HP6268B Restoration

[AMR Labs]

I got a HP6268B off the e place and it was listed as untested. In the few published pictures it looked like it was in decent shape, and everything there that should be at least externally. I sent in a low offer and it was accepted, best of all shipping had been listed as free. Or was it...

As you can guess that free shipping option was probably not a good idea after the seller found out what it would cost him to ship 80+ pounds. So he put it inside a flimsy cardboard box and a little cushioning on top and bottom of the box, but really very little on the sides. Needless to say it got damaged along the way. Probably some pissed off UPS worker that really was unable to handle the box must have dropped from at least from 3ft. Can't really blame this person, as the box which was barely larger than the power supply itself, no handles and poorly packed, was really unmanageable. During the fall the sharp edges of the front panel chewed its way out of the box so what I received was a big mess. Front panel bent at two corners, one handle bent slightly sideways , plus various dents, scratches and bent metal at several other areas. The rear junction block where the remote programming and sensing wires connect is bent upwards, but luckily nothing broken nor has the PC board suffered any damage. Fortunately the meters where still intact, and the front knobs and more importantly the controls seemed to be ok. Two of the knobs seem to have lost its cap but that was the way it was shown in the listing, but the two other knobs survived.

There was also some sort of rattling like loose parts towards the rear when the equipment was turned over, but I could not see anything moving around or make it fall out. This turned out to be inside the red RFI assembly box, where 3 out of 4 spacers where broken off and there where some loose screws as well, along with various washers. So far found everything except for the last screw that holds that cover in place. Out of the 4 screws that hold that cover 2 where missing. Its like some hamfisted dolt forced his way in there and in the process broke several threads made of soft aluminum, just to then hastily threw everything back together. This could not have happened due to the recent tumbles in shipment.

I noticed that one of the broken off spacers is what actually completes the connection between one of the inbound AC sides and the Triac that is part of the preregulator, so this power supply could have never worked this way, I wonder if this might have been be the problem and the reason it was listed as "untested". I checked the Triac and it is ok. On the board inside the preregulator red box there are 3 RIFA caps which looked very crazed, one even having a gash on the side.

First thing was to re-strap the power supply for 115V for testing, as I don't have 230V at my bench. this involved reconfiguring both transformers T1 the (huge!) main and T2 the bias transformers. Also two jumpers need to be added and one removed, plus a 390 ohm 2W resistor needs to be added as well on the RFI assembly board.

Then tested the power supply by substituting the triac with a light bulb to prevent any big surges or possibly any shorts from causing fireworks, and it basically came up and regulation seemed was working fine as I was able to vary the voltage from zero to almost up to 40V. I was thrown off for a while and stuck at zero output because had not noticed that both current controls where set to minimum, which basically prevents the supply to provide any output voltage at all. So this fixed that, but then the overvoltage started triggering, and again by just setting the front panel adjustment to fully CW solved that. I was even able to connect a 12V car bulb that draws 0.5Amp and it was holding fine, but anything above 1amp caused the voltage to slowly fall back. Given that the whole preregulator is out of the loop, and the light bulb in place of the triac was only allowing about 75V AC to reach the transformers, that should be normal for now.

So needed to order the replacement for the 3 RIFAS, and also all the electrolytic caps on the A1 main board. One of the bigger caps (orange one) on that board clearly has outgassed on one side, but no signs of corrosion. There are in total perhaps under a dozen caps on the board total that will be replaced. They all look to be from the 80s given their size to value ratio.

Some nudie pictures attached.

[AMR Labs]

Some preregulator pictures inside the RFI assembly A2 board, as found with rattles inside.
Crazy crazed Rifas and broken off spacer threads.

[wn1fju]

It's amazing just how much damage can be done by mechanical shock.  I received a piece where the transformer broke free and tumbled around the unit ripping away some components that happened to stick up out of the board.  I've gotten pieces where large electrolytic capacitors have ripped out of the board.  Relays in a socket without a hold-down strap?  No way they stay put.  I even had one piece, which consisted of a shallow rectangular box with about four large PC boards mounted on standoffs, where every single board had ripped off their standoffs.  Now I can't know for sure whether any of these faults existed before the seller put the thing in the shipping box, but I kind of doubt it. 

Not to mention the FedEx driver who rather than walk the extra 10 feet to my front porch, is fond of placing small packages right in back of my car (where I don't usually look) maybe in the hopes that I will back out and run the packages over!

[AMR Labs]

Crazed RIFA caps have been replaced in the RFI assembly preregulator, and 115V conversion steps had all been completed. Also replaced all the electrolytics on the A1 main board. In the rush to order parts forgot to include the 390 ohm 2W resistor that is part of the 115V conversion, but was able to put together a temp replacement combination with what I had.

Otherwise all went together as planned. Had to disconnect the power to the bigger fan attached to the heatsink assembly, the constant noise was driving me crazy. I will implement a switch probably on the back panel to turn it on only when really needed under high power long term requirements. On normal short term or lighter loads the power supply stays cool as a cucumber. Its absolute madness to have those fans running at full speed all the time for hobby lab use. I am thinking of doing it for both fans on one switch. The smaller fan actually just sucks air into the cabinet towards the general direction of the main transformer which is some distance away anyhow, and runs constantly as well. Not needed most of the time I expect, and it will save some dust accumulation and my nerves. Worst comes to worst I can reverse the flow of the smaller fan to extract air and leave it running as it is much less annoying.

Also pulled out the whole panel to try and straighten out the bent corners and give it a good clean, and set the voltmeter pointer that needed a tiny tweak back to exact zero and the adjustment screw is hidden behind the panel anyhow. Given I don't have access to a 10 ton hydraulic press I think the corners look quite acceptable for now.

Found another something rattling around somewhere under the heatsink assembly so had to remove it. It was a royal pain in the posterior the get this to come out as there are clearance issues on almost all sides. Manual says to just remove two screws that hold it to the chassis on one end, slide forward to disengage two additional holding pins and then bring out by "sliding heatsink down about 1/2" and slightly away from chassis". Come again, say what? I spent almost an hour but various things would still hit the chassis or other areas, and the fan on the end of the assembly was already at its end of travel as it was bumping into the output terminals just in front of it, while the other end where the two screws where removed was still half way under the edge of the chassis. In the end decided to just remove the rear panel by taking off 3 screws on each side and then things started to happen much easier. I think HP forgot to mention this step in the instructions, or the Marquis de Sade got involved. So worth noting if you ever encounter yourself trying to pull out one of these heatsink assemblies. In the end the rattling was yet one more loose screw from the RFI assembly red box.

So was doing some load testing with my HP6060B electronic load. Everything checked out nicely, the power supply is super stable, practically no drift that I can detect on a 4.5 digit DMM once you set the output voltage. Both analog meters are dead on accurate, that is as accurate as a needle on a scale can be. Tested the supply under various loads, starting with 5amps, then went to 10, 20, 25, and finally 30 amps. Absolutely rock solid. The voltage actually raises a bit under heavy load by about 50-80mV. For example if I set it to 12.00 volts and apply a load of 25amps, the voltage raises to 12.08V or thereabouts (DMM always right at the output terminals). About the same as well on 10-20 amps loading. I did not see any adjustment in the service manual calibration section that might be able to compensate for this, but I guess it might be normal within spec? Will have to check the load regulation specifications.

That's it for now.

[factory]

In case you missed it, for 115V operation the mains circuit breaker is replaced with 30A one, 230V operation has a 20A one.
And two of the original capacitors were type 150D, these are wet tantalum types.

Re the voltage error, have you got the sense connections wired up?

David

[AMR Labs]

In case you missed it, for 115V operation the mains circuit breaker is replaced with 30A one, 230V operation has a 20A one.
And two of the original capacitors were type 150D, these are wet tantalum types.

Re the voltage error, have you got the sense connections wired up?

David

Yes I am aware but have not changed yet the breaker, for now not needed. Besides I doubt I can get 24 amps/3200W out of the 115V wall outlet anyhow. So for now the lower rated breaker will stay in place, and this would also provide a better margin of protection as I do not plan on using this power supply on high power and/or for extended periods of time.

Yes the sense wires are connected as it came I suppose from the factory. See OLD picture, when the junction block was still bent out of shape.
+S -> + Out, -S -> - Out.

Besides if sense where not connected regulation would not work at all, and output would shoot up to maximum voltage.

Thanks.

[bastl_r]

This looks very similar to my 6269B. I wouldn't let it work on 110V.
If you drive the power supply at 40V output voltage into the current limit and terminate it, sometimes the 16A circuit breaker in the fuse box drops when recharging the electrolytic capacitors.

Crazy crazed Rifas and broken off spacer threads.

Here we call them "Knallbonbon"
I also have a 6448B wich contained more than 10 of them.

bastl_r

[AMR Labs]

This looks very similar to my 6269B. I wouldn't let it work on 110V.
If you drive the power supply at 40V output voltage into the current limit and terminate it, sometimes the 16A circuit breaker in the fuse box drops when recharging the electrolytic capacitors.

You are of course aware that the 6269B does not even have the 115V option on the transformer. Actually one of the reasons why I would not go for that model instead, and besides I really did not need more than 20-25amps for short periods of time. Just planning on power testing some devices that I am designing and that require either 12V or 24V. So this model really was perfect for that purpose. But I never imagined that it would be so big and heavy, pictures don't really do it any justice.

Crazy crazed Rifas and broken off spacer threads.

Here we call them "Knallbonbon"
I also have a 6448B wich contained more than 10 of them.

bastl_r

Tell me about it, but at least mine did not "knall" when the one Rifa cap in my 2465B scope went up in smoke, but for a few seconds it sure sounded like someone had punctured a tire, or some high pressure soda fountain gas was escaping.

[bastl_r]

You are right.
The term "Knallbonbon" (firecracker in english)  is a colloquial term. Popping would probably be more correct for the description of the bursting.

[AMR Labs]

Was testing PARD at output, under load, and I see what is on the attached images 1 and 4. But specs (image 3) say should not be more than 5mVpp, but getting between 165 to 175mVpp at 60Hz. Wondering if this is a ground loop effect, as explained in the next image from the service manual. The preregulator phase angle switching is also clearly visible on the waveform. Probe was connected directly across output terminals. Sorry for the somewhat dark images of the scope but it is tough taking a picture with enough light to get a fast enough camera shutter speed without washing out the trace.

I am using a single ended (P6137) probe, so ground loop current might indeed be flowing through the probe's ground lead as described in the manual (image 2).
Time to get a differential probe? I might dig out the Tek 222 scope and measure this again, and while on battery operation so it will be fully floating from ground.

Opinions & comments welcomed.

[AMR Labs]

I finally received the KSD-01 bimetallic thermal switches housed in a TO-220 package to implement the fan temperature control, instead of them running constantly which is both quite annoying and not necessary for my requirements as I will rarely push the power supply hard for a considerable amount of time.

Since both fans inside the HP6268B run on 115VAC, and even tough the metallic mounting flange of the switches seems to be adequately insulated from the switch itself, I was not entirely comfortable running 115V on the main A4 heatsink and to lesser degree at T1 the main power transformer. So I decided to build a small board with a tiny 12VDC converter that takes AC voltage directly, and two 12V relays that would be controlled by the thermal switches, one to control each fan.

For the A4 heatsink assembly I drilled a hole to mount the switch (60-C) directly onto the heatsink, and in the case of T1 I managed to just slip the switch (55-C) between two layers of the cardboard wrapping being careful not to disturb or scratch any of the transformer wingdings. Not sure yet if the temperature trip points of these two switches might need to be adjusted up or down depending on my usage. Guess I'll find out.

[Bileeg]

I have a 6268B that was originally setup in 208 but I was able to configure to 230VDC but can't get it to set in constant voltage or current mode like described in the manual. All the jumpers are in across the terminals as recommended from the factory in normal mode as well. But when I disconnect one of the sensing leads The unit will drive to max voltage. Just as a test I placed a 12 vdc motor on the output as a load and the motor began to ramp up. the only way to control the output was to turn the over voltage screw counterclockwise but it did not go to zero just limited to 15 volts/3 Amps and the Over voltage light did not come on. Not sure what the issue is since I am just a novice at this kind of stuff. Thanks!