Another Fluke 5440B/AF Repair

[essele]

After a break from all of this volt-nuttery I got a bit carried away on eBay and got myself a really beaten up 5440B ... it's definitely time for a project!

It's in a bit of a state ... it looks like it may have some water damage (or at least corrosion due to high humidity), plus there's a huge amount of grey dust in there, not sure if this is just dust or whether it's the result of some burning -- I haven't inspected many of the boards yet.

I've only done an initial power on test and it powers up, there was something on the display, but the fan wasn't running and there was a noticeable whine which sounded like it was coming from the main transformer (is that normal?) ... so I switched it off again pretty quickly and decided I should clean it up and at least replace the fan before going much further.

To get at the fan (and mains filter) I've removed the back panel and the A19 Outside Guard Terminator PCA (which seems to be just a set of unregulated power supplies) ... on the board there was an area which was quite darkened, mostly around CR2, CR3, and F2. Some tracks on both sides were missing their solder mask and I think the plastic coating that would have been on CR2/CR3 appears also to have burnt off! F3 had some damage to the sheath underneath where it was close to a track which appears to have overheated.

All of the fuses are intact and don't look like they have been replaced (they do now because I messed with them to apply some new solder mask underneath -- and looking at the photo I need to redo my soldering!)

So I am a little confused ... CR2 and CR3 feed the unregulated +5V supply, both are fused with 5A fuses, both diodes appear to be ok. However something has caused them and the PCB tracks to get very hot but it hasn't blown the fuses! R2 measures fine, and C2 seems to be ok from an external timed charging test.

I'll obviously check the voltages once I get it all back into the chassis.

Any thoughts?  Is this just to be expected with age? Should I replace CR2/3 and the fuses?

This is probably the simplest board in there --- this is going to be a long journey ;-)

Thanks,

Lee.

[dietert1]

That design is from a past when engineers thought human ingeniuty could surpass physical laws. Those diodes have no other cooling than the board. They last but the board won't, pretty obvious. I would remove the diodes and replace them by shottkys in TO220 with a clip-on cooler. Temperature will be a lot lower.

Regards, Dieter

[essele]

Thanks Dieter. I was thinking about excess current heating the traces rather than simple heat transfer from the diodes -- doh!

I've done a quick test, at 3A those diodes have a Vf of just over 1V, some others (schottky) I have to hand are down at 0.8V, and I've ordered a couple that should be under 0.6V .. they should bring the heat generated down fairly significantly without me trying to bend TO220 leads and fit heatsinks in.

I've cleaned up the the boards in the rear section as well as the two filter boards. I've fitted a new fan (along with a 3d printed fixing bracket) and mains filter. I'm waiting for an inline connector before I can connect the fan.

I have powered it up again ... the front display just says "Fluke DV Calibrator" and "Front Panel Running", the cursors seem to work but no other key does anything. There also seems to be a somewhat inconsistent set of red LEDs lighting up on the various boards.

A first batch of caps ordered, and on to check voltages...

[MegaVolt]

You have to make sure that all 4 diodes on the A16 board go out one by one. If this is not the case there is an error on board A16. Until you fix this there will be no communication with the front panel.

[essele]

Thanks MegaVolt ... that's helpful.

I've checked the outside and inside power supplies and most are ok. The outside +28V is actually +24V but I'm going to hold off adjusting anything until I've recapped that PCB.

The inside regulator has two failed rails TP7/8 +30V (measured -0.9) and TP11/13 -15V (measured +0.9) ... a quick look makes me suspect the LM317 and 7915 respectively (although I'm not sure I understand how the polarity could be wrong), the caps look ok, although I'll replace all of them once new ones arrive. I doubt any of that is related to the A16 issues you mention.

Quite worryingly I get different behaviours each time I turn it on ... sometimes all 4 LEDs go out and stay that way (still with the same front panel behaviour), sometimes it does that but then seems to get into a reboot cycle every 5 seconds or so, other times different LEDs don't go out.

I'll get all the power rails sorted and some of the recapping done and see where that gets me.

[MegaVolt]

Quite worryingly I get different behaviours each time I turn it on ... sometimes all 4 LEDs go out and stay that way (still with the same front panel behaviour), sometimes it does that but then seems to get into a reboot cycle every 5 seconds or so, other times different LEDs don't go out.

The A16 board has a watchdog that receives reset signals from the A2 board. If for some reason these signals do not come, the board A16 resets again.

Also pay attention to the LED on the A2 board it goes out after the internal checks of this board are passed.

For a successful operation of the digital part you may remove A4...A10 and A14. Everything should work and show a communication error with board A14.

[essele]

Thanks again MegaVolt -- really helpful, and I now have some good progress!

I re-examined each of the digital boards and found quite a crusty almost calcium like build-up on many of the fingers on the edge connectors, so a careful scraping and cleaning of each of them and I now have some progress ... it seems consistent, the LEDs go off, the front panel now says "power supply fault", "check guarded power" ... and I already know I have two rails on that with problems ... so this is great!

Thanks!

[MegaVolt]

"power supply fault", "check guarded power" ... and I already know I have two rails on that with problems ... so this is great!

This error is for the A10 board. It must be inserted back into the unit.

[essele]

This error is for the A10 board. It must be inserted back into the unit.

Yes, thanks, it is back in ... and it's the one with the two failed rails, I'm just waiting for an LM317 as I don't seem to have any.

[MegaVolt]

Yes, thanks, it is back in ... and it's the one with the two failed rails, I'm just waiting for an LM317 as I don't seem to have any.

That's all that could be done so far. We have to wait

[essele]

OK .. some great progress.

I replaced the LM317 and associated caps, however the behaviour seemed the same ... more investigation revealed that with no other cards in both "failed" rails seemed to be ok, as soon as you put any other card in they dropped to around 1V. I couldn't find any obvious shorts.

It turned out to be related to the crap on the card edges again ... a good clean of the inside regulator and the +30V rail was all ok, not sure I actually needed to replace the LM317, but no harm done. After cleaning the -15V rail was now at -20V and causing the error LED to remain lit (I did a load of comparator and logic probing before I went back and remeasured the voltages) ... so I replaced the 7915 and 22uF cap and it came good, the LED went out, and front panel was giving board errors (because they weren't in there!) ... so power problems seemingly solved. (Still a load of re-capping to do!)

The next error was "board ack fault", "check output board" ... so I cleaned the edge connector for A4 which (a) needed it, but (b) was the wrong board ... should have RTFM and cleaned A5. Once A5 was cleaned up I get the "VOLTAGE MODE" message and it seems to be getting through the power on sequence ok!

I tried a 1V output and got 34V (I think from memory) quickly followed by an over-voltage warning. But I still have 4 cards to properly edge-clean so I'll wait to do that before going much further.

There were a few relay clicks, which from how they look is a miracle ... I think they will need quite a bit of attention.

Thanks for the help so far ... assuming I can get this back to a working state does anyone have any ideas about replacement key caps?

[essele]

I think I'm going to have a major problem with the relays ... they all look badly rusted.

I've just opened one and cleaned off the worst of it with IPA, it all moves ok, but there's a lot of corrosion I just can't see these all working to a satisfactory level and I know proper replacements are prohibitively expensive.

Is there a reason that immersing in IPA is not recommended? I think that doing that followed by a careful contact cleaning is my best bet at this point?

Otherwise I'm going to be looking at replacing these with custom PCBs and lower cost alternatives (I certainly don't have the patience to do all the wiring detailed in the link below!)

I have seen both of these posts which are helpful...

https://www.eevblog.com/forum/metrology/replacing-relays-in-a-5440b-or-5440baf/
https://www.eevblog.com/forum/metrology/fluke-5440b-_repair_/

[essele]

When I said "it all moves ok" that was from manually prodding, when powered it was only moving a small distance and not properly making contact. I've had to scrape a load of crap out from between the magnet and the plate that it attracts (not sure of the proper terms) and now it seems to be properly making contact.

Everything is covered with that brown dust so I really feel like I need to soak these in IPA to have any chance of getting rid of it.

Should I just give up and find another solution or are these salvageable?

[dietert1]

I suspect the dirt you see is from the previous use of some kind of deoxit.
Don't get desperate, i've seen worse. You can certainly salvage the relays, but i would try to keep the coils dry, as the wire is really thin and you would not like a solvent cause shorts.
As far as i remember the contacts can be unscrewed (after taking some pictures!!). Then use a toothbrush with soap. When everything is clean and dry again, it should work.
In the future, don't use deoxit, but some software to exercise the relays regularly. Same when beginning measurements on the Advantest R6581T. For 10 V DC measurements, as far as i understand it has three relays that need exercise: One for front/rear, one for high voltage ranges and one for ohms ranges.

Regards, Dieter

[MadTux]

Is that the one that was on ebay for $699??
Kinda lucky I didn't pull the trigger on that one, corrosion/water/fire/smoke damage are the worst.

I'd fully disassemble that instrument and put all the parts into the dishwasher, except the transformer and maybe the reference board, the LM399 inside my Keithley2010 didn't care about the dishwasher though, still agrees to my other K2015/HP3456As within a few ppm.
Then use compressed air to blast away all moisture directly after dishwashering, so a minimum of salts remain on the boards, at least that's what I usually do in such cases.

Relays actually look like they've spent some time partially filled with water, looks a bit like first smoked by fire, finished with water from firefighting, discarded, salvaged/dumpster dived and put back on ebay :-)

[essele]

Hi MadTux ... yes, although I got it a bit cheaper ... he was willing to accept a fairly low offer, makes sense now as I think your summary of fire/firefighting/resale is about right!

I've managed to completely wreck that first relay ... got it working reasonably well with one pole higher resistance than I'd like and one not working, but then tried the "removing the screw" option after which I found it completely impossible to reassemble in a working way and eventually broke the coil terminations.

That said, even with a missing relay, it's partially working ... output above 22V seems to be fine ... checked a few spots between 25V and 400V and all seem ok.

I can't believe many of those relays are going to be functioning properly (although some of them must be at least partially working!) ... so I think I'm going to have to do something with them first before trying to diagnose much further. I've designed a PCB that should allow me to mount three DS2E-S-5VDC's (or similar) in the space of the normal 6PDT relay (using the same approach as @wn1fju) but without the painful wiring. If all goes to plan these should fit into the existing sockets. If that works I can also look at the other configurations. I'll slowly work through recapping while I wait for the PCBs.

I'm quite nervous about the dishwasher option, any gotcha's I should be worried about (apart from the wife moaning)

[Kleinstein]

The normal detergent used in the dishwashers is quite aggressive to aluminum. So it would not be a good idea to use it with the AL electrolytic caps.
It also takes some time to get the dishwasher clean afterwards - there often remains some of the dirty water in the sump. So do some rinse with extra water and pump empty from the machine.

One may as well to a manual wash in the sink with warm water,  more normal detergent and a brush. Then follow with quite some clean water and air to blow dry and maybe than some IPA. After a water damage cleaning with clean water is definitely a good ideal, as some of the dirt can be water soluble and not alcohol soluble. Some parts also may need just the mechanical action.

[TheSteve]

Nothing causes more grief in a 5440B than those relays, even when they look pristine. I would plan to replace them all.

[essele]

Nothing causes more grief in a 5440B than those relays, even when they look pristine. I would plan to replace them all.

I have seen quite a few painful stories about the relays ... if my PCB/off-the-shelf relay approach works then I'll plan on doing a complete replacement ala @wn1fju.

In the meantime to allow me to get a bit further with fault diagnosis I've figured out a reasonable routine to get the existing relays cleaned up a bit, I'm breaking all the rules, but it's a choice between just throwing them away or having a go!

Bear in mind that almost all of the relays I've looked at so far are not even able to move enough to make physical contact when applying 5V, let alone low resistance electrical contact.

So I give them a 2 minute clean in my ultrasonic cleaner but in a small container of IPA ... this is generally good enough to start them properly moving again. At this point there are usually some good contacts and some bad ones.

I gently scrape (caress?) the relevant contacts with a scalpel (although I seem to spend more time cutting myself at the moment) and repeat until I get sub 1R resistance on everything.

Then I setup my PSU to do 3 second on/off cycles and monitor the resistance as I've noticed a few creep back up again if they aren't properly clean.

If all else fails then a squirt of contact cleaner and another 2 mins in the ultrasonic/IPA seems to do the job. It's a bit painstaking but it seems to be getting there.

Probably not a good thing to do to ensure long life, but as per the above these things are a nightmare anyway!

Photo attached ... three done so far!

[Kleinstein]

With the relays it is not just about getting good contract, but also possible problems with leakage currents. So a cleaning step with clean / demineralized water at the start may be a good idea to remove salts.

To protect the iron parts one may consider wetting the iron part at the magnet with some oil. A fully degreesed iron core may start to corrode again rather quick.

[wn1fju]

I am the poor soul who replaced all of the relays a few years ago in my 5440B with a set of 62 DS2E miniature sealed relays.  I am happy to report that the 5440B is still working just fine.  And the best part is that I don't hear big "clunks" when changing 5440B settings.  All I hear are very tiny muffled "clicks."

With the original relays, I tried all kinds of cleaning fluids.  I also tried a "mechanical" cleaning with a burnishing file.  I also tried "electrical" cleaning with a few hundred mA across the contacts.  Although things would improve for a few days, the problems would always come back.

I also have a Rotek 320 multi-function calibrator.  Same problem, same attempt at repair, and finally, the same solution - replace all (21) relays.

So it has been my experience that once calibrator relays start to degrade, they really can't be brought back to acceptable functionality.  Perhaps others have been successful, but I haven't.

[essele]

One step forward and one step back .. I ran the self tests, the digital one passed, but the analog and HV ones both failed, however I think the HV one has done some further damage as I now can't get any good output on any ranges.

I've finished cleaning the relays and think they are working ok for now, however I'm seeing consistent problems even in standby so I don't think that's relay related.

The DAC output is fine, however the pre-amp output isn't, it's pegged at the negative rail (and then clamps by the zeners on the output PCA) which then also causes U7 to get very hot.

I've managed to get the preamp up an running on the bench and I see similar behaviour (I've added my own feedback path from U7 back into a dummy PREAMP-SJ) and I'm just working through comparing results to a falstad simulation.

I am totally confused by the different reference points (A and R on this board) and how they relate. They are tied together by CR19 and CR18 which I assume allows for some variations without letting them get too far apart (not sure of the reasoning behind this!)

One thing I'm noticing which I find very strange is that if I measure the voltage drop across CR18 or CR19 I see different results depending which way around I use the test meter leads ... for example with a 25mV test input signal (to PREAMP-SJ) I see 115mV drop in one direction and 9mV drop in the other ... how can that be?

(Also, BTW the schematic is wrong ... CR18 and CR19 do NOT connect to the +17 rail, they actually connect to S COM which seems more sensible and wasted many hours trying to get the simulation to work!)

Any help appreciated ... I'm well out of my depth on this!

Lee.

[dietert1]

The diodes are a protection in case of missing ground connection. If all ground connections are normal, the voltage difference should be some mV at most, ideally below 0.1 mV. The connection should be near the output terminals, maybe there is another relay or a jumper for 2-wire/4-wire mode.

Regards, Dieter

[essele]

Thanks Dieter, that helped simplify my testing.

So good progress now ... discovered that the UA714's were not a suitable replacement for the LF356's (U6 and U7) ... with proper parts in place I'm now getting good results on everything up to 22V, after that it goes mad and I see some quite alarming voltages on the output.

So it looks like an HV issue ... onwards and upward!

[essele]

So it was an HV loop problem ... caused by K1 on the pre-amp PCA which although working fine a couple of days ago again wasn't playing ball ... so I completely agree with wn1fju that these relays really do need replacing! (My first batch of PCB's should arrive tomorrow for the 6 pole variety, if they seem ok then I'll get the other variations designed and ordered.)

Once that was 'fixed' I now get output above 22V, however it's low ... a quick run of the analog self test and you can see the issue, test M34 should be +20V (on the 250V range) and I get 12.625V...

Code: [Select]

[...]
     M31: -.093960V
     M32: +19.8853V
     M33: -.000760V
     M34: +12.6588V

END OF ANALOG DIAGNOSTICS
I'm assuming there is another relay somewhere messing up the loop path ... I seem to be getting around 30% of the requested value on the 275V range and 60% of the value on the 1100V range...

100V => 34V
200V => 67V
300V => 180V
400V => 290V
600V => 360V

... so I'm going to go through and check all the relays that should be on for both ranges ... there is the possibility that another relay is stuck on when it shouldn't be, but I've not seen that at all as I've gone through testing them, so seems a much less likely scenario.