[time-nuts] HP 58503B PSU replacement

[Gerhard Hoffmann]

Am 14.11.18 um 18:41 schrieb Dr. David Kirkby:
> On Thu, 8 Nov 2018 at 21:43, Dr. David Kirkby <
> drkirkby at kirkbymicrowave.co.uk> wrote:
> On Thu, 8 Nov 2018, 14:53 George Atkinson via time-nuts <
> time-nuts at lists.febo.com wrote:
>>> Last month I was at the UK “National” hamfest and in the flea market area
>>> a seller had a HP 58503B for sale. It was missing the cover and power
>>> supply (AC mains version)including mounting. However it was only £3 (<5$)
>>> so it was a no-brainer to buy it. Research shows that PSU failure is fairly
>>> common on these units.
>>>
>>> I had the PSU fail in my 58503A. It keeps trying to start, then stops. I
>>> have checked capacitors with an ESR meter, but could not fix it.
>>>
>>> Three is a good chance your unit takes the same PSU. If so
>>>
>>>
>>> https://www.ebay.co.uk/itm/BEL-POWER-MAP55-4003-55W-Quad-Output-Embedded-Switch-Mode-Power-Supply-SMPS/132833037734
>>>
>>> may be of interest. I had an offer of £45 accepted yesterday.
>>>
>>> Dave.
> Well, the bad news is the PSU I received was faulty (no output whatsoever),
> and the seller has checked another, and that too is dead, so it looks like
> he has rejects. He has refunded me and does not want the one he sent back.
> So now I have two non-working supplies. The new one does not power up at
> all. The old one does, but cycles on/off.

If more than one replacement unit does not work, then it is not

improbable that the switcher needs a minimum load to start.


Also, a shorted capacitor can show excellent ESR.


I once had a cycling power supply in a W&G TSA-2 spectrum analyzer.

It delivered power for some seconds, and that was enough to trace

a voltage to a faulty feedthrough capacitor. A DVM with enough digits

helps a lot. Towards the short, the voltage sinks mV by mV.


regards, Gerhard