a DMM probe onto it and do a continuity test. Test reveals that the switch works just fine. So, some idiot not me, probably tried to "fix" the switch by relocating the wire to another terminal ! Hey, all terminals are interchangeable in a switch, no ? No ??
So I replaced the switch and added a piece of heatshrink tubing on the middle pin, as was done at the factory, if I am to take the other / working meter as a reference.
Works fine now.
The front panel assembly is easy to get out. Just 3 easily accessible screws, a connector to undo underneath the chassis, and the assembly comes out and gives you great access to all the switches.
This evening got a message from ebay... one of the two ebay sellers is telling me he did some digging and excavated some MORE of these 10mm DJET switches ! Asked if I might be interested...
https://www.ebay.fr/itm/125268668056I might well be... 6 switches, 4 of the toggle type, and two push-buttons... the latter I have only one of, and they appear to be much harder to come by than the toggle form. So... at only 10 Euros + 6 shipping... I think it's a good deal... His ad allows for offers, so an offer I will make...
Dodgy solder joints on back plane PCBYes.. the meter is "free running" again... my fix didn't last long did it
This time wiggling that #2 Nixie board, and the ground wire on the back plane, does not cure it... so it's a new problem this time.
I hesitated between committing suicide or persevering. I chose to persevere.
So I redid all the joints, 100.00% of them, on BOTH sides of the back plane PCB. I already did those in the back but only touched them up. So this time I sucked them all with my trusty solder sucker. Took a while, an hour at least.... THEN I could definitely see which joints were good or bad. And I did find like 10 or so which were bad, where, again, the tinning didn't stick to the copper pad
This PCB is a disaster...
I redid 100% of the joints with silver solder, because I am generous and want to help the stars align in my favour, hopefully.
Getting to the other/inner side of the PCB obviously required to pull it out. 10 screws, 10 wires or so to desolder (do take pictures to remember where it all goes....) and there you go. Again I sucked each and every joint, and found ONE bad one. Looked it up in the schematic... hmmm.... interesting one : it comes from the Vref board under the chassis, and supplies the Vref that the ramp generator uses to charge its timing cap.... hmmm.... remember I was going crazy because my GOOD comparator board now would not generate a ramp any more ?! Fast forward.... YES, the board is 100% fine and it was this joint at fault !
One problem solved, I have now my good board back ! A reference board I can cherish and use to help me fix the bad board !
Power socket replacementOnce I had the back plane PCB, I now had good access to the power socket and could replace it.
I searched my box of 3 prong IEC sockets and found one that might do it.
Two mounting tabs on the sides, like the old obsolete socket.
Guess what ? Excellent surprise : the modern socket fits LIKE A GLOVE inside the existing hole / cut out in the PCB, and better yet, the two holes for the mounting screws.... EXACT same location !
Couldn't believe it. I was in heaven... zero trimming or drilling needed to be made to get the new socket to fit. It's plug and play.
Well I did have to buy new screws though, because the old ones had a flat head. Also, the tabs on the new socket are 3mm thick, versus super thin on the old socket. So I had to buy counter sunk screws, that are 3mm longer than the old ones. Old ones were 8mm long. So I needed 11mm or a bit more.
H/W store had 12mm long in stock, just perfect.
So I was very pleased indeed. My new socket is not perfect though : it's meant for PCB mounting, so right angled downwards.. the pins interfere with the chassis/deck. So I bent them backwards 90° to make them straight. Now they clear the deck, but not by much. So I added heat shrink tubing to all 3 pins, and also a bit of Kapton tape on the deck by the socket, just in case.
Will buy more sockets to do the other meter, and the frequency counter as well. Of course I will get proper terminals this time, exiting straight not angled.
While putting it together, I lengthened a few wires because they were too short. Some broke, some were too tight, not enough slack to let you put the PCB flat on the bench to work on it.
Of course I had to add an earth wire that did not exist before, since the old socket was using its metallic body itself as the earth "pin"... The mounting screws were pressing the body onto a wide track on the PCB, itself in contact with the cabinet once you put all the cabinet screws in place.
So I added an earth wire that I soldered to that track.
Overall result is very nice, am happy. Clean, no destruction / trimming / drilling involved... a painless upgrade.
Can now use a modern / standard power cable.
A modern cable, a working power switch... I am bordering luxury here !!!
Trouble shootingSo... I put it all back together, fired it up and resumed troubleshooting.
Result ? Free running still there... but not always, it's hit and miss. When you power it up, sometimes it's free running, and sometimes it displays stable and valid data. Swapped the good board... meter works fine, the trigger switches on the front panel as well.
OK will try to make it short, I am tired it's past midnight.
In short, been playing with the thing, scoping, and I think I am getting somewhere at last. It appears that there are two distinct issues combining together, which sure didn't help me get a clear picture of what was going on at first ....
So :
1) Trigger circuitry is now 100% kaput, no matter what you do you can't get this freaking board to generate a trigger signal to get the ramp generator to do its thing, hence of course it can't do any measurement. So I need to investigate that.
2) Second problem : when the counter is free running... it is now ACTUALLY free running : I can see the clock being pushed to the counter input.
3) When it does so, because there is no ramp going on, I probed the inputs of the gate circuitry, and hey presto I saw something wrong. Normally the two inputs (one for each comparator, Start / Stop signals), are both low when you are not asking for a measurement. I.e you set the trigger to "Manual", and you do not press the push-button.
Well...here, the two gate inputs are NOT both low. Only one is, the other is high. So... one of the comparators must be misbehaving ! So I probed a bit more upstream, closer to the relevant comparator and... yep, it's throwing garbage ! Not only is it high instead of low, but it's also sending pulses ! It's "oscillating" ! 10us pulses every 1ms or so. So I need to investigate that.
When this happens, I had the idea to ground the output of that comparator, to force a stable and correct logic level, so that the gate logic has valid inputs. As soon as I do that, instantly the meter stops free running ! I am on to something here it looks like !
Also, if I then remove my grounding lead... the comparator now stops misbehaving ! No pulsing anymore, and correct logic level !
I am not too clued as to how the comparator could throw pulses like that... any idea ?!
A dogdy transistor again ?!
I will try to now scope the guts of the comparator to try to pin point what part of its circuitry is producing these weird pulses...
So... making progress I think. Now have a plan of attack. Work on that dodgy oscillating comparator, and fix the non-working trigger circuitry.
No, no need to thank me, it's always a pleasure to deliver these updates to you all !