Good 9 years ago we acquired and brought to working order old but still quite capable HP 4263B LCR meter. It was rarely used, mostly powered up only on occasional need to measure some capacitor for nominal value determination and checking Rs/Rp resistances of capacitors.

This week I’ve powered it up for a SMPS repair project to test old capacitors and new replacement, but LCR failed. LCD was barely lit and 4263B refused to boot up. I tried multiple times, and after few attempts strange sizzling noise could be heard from the unit and even LCD now was completely dead. Sounds like a power supply issue?

None of the internal PSU capacitors in the instruments were replaced back in 2014, as everything looked good and power rails measured nice and clean back then. Perhaps first mistake leaving original capacitors in the unit for this long? One should think twice of not doing mandatory recapping of old test equipment, especially after learning of destruction of so many Keithley 2001 and 2002 DMMs due to leaking caps.

This module has part-number HP 04263-80901. It is taking +14 VDC voltage from linear transformer and diode bridge and generates +12 , +5 and -12 V rails for instrument use. These modules are listed on ebay for some absolutely insane prices, like this:

And like typical for HP instruments made by YHP division these come with quite limited service information and no schematics. So our option is to attempt a repair on this module without buying a replacement.

Module is mounted to metal plate, used as both heatsink and chassis. All active TO220 devices used plastic case without exposed metal to provide insulation.

Sadly this module was built with poor looking Marcon capacitors. Marcon is a subsidiary of Toshiba from 1942, later gobbled up by Chemi-con in 1995. I’d rather prefer to see here some Rubycon, Vishay or CDE’s to be honest.

In any case, the capacitors are very very dead with guts spilled all over the circuit board.

After some angry time with screwdriver the PCB with electronics was liberated out of the metal heatsink/frame. It revealed nasty looking goo on the top of the PCB around capacitors and transformer. Yes, this is a switch-mode power supply module.

Back of it is not any much better. Designers didn’t bother using connectors for cables and just soldered wires here and there at random places. Layout engineer was on the vacation as well, so few missing connections were jumpwired with few extra cables too.

Just few swabs with IPA-soaked Q-tips take a lot of green crud out of the PCB. Much more cleaning is required to revive this module. Electrolyte is corrosive so even soldermask is coming off the copper tracks together with IPA.

Repair would include removing ALL thru-hole components, wires, active parts and probably most poor looking SMD passives. Then actual PCB cleaning can begin, in multiple iteration till all of the electrolyte is removed.

PCB around L266 4-pin device look quite damaged and discoloration of that component black body hints on overheating event. Perhaps PCB is charred and will need to be drilled out.

BOM for parts to replace:

Anyhow, hopefully we can make this module work again soon!

Stay tuned and let us know your feedback on this post! Discussion about this and related stuff is also welcome in comment section or at our own IRC chat server: xdevs.com (port six-zero-ten-zero, channel: #xDevs.com) or via e-mail.