HP 5385A freq counter repair

[Charles]

It quit working this week and found the 2 A fuse blown. Not in stock, so
had to order some. In the meantime, took it apart and powered it from a
bench supply and found it OK but only if the bench supply was turned on
after the front panel switch was turned on. Turning it on or off with the
front panel switch activated the crowbar circuit.

I suspected a switcher problem and probing in circuit showed C36 (the main 5
volt switcher cap) to be measuring about 1 uF and the schematic shows 390
uF. Replaced it and all is fine now. Fuses are due tomorrow from Amazon
(hate to make those folks work on Sunday).

Lucky that the manual with schematics is a free download. Google showed it
about 6 choices down. NEAT!

[Phil Allison]

Charles wrote:
>
> It quit working this week and found the 2 A fuse blown. Not in stock, so
> had to order some. In the meantime, took it apart and powered it from a
> bench supply and found it OK but only if the bench supply was turned on
> after the front panel switch was turned on. Turning it on or off with the
> front panel switch activated the crowbar circuit.
>

** The HP counter has a conventional PSU delivering about 14VDC followed by a switching regulator for the 5V rail. 2A fuse is between the bridge rectifier and a pair of 2200uF electros. If the 5V rail goes out of tolerance, an SCR is triggered to short the electros and blow the fuse.

> I suspected a switcher problem and probing in circuit showed C36 (the main 5
> volt switcher cap) to be measuring about 1 uF and the schematic shows 390
> uF. Replaced it and all is fine now.
>

** The 390uF electro is the filter for the 5V rail and comes after a choke, high speed diode and switching transistor.

With a value like 1uF it could not possibly do its job and the 5V rail would be full of HF switching noise which could trigger the SCR.

Nice of HP to fit that crowbar and save all the digital circuits.

I have worked on expensve gear that had massive damage or was a complete write off because of a wayward 5V rail.


.... Phil

[dcjd...@gmail.com]

Thanks Guys, your post got me back up & running quickly!

[Bob Engelhardt]

On 9/24/2016 11:18 PM, Phil Allison wrote:
> ... If the 5V rail goes out of tolerance, an SCR is triggered to ...
> ... the 5V rail would be full of HF switching noise which could trigger the SCR. ...

How would the monitoring circuit work, in general terms?

[Phil Allison]

Bob Engelhardt wrote:

--------------------------

>
> > ... If the 5V rail goes out of tolerance, an SCR is triggered to ...
> > ... the 5V rail would be full of HF switching noise which could trigger
> > the SCR. ...
>
>
> How would the monitoring circuit work, in general terms?
>

** Essentially, just a 400mW Zener diode rated at 4.7 or 5.6 volts.

The SCR has a resistor of say 100ohms from gate to ground and the Zener connects from that to the 5V rail. Takes about 0.8V to trigger the SCR so about 8mA of Zener current.

I have added such to several linear & SMPSs to protect digital boards from faulty PSUs. In one case, I had to replace about 100 digital ICs after a linear 5V supply failed to circa 10V DC.

Luckily they were all in sockets.



.... Phil

[Bob Engelhardt]

OK ... I was thinking that the AC switching noise itself would have to
be detected, but it sounds like you just need to look for the noise
peaks, which would exceed the zener htresho0ld.

Bob