Date   : Wed, 06 Dec 2006 19:44:25 +0000
From   : splodge@... (Richard Gellman)
Subject: [OT-ish] A3020 won't turn on

Hi,

!*"?^%!%?%!!!!!!

Top tip! When your Acorn won't switch on... CHECK THE DAMN SWITCH HASN'T 
SCREWED UP! :)

I did some prodding about, voltages here, there, etc. Not a sausage 
coming off the transformer.

On a hunch, I tested for some continuity between certain points... Fuse 
no thats fine... tracks to transformer... fine there... neutral 
switch... fine... live switch.. f.. er...

Nothing. No connection when on. The entire neutral side of the mains 
switch no longer functions. I haven't a clue how on earth that has happened.

One disassembly of the worlds hardest to get into casing, and a 
quick-fix soldering of a wire across the aforementioned switch 
terminals... oh look. It powers on now.

*sigh*

Back to work then...

Thanks for the information, I shall keep it handy for when a power 
supply has a more genuine fault ;)

-- Richard

Robin Commander wrote:
> I thought as much - looking at the data sheet for the L4974A it is a
> switcher.... with an internal drive transistor, hence why there's no chopper
> transistor on the heat sink. Now I'll be able to sleep tonight ;-) That
> chip's size belies it's current capabilities.
>
> Wish we had a schematic for the psu.... can you hear any ticking or
> squealing sounds from the psu area ?
>
> If there's a single bridge (4 diodes) arrangement for the +-12v supplies
> that will account for 3 of the 5 pins on the mains transformer secondary
> side, there'll be a centre tapped winding, the ends of which feed the bridge
> and the centre tap is effectively the ground (0 volt) connection. In this
> case there'll be a separate electrolytic cap between 0v and +12, and 0v and
> minus 12v.
>
> The other two pins on the transformer secondary will probably be the winding
> which supplies the +26v supply that Jules has alluded to. Unless the L4974A
> accepts an AC supply (doubtful)there must be at least a single diode (or
> possibly a bridge) doing the rectification here, along with another
> electrolytic capacitor as the reservoir. If this diode fails shorted I'd
> expect something like you're seeing. It could also be a failure of the
> L4974A chip, or dry joints on its pins....
>
> Do any of the electrolytics capacitors have signs of bulging can at the top,
> or brown gunge around the bottom where they meet the pcb ? This is always a
> good sign of a capacitor which is past its sell by date.
>
> HTH.
>
> Regards,
> Robin
>
>
>  
>
> -----Original Message-----
> From: bbc-micro-bounces+robincommander=blueyonder.co.uk@...               
> [mailto:bbc-micro-bounces+robincommander=blueyonder.co.uk@...               
> ] On Behalf Of Jules Richardson
> Sent: 06 December 2006 15:37
> To: bbc-micro@...               
> Subject: Re: [BBC-Micro] [OT-ish] A3020 won't turn on
>
> Richard Gellman wrote:
>   
>> I don't know if this'll help at all, but putting a voltmeter on the 5 tap
>> outputs of the transformer read as 0.01 V. This may be a red herring
>> however, as I forgot to change my multimeter from DC to AC reading...
>>     
>
> Yep, that's probably just the meter picking up noise...
>
>   
>>> I don't have a photo from the right angle handy, but in the one that I
>>> have it
>>> looks like there's probably a chopper transistor mounted on the heatsink
>>> which
>>> separates the transformer from the bulk of the machine.
>>>       
>> Negative. There are no heatsinks in the whole machine. The metal upright
>> to the right of the transformer is merely a seperator.
>>     
>
> Hmm, you made me go dig out my machine... right you are (on the A3010 that
> is 
> a heatsink with a regulator IC bolted to it), and it does look to be a
> linear 
> PSU. Surprising on a machine so recent (relatively), I would think.
>
> There's a L4974A regulator IC for the +5V rail, driven via a diode bridge 
> (D5/D6/D7/D8) and capacitor (probably C200) for the smoothing.
>
> The +12V and -12V rails are totally unregulated and use diodes D1/D2/D3/D4
> for 
> rectification (and probably capacitor C201 for initial smoothing); they're 
> fused via FS2 and FS3.
>
> The datasheet for the L4974A is online if needed. As a start, I'd measure
> the 
> DC voltage at pin 11 (relative to the 0V contact pad on the PCB) as that 
> should be the unregulated DC from the transformer. On my A3020 it's +26V, 
> although it'll vary a little between machines. If you've got 0V there then 
> there's something up with the rectifiers / transformer / smoothing cap / 
> solder joints / power switch (as you've tested fuses you know there's not a 
> dead short within the chip or elsewhere as the main fuse would have gone)
>
> Pin 13 of the chip is the voltage reference, and should be a little over
> +5V. 
> Pin 20 is the output, and should also be around +5V, but you already know
> it's 
> not :-)
>
> Don't rule out bad solder joints to the transformer; it's a big heavy lump
> to 
> mount on a PCB and no doubt introduces all sorts of thermal stresses as it 
> warms up / cools. Unfortunately separating the PCB from all that shielding
> is 
> a right pig (hence my 'probably' regarding which smoothing caps are which 
> above as I didn't bother!), but it may be worth you checking.
>
> cheers
>
> Jules
>
> _______________________________________________
> bbc-micro mailing list
> bbc-micro@...               
> http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
>
>