HP 6114A repair (fault: regulation seems to work, voltage low, output ripple)

[URI]

My HP 6114A arrived today.

That said I directly moved it from the cardbox package onto my bench.

After switching it from 115V~ to 230~ input I switched it on and it behaved just like the seller had described it:
Problem: It put out voltage some volts lower than the voltage dialed in even though the voltage and current regulation seem to be working basically. And it had 50mV ripple on the output.

Ok, first, I took it apart.   

Turned out that one rail for the control board (+28V) was at ca. 19V DC with 9V AC ripple on it. The brave 6114A regulated that down to 50mV ripple on the output.
The other rail (-28V) was fine: 28,1V and ca. 80mV ripple.

First I suspected one of the rectifier diodes (CR1 or CR2) to be dead -they all were intact.
After that I powered the faulty rail with 28V from my Oltronics B300D supply and -tada:
The 6114A came back to life and worked as it should -despite being completely decalibrated by the seller -he must have tried to solve the problem through recalibrating. I'm glad he didn't touch anything else inside.

With this finding my next suspect was the electrolytic capacitor for the +28V rail (C1).
Some further mechanical dismantling it turned out that it is completely dead:

I had checked the one for the negative rail and that had more than 800µF (490µF nominal value!).
Strange that one of two similar capacitors is completely dead while the other is as good as new. 

I only had a bunch of 220µF/400V capacitors available for testing so I bodged two of them in parallel into the 6114A:

That brought it back to life again. Winner! 
Some calibrating later (quiet easy with modern equipment like electronic loads and DMMs with >> 10megOhm input resistance  )
the 6114A was spot on again and also the hefty ripple was gonski.      



This repair was a quick hack because I couldn't wait.

I'll check all the other electrolytic capacitors as well as the tantalums and will replace some or all of them.
Don't know yet. Stay tuned.   

[BravoV]

Thanks for sharing 

Btw, I made a "stiched & colored"  schematic in single page for this PS, you can get it from -> HERE , beware, its quite big image.

[bd139]

Looking good. Those old capacitors are trouble. I’d do the lot if you can.

I have considered making adapter boards for them in the past but I usually crudely drill another hole in the board.

[wn1fju]

Keep an eye on the voltage selection switches.  Over the years they crud up inside and the result is considerable "noise" on the output.  There was a
tutorial (on eevblog) on how to disassemble the switches, clean them, and reassemble.  I had this problem with my similar HP 6115A and the cleaning
procedure cured all ills.

[URI]

Thanks for sharing 

Btw, I made a "stiched & colored"  schematic in single page for this PS, you can get it from -> HERE , beware, its quite big image.

Ah, thank you!
I had that already as a printout together with a copy of the manual and I got it from the seller together with the 6114A. 
Was very helpful, indeed.   

I also read robrenz's restauration report:
https://www.eevblog.com/forum/testgear/hp-6114a-precision-power-supply-restoration/

..and this thread (not related to the problem I had with my 6114A but interesting, too.):
https://www.eevblog.com/forum/testgear/hp-6114a-help-needed/

..and this report with lots of different faults repaired:
https://www.amplifier.cd/Test_Equipment/Hewlett_Packard/HP_other/6114A/6114A.html
(Interesting but written in German with shorter English translated parts. No problem for me, but..)

[URI]

Keep an eye on the voltage selection switches.  Over the years they crud up inside and the result is considerable "noise" on the output.  There was a tutorial (on eevblog) on how to disassemble the switches, clean them, and reassemble.  I had this problem with my similar HP 6115A and the cleaning procedure cured all ills.

Yes, thanks, I read this tutorial by robrenz already (link in my last posting above).
It looks like I don't have to go so far by now, the unit behaves well and was in a far better (cleaner) condition than robrenz's.
But I'll keep an eye on it.
First I have to get some spare electrolytic and tantalum capacitors to replace the old ones.

[URI]

I'm checking the capacitors currently and am facing some surprising things.
Disclaimer: This is true for my 6114A with a serial # beginning with 1928A. 6114A's with different serial # beginnings might be (as expected) different.

First of all, I had found capacitor C1 completely dead causing the 6114A to not regulate the output right and with a ripple completely out of Spec (see first posting of this thread).

I checked C2, C3 and C12 now (really hard to get them soldered out the PCB even with a desoldering gun..) and found C12 completely dead also (120pF left over from 490µF).
Having all big caps out I decided to bodge in capacitors like I did it before for C1 temporarily.
I'm keeping the good C2 and C3 for spare parts. Could be another 6114A or a 6104A be hitting my Test Equipment stack. 
Upcoming TEA once more..   

Inside of my 6114A it's now (temporarily) looking like this:

The failed capacitor C12 (with a 0.25Ohm series resistor) is connected between terminal A8 and -S(ense).
With A8 connected with a bridge to + and -Sense to - normally C12 is connected to the +/- output of the power supply.

In the manual it's function is described (error corrected..):
3-59 An internal capacitor (CI2) connected across
the output terminals of the power supply. helps to
supply high-current pulses of short duration during
constant voltage operation. To reduce current
surges. this capacitor can be removed by unstrap-
ping terminal A7 A8. Any capacitance added external-
ly will improve the pulse current capability. but
will decrease the safety provided by the constant
current circuit. A high-current pulse may damage
load components before the average regulator cur-
rent is large enough to cause the constant current
circuit to operate.

A failed capacitor C12 is therefore nothing else then an open connection between A8 and + output. 

After I had replaced C1 (and with it repaired the main fault of my 6114A) there was still a crackling sound (50Hz) noticeable that was really annoying to me. After some exploring I found this guy across transformer terminals 12 and 14 making this annoying sound:


It's a foil capacitor 1µF 220V~ still measuring 923nF, 0.1% loss, 0,14Ohm ESR.
It's not getting hot, not even warm but I'll replace it anyway to get rid of that crackling sound.
Capacitors of this kind should be absolutely silent.
The legs were wound around the transformers solder terminals and it was a little difficult to desolder it before I found out. 
The smallest replacement I found is slightly bigger (31,5mm x 11mm x 21,0mm).
I wonder how Sprague managed to build this capacitor in such a small form factor. Perhaps that's the reason why it makes a crackling sound now. 

Then I found a capacitor C1 on the A5 interconnection board that's not mentioned anywhere, not even in the change/errata notes I got with the manual from ArtekMedia:

(one pin desoldered for measuring purposes)

C1 on the A5 board is connected directly to + and - output. And it's nowhere mentioned/shown in the schematic nor the changes/errata section.
I discovered that in my 6114A C2 is not connected to the A3 terminal but to the A1 terminal.
Last but not least the output capacitor soldered directly over the front output terminals is C3 in the parts list but C13 in the schematic. I corrected the schematic for my 6114A:

[URI]

Hmm. I got a little stuck and did only do few things related to my 6114A.
I nearly measured all caps and began to search for replacement parts. The parts list isn't complete though.

I bought some surplus parts (see my posting https://www.eevblog.com/forum/chat/what-did-you-buy-today-post-your-latest-purchase!/msg1461785/#msg1461785).

I found a 1µF foil cap suitable to replace that foil cap making the crackling sound.
With 450V~ it's a bit overdesigned but I like it:
With its yellow foil with the caps data on it looks like the big brother of the cap it's replacing. 


The cross section of the two caps (old one left..):


The new one is slightly bigger.   
I'm glad I found something looking kind of original and wasn't forced to put a (kind of) giant WIHAWIMA foil cap in there. 

With the surplus parts EPCOS 470µF/100V electrolytic caps hit my bench. Same diameter like the old ones but only 2/3 tall.
The time stamp (04/2004) shows why they were surplus parts..
Nevertheless I selected four with 470µF to 480µF (the worst had 468µF  ) that I will fit in place for the 490µF ones (Board A1: C1, C2, C3,C12).
470µF is well within the tolerance of the original 490µF ones -they will all do with their ESR smaller than 60mOhm.
The old ones that measured 580µF to 760µF have an ESR >200mOhm.

[bd139]

Looking nice  Thanks for posting updates - interesting to read.

The Sprague 492P caps were early polyester film caps. It was crackling probably because it was conducting through some broken down film. They suddenly explode so it was a good job to replace it. Usually they replaced them with much smaller X2 class capacitors if they are at line voltage or so. Same thing, just as flammable, well the RIFA ones were anyway

I always go the Heathkit route these days. I have some 2Kv Mallory-RMC ceramic discs I got off eBay which are 1970s dated. They never seem to go wrong or explode!

[URI]

I decided to replace the old filter caps with NOS electrolytic caps with 470µF/100V in March. The project paused a little while..

When I came back to it two weeks ago I decided against those NOS caps (all measured >>470µF).

The four filter caps in the 6114A with 490µF nominal value measured 580µF to 760µF with an ESR >200mOhm.
To replace those I didn't take 560µF (nominal) caps.

Instead I ordered new 820µF/200V electrolytic caps.   



Measured them with my HP 4263A at 100Hz and these 820µF caps all measure between 720µF to 740µF.   
Ok, this is well within their tolerance (+/- 20% - 656µF to 984µF).

But I do remember times when fresh electrolytic caps had a capacity of >10% regularly and therefore had a good reserve capacity for many years of circuit life to come naturally loosing their capacity over time.
Back in those days a fresh electrolytic capacitor with -10% would have been discarded for production.

The Times they are a-changing...

Nowadays electrolytic caps seem to be built to price only.
If you want a little reserve capacity for a long working life of the circuit you are building take a value one or two steps up.      

[URI]

Looking nice  Thanks for posting updates - interesting to read.

Thanks!   

The Sprague 492P caps were early polyester film caps. It was crackling probably because it was conducting through some broken down film. They suddenly explode so it was a good job to replace it. Usually they replaced them with much smaller X2 class capacitors if they are at line voltage or so. Same thing, just as flammable, well the RIFA ones were anyway

I checked the old  with polyester film cap my LCR meter (HP 4263A): It nearly has its nominal value (less than -5% ) but its serial resistance measured >2Ohms. Bad. I'm glad I removed this ticking time bomb.   
I'll do the same with my second 6114A I acquired some weeks ago. 

I always go the Heathkit route these days. I have some 2Kv Mallory-RMC ceramic discs I got off eBay which are 1970s dated. They never seem to go wrong or explode!

Yep, some of the ol' shit is really built to last.   

[URI]

I discarded some ideas how to replace the filter caps of my 6114A now and decided to retrofit new caps inside the aluminium cups of the original but defective caps.

For a while I thought it would be easiest to do an adapter board that adapts the pinout of the original caps to the pinout of the replacement caps. I required surplus caps that were ideal for a replacement in my opinion then. I changed my mind on that.
The three-pin pinout of the replacement caps is a PITA:
Pinout (sorry for my camera not focusing on the pins in the upper half.. ) and prototype adapter board:


As you can see I really had to fight to put the two groups of pins into that prototype adapter board.
I would have loved to do this in KiCad or the like but 1. I don't have any experience working with KiCad and 2. I didn't want to buy a stupidly licensed Eagle that I used years ago when I could really buy it and 3. I think I would have had to define both layouts myself -PITA again..   

So I did the next step: retrofitting new caps in old cap cups.   

[URI]

Retrofitting new caps into cups of old caps:

First I dismantled the old cap. I hadn't come across sites on the web where this process was already described when I started my first attempts. I went down one dead end street by trying to drill the interior to get it out. That didn't work despite leaving the drill bit covered with a viscous black coat. Uuhrrghh.   

Finally I opened the cup by bending back the border at the bottom side of the aluminum cup.
This is what was left of the old caps interior (metal ring with ground pins not shown) besides the cup:


I did a little jigsaw job copying the original (but drilled) hard paper plate. I isolated the area around the new positive pin from the rest of the copper area of the new plate. To replace the lost positive pin I cut one pin off a power diode, put it through the pin hole of the new plate and soldered it.


I connected the metal ring with the negative pins by drilling two holes and putting in short pieces of silver wire and soldering those to the negative copper plane on the back. After that I connected the new plate with the new cap using short pieces of wire:


To put the new cap into the original caps cup I had to add two old pieces: A transparent plastic plate and the former seal. This is important: The stack of these pieces fills the thicker end of the cup. For re-bending the border on closing the cup again this is crucial.


After cleaning the inside of the cup I put the new cap in and closed the cup by re-bending the border. This is the result, next to an original cap:


Bottom view of original cap and retrofitted cap. The bent and again re-bent border of the aluminum cup doesn't really look good but when it's put back on the PCB it's nearly invisible.


I retrofitted a small 470µF cap as my prototype first; I'll use it for cap C12 -that's the output capacitor that can be unconnected by removing bridges on the back of the 6114A. For the filter caps C1..C3 I'll use the 820µF electrolytic caps I bought some weeks ago (see posting above). These have the same cross-section as the retrofitted cap shown above.

[Tomorokoshi]

Looks good as a practical solution when faced with non-ideal parts.

One suggestion: label the outside of the retrofit capacitor with the markings of the replacement internal capacitor, or at least with a message about it having an internal replacement.

[Specmaster]

Nice work but personally I'd have just installed the new cap as it was and not bothered about restuffing the old cans.

[URI]

Looks good as a practical solution when faced with non-ideal parts.

One suggestion: label the outside of the retrofit capacitor with the markings of the replacement internal capacitor, or at least with a message about it having an internal replacement.

Thanks!   
Fixed that:

Nice work but personally I'd have just installed the new cap as it was and not bothered about restuffing the old cans.

If it was such as easy... 
The layout to fit those original caps has three pins for minus arranged on a circle with the pin for plus located half way between two of them. There's no new (modern) replacement cap that fits both these holes and the outline of the original cap on the PCB the same time. That's the reason why I first wanted to do an adapter. I also liked the idea of not changing the appearance of the interior too much.   

[URI]

Retrofitting new caps into cups of old caps pt. 2.1:

This time reusing all parts of the original enclosure.
The first image shows the bottom of the canned cap. After pulling out the innards there are: The can, the cap itself, a spacer/insulator plate, a seal, the top cover plate with the positive solder pin and the a ring with three negative solder pins.


The seal is obviously cast at production inside the cap and clearly shows casting fins that have to be removed. This doesn't have to win a price..


I cleaned the inside of the aluminium can from a sticky, tarry material. Before, after and then all cap innards removed:


To be continued..

[URI]

Retrofitting new caps into cups of old caps pt. 2.2:

After cleaning the inside of the aluminium can the new cap fits smoothly into it.
Now the electrical connections and the mechanical retro-fitting have to be prepared. The positive solder pin can stay where it is, it'll be reused. The solder pins conflict with the isolation plate and the seal, therefore two holes are drilled into them. After that, these two pieces fit onto the new caps solder pins. For connecting the new cap to the outside electrically the negative solder pin gets extended with a piece of silver-coated copper wire. The tip of the positive solder pin is bent to be soldered to the original positive solder pin.


To be continued...

[URI]

Retrofitting new caps into cups of old caps pt. 2.3:

With the new cap soldered to the original positive solder pin sitting in the original cover plate, the new cap is ready to be re-fitted. The first photo shows just this. Before sealing the aluminium can again the ring with the three negative solder pins is added and the stack of the new innards is pressed down to fit against the step of the aluminium can. The sealing of the can is done by crimping the border of the aluminium can down again. First by hand using flat pliers and in a second step with a hammer and something I know as a "Nagelversenker" babbel-translated a "nail countersink". The foil covering the aluminium can is a self-shrinking one and after using a torch it looks much better. As a last step the extension of the negative pin of the new cap is cut and soldered to one pin of the ring of solder pins. The check using my chinesium LCR/Transistor/Diode-Tester was positive on all three retrofitted caps. 



To be continued...

[URI]

Retrofitting new caps into cups of old caps pt. 2.4:

Labeling is the last step, not to be forgotten!     


The way described in the three last postings I finished all three caps left. Now all four are ready to be re-fitted into the 6114A.
From above (as seen through the vent holes of the housing) everything looks untouched.   


Seen from the side the bottom border of the cans is not completely smooth, the bending-open and close again has left some dents and noses. But it's marginal to me, I can live with that. 


end.

[URI]

Re-fitting the filter caps of my 6114A I came across something that surprised me.

One of the ripped-out foil-cap-windings made of aluminium foil and a kind of paper still measured correctly as a capacitor with it's value that it had before being ripped out:


The upper end looks as if there must be a short somewhere -but there isn't one obviously.

[bd139]

Those restuffs look pretty excellent.

Much better than the shitty job I do 

[URI]

Thanks!

Once being caught it's hard for me to do less.      

[Specmaster]

Nice but TBH I'd have just installed standard radial ones in their place, not bothered about keeping internals looking original, more about lets get it working and usable.

[URI]

Still to do:

1. recapping those tantalums and checking all other electrolytic capacitors (list of components in the service manual is already coloured for that purpose)
2. going through all changes mentioned in the service manual I got from artek media. There are some to the schematic..
3. repeating all this with my second 6114A..