Help with ICL power supply fault (Farnell SMPS)

[John S]

Hi,

I've started to setch the output circuitry, not 100% accutarely as my meter shows all the transformer windings as 0 ohm, so I have some points that measure 0 ohm to ground which may be ground or maybe a transformer output.

Thanks Tony for your reply, quote below.

Tony wrote:
>John wrote>

>> I have an ICL PC2 CP/M box (like this:
>> http://www.old-computers.com/museum/computer.asp?c=752) which has a
>> faulty PSU, I am hoping someone here can advise me.
>>
>> The PSU is a Farnell N100/F4190 SMPS, looks like a high quality unit
>> with nice screw terminals for mains in and DC out.
>>
> When fired up with 240V AC and a dummy load the output voltages are:
>>
>> 12V output=1.1V
>> 5V output=2.3V
>> -12V output=-5.9V
>
>The first thing that 'bothers' me is that the 12V output is lower than
>the 5V one. If this was a simple regulation fault, I'd expect all outputs
>to track.
>
>Do you know which output the crowbar operates on (not which one triggers
>it, but which one it shorts out?).

Now I have some of the output circuit, I have found their is an SCR (2N6400) connected as follows:

Anode - to +12V output
Cathode - to ground
Gate - to a resistor divider, which inturn is connected via a diode (assumed to be a zenner) to the +5V output.

With the power off I have tried applying +5V to the +12V output, and this drew 250mA - ie around 20 ohm. most of this current turned out to be going through the SCR, as when I removed the SCR the current reduced to around 20mA. I think the SCR has been damaged, it is showing 80ohm between gate and cathode with the meter either way round (I think this should be a diode!). I guess the SCR must have taken the brut of the 80W or so of excess power taken by the PSU when I was testing it earlier. I'll buy a couple of replacements.


> It's not uncommon for that to be
>something other than the main output (simply becuase it's easier to pull
>down), it's possible there's a crowbar thryistor on the 12V output, and
>that that's firing.

Exactly right - thanks for the tip.

>I assume this thing doesn't have external sense inoputs for the main
>output, or if it dows you've conencted them to said output.

No sense inputs.

>I think it's reasonable to asusme the chopper is working (otherwise you'd
>get no outputs at all). Does it seem to be running continuously, or do
>you get the 'tweet tweet tweet' of a PSU that's starting, detecting a
>fault, shutting down, and repearing?
>
>Are tyhe otuptus steady at those votlages? I find an analogue meter best
>for this, you can see the needle twitch if the PSU is starting and
>shutting down.

When I tested it earlier I applied power for around 5 seconds at  a time, outut voltages were fairly stable.

Next thing I'll look at is the feedback circuit from the +5V output via some resistors and what might be a FET into an opto-isolator to a circuit on the hot side involving a compartor, transistors, resistors and diodes. This looks fairly complicated (to me, probably a piece of cake for Tony!). Pity there isn't a nice IC controller like on the HP-85 PSU :-)

Regards,
John
_________________________________________________________________
Get Hotmail on your mobile. Text MSN to 63463 now!
http://mobile.uk.msn.com/pc/mail.aspx