On 2017-Aug-20, at 4:45 AM, Adrian Graham via cctalk wrote:
> I think I’ve mentioned this PSU before but the question I have might be
> applicable to other cheap switchers. This lump for the original Atari 520ST
> is run by an NE5561N control chip with a D45H1 switching transistor. It’s
> supplied by a multi-tap transformer giving 2x2 feeds of 6VAC and 14VAC with
> an eventual output of 5V@3A, 12V@30mA, -12V@30mA.
>
> This one will run under load for around 10 minutes then the 5V rail will
> gradually collapse over the next few minutes. Turn it off and leave for a
> minute or so and it’s back to normal, repeat as above.
>
> Unfortunately there’s no schematic online but I’ve done a mix photo of the
> control circuits:
>
> http://www.binarydinosaurs.co.uk/atariPSU1Hybrid.jpg
> <http://www.binarydinosaurs.co.uk/atariPSU1Hybrid.jpg>
>
> The 4700uF cap in the centre is on the output stage so I changed that
> yesterday which just makes the eventual collapse longer.
>
> Could the actual transformer lump break down over time and recover that
> quickly? Its 6VAC outputs do go a bit erratic. Also I measured the GAIN input
> on the NE5561N pin 4 and during the collapse it goes over twice what it is
> when the 5V rail is stable (7V instead of around 3.4V). I have another
> running supply to measure against.
>
> Any clues appreciated. It’d be nice to feed the 6VAC input from another
> source but I’m not sure my VARIAC will go that low.
The gain pin is presenting the output of the error amplifier, an increase there
is what you would expect as the output V falls, as the difference between the
ref and target V increases.
It's 'trying harder' to raise the output V to correct the error.
It implies that the reference and the error amplifier are functioning.
"6VAC outputs do go a bit erratic" is not a helpful description.
A wide range of stepdown transformers .. 8V, 10V, 12V, 18V, 24V .. could be
used in conjunction with your variac to get 6VAC for an alternate feed.
If you have a variable DC supply of adequate current you can feed approx. 6V *
1.4 = 8.4VDC into the 6VAC input instead.
Is the failure load-level (current-draw) related?:
Does the out V remain stable if the supply is unloaded or very lightly
loaded?
Does it take longer to drop at a lesser load?
If so it's likely thermal-related.
Is anything heating up unduly?
If the transformer were faulty and heating up internally, it would seem
unlikely that it would cool off that quickly to recover, IME it takes some time
for a transformer to cool off.
