Debugging power supply problem in HP 215A pulse generator

[JuiceKing]

Hi,

I picked up a broken HP 215A pulse generator on eBay and have been trying to restore it to working order. I was able to adjust the power supply to specified values for all rail voltages except the +40VDC line. That one won't go higher than 35V or so, and has excessive ripple voltage (>2V) as well.

Can you take a look at the attached schematic and offer suggestions of in-circuit tests I can do to isolate the cause? I would prefer not to replace all electrolytic capacitors just because they are old, but my guess is that at least one is bad. I checked to see if C302 is shorted, and it was not. C303 is the big cap on the +40V line. At least on the surface, it looks fine, as do all the components on the board.

Happily, the -40V side of the power supply is working fine and has a close symmetry to the positive side, so it's easy to compare behavior of components on one side vs. the other. However, I want to avoid removing components from circuit if at all possible.

Any suggestions?

Thanks!

- Ken

[free_electron]

C301 is suspect no1...

[JuiceKing]

C301 is suspect no1...

C301 is one of the big 4 electrolytic capacitors that are riveted to the assembly and have multiple soldering connections to beautifully dressed wires with zero slack. It looks very hard to remove.

Is there a way to test it in-circuit to see if it's the source of the trouble? I checked for shorts and its resistance was in the 150-200 ohm range, and so it was for the other 3 big caps.

- Ken

[Tube_Dude]

C301 is one of the big 4 electrolytic capacitors that are riveted to the assembly and have multiple soldering connections to beautifully dressed wires with zero slack. It looks very hard to remove.

Is there a way to test it in-circuit to see if it's the source of the trouble?

For me C301 is the principal suspect too, if you don't want to disconnect it, try solder another cap in parallel to see if things improve...

[SeanB]

Place a 1000uF 63V electrolytic in parallel with C301, leaving the old one in place. As well disconnect the connections to the transformer at points 10 and 12 and check the 2 diodes CR301 and CR302 if they are either open ( most likely one is) or leaking. Replace C302 with a 22uF 63V one anyway, and C303 with a 220uF 63V one. Thast should solve the problem, as this is the most common fault in a power supply that still produces an output.

They really wanted to use only one type of power device there, such a convoluted design to accommodate this.

[free_electron]

is the goal to fix this thing and use it ? or restore it to its lack-and-lustre , put a perspex cover on it and put it in a museum of modern art ?

if the cap is riveted : take the drill and hammer out .

[JuiceKing]

is the goal to fix this thing and use it ? or restore it to its lack-and-lustre , put a perspex cover on it and put it in a museum of modern art ?

if the cap is riveted : take the drill and hammer out .

I'm keen to fix it, but those big capacitors that are riveted into the center of the chassis look really hard to remove so I'd like to test and/or replace them without pulling them out. The ones that are simply on the power supply assembly board are easy to replace.

Fortunately, new capacitors are more compact than these old ones. I might be able to tuck them into empty space in the bottom of the unit (mostly taken up by a long coil of delay line) and connect them with 5-6" wires to the points where the original caps are mounted. Obviously, I'd clip the connections to the original caps, but leave them in place. Any reasons not to do that?

- Ken

[david77]

You can do that in theory, yes. 5-6" of wire to the cap might reduce the effectiveness of the new cap, though. The rule of thumb is to have the caps as close to the power supply circuit as possible for best ripple rejection and in line with the supply circuit as opposed to have it sitting at the end of a spur taken from it.

I would advise to disconnect the old cap in any case. It may have low resistance now and could draw a significant amount of power or it might deteriorate in future until it's a dead short, then you'd have to go in there again.

[M. András]

can we see some pics of the internals?

[JuiceKing]

can we see some pics of the internals?

Here are some pix: front, top, top (power supply detail), bottom, bottom (capacitor detail).

- Ken

[mikeselectricstuff]

Rivets are easy enough to drill out.

[Tube_Dude]

Why not just disconnect the wires from the capacitors, connect them to a new one and see if things improve?

[T4P]

At first when i saw it i went OOOOH AAAAH and gaga over the black TO-3 transistors

My word? same as tube_dude, just cut it up or bodge the board

[tekfan]

Rivets are easy enough to drill out.

You dom't even have to drill out the rivets to remove the large capacitors. Just unsolder the wires from the bottom and then twist back the four eyelets holding in the capacitor. You should then be able to remove the cap from the top and the brown pertinax holder will stay riveted to the chassis. I've experienced the same problem with older test equipment. The electrolyte inside the cap dissolves the aluminium foil connection to the solder tags.

If you really want to preserve the oldschool look of the generator then open the bottom of the removed cap and pull out the 40 year old insides. Then clean the inside of the empty can and put in new capacitors. Close it up and solder it back into the generator. Easy as that.

Similar process: http://home.comcast.net/~hanksradioland/tips/mkcan/mkcan.htm

[T4P]

In fact what tekfan has mentioned is i think done everyday in china ... You get what i mean automatically

[JuiceKing]

Rivets are easy enough to drill out.

You dom't even have to drill out the rivets to remove the large capacitors. Just unsolder the wires from the bottom and then twist back the four eyelets holding in the capacitor. You should then be able to remove the cap from the top and the brown pertinax holder will stay riveted to the chassis. I've experienced the same problem with older test equipment. The electrolyte inside the cap dissolves the aluminium foil connection to the solder tags.

If you really want to preserve the oldschool look of the generator then open the bottom of the removed cap and pull out the 40 year old insides. Then clean the inside of the empty can and put in new capacitors. Close it up and solder it back into the generator. Easy as that.

Similar process: http://home.comcast.net/~hanksradioland/tips/mkcan/mkcan.htm

Aha! That's what those four twisted eyelets are for!

[JuiceKing]

Aside from meeting the voltage and capacitance specs of the original parts, do you guys have any other recommendations on what specific kinds of electrolytic capacitors I should use for these replacements?

If the PSU diodes are leaking (CR301, CR302), can you suggest a replacement? The HP service manual refers to them by an old HP part number and no generic substitute.

Thanks!

- Ken

[JuiceKing]

You dom't even have to drill out the rivets to remove the large capacitors. Just unsolder the wires from the bottom and then twist back the four eyelets holding in the capacitor.

Hmm...as you can see in the photo, there are leads soldered to some of the eyelets. What are they for if not simply mounting points? It looks like there are two leads in the center (these I thought were the capacitor leads) and then four eyelets. If I disconnect the wires currently connected to the eyelets, what should I connect them to after I replace the capacitor with a new one? In the example you showed, the eyelets connect to the can, and the can is marked to be negative. Is that typical?

[T4P]

Typical, in fact a modern day capacitor's metal body is at negative potential which means their negative pole
The plastic around them helps in some way or so, can't really remember for heat dissipation
And most importantly keeping the negative pole voltage from the rest of the cap.
Which means you can probably have a short from the top of the cap but you can't do so from the rest of it

[tekfan]

There are actually multiple completely separate electrolytic capacitors in one of these cans. They all have a common negative pole which is connected to the can of the capacitor and the four mounting eyelets. The other leads are the positive leads for each separate capacitor inside the can. If you look closely you will see a semicircle, square and a triangle on the bottom of the can. These denote the positive leads. If you now look on the side of the can you will see several ratings and a symbol on the end that tell you which pin the rating is for.

so for this cap there are actually three separate ones inside with a common negative connected to the can
semicircle is 120uF / 200V
square and triangle are 40uF / 200V

[JuiceKing]

C301 is suspect no1...

And you would be correct!

I attached a new 500uF capacitor in parallel with C301 and now (at least for positive pulses) the pulse generator is working beautifully!

Thank you!