H7140 PDP 11/24 Power Supply Advice

From: Peter C. Wallace <pcw_at_mesanet.com>
Date: Tue Nov 13 20:49:09 2001

On Wed, 14 Nov 2001, Tony Duell wrote:

> >
> [Learning electronics]
>
> > I strongly recommend doing so if you plan on working on pre-PC era
> > machines. They do break. But repairing them can actually be quite
>
> I strongly recomend you learn electronics no matter what sort of machine
> you work on. Of course I'm biased :
> (1) I like to fix the actual fault, know I've fixed it, and know it's
> going to stay fixed. Not just swap parts until the machine seems to work
> again.
> (2) Electronics is fun. I'm an electronic hacker first and a computer
> person second. I tend to regard computers as being interesting electronic
> circuits [1].
>
> With anything older than a 486, you're probably going to have to fix it
> yourself. There aren't that many people who will work on older machines
> now, and those that do tend to be expensive (for 2 good reasons, (1) it
> can take a long time to track the fault down and (2) there are few enough
> people doing it, so there's little competition). But I can assure you
> that fixing such machines is fun and _very_ educational.
>
> [1] OK, I think of electronic computers as being interesting electronic
> circuits. Babage's engines and other mechanical machines are very
> interesting too, but they're not electronic circuits ;-). But I don't
> generally get to repair such machines
>
> > And, sometimes, recognizing circuits can be good for your health. Although
> > I am certainly no expert on power supply design, I can recognize a +300V
> > voltage doubler when I see one, and realize what that might mean were I to
> > touch it.
>
> You won't feel a thing. Ever again. In other words you'll end up very dead.
>
> Switch mode power supplies are not particularly nice to work on. In
> general a lot of the internal circuitry is directly connected to the
> mains, so touching just about any point on the 'hot' (primary) side of
> the power supply is equivalent to sticking your finger into a mains
> socket. Which is never a good idea! You can't clip a 'scope onto the
> primary side either (not unless you're running the PSU off a double-wound
> isolating transformer for testing) since the 'scope ground is effectively
> much the same thing as the neutral side of the mains. Clipping that
> 'scope ground lead onto a point on the hot side of the PSU is similar to
> shorting ou the maisn. Again never a good idea.
>
> I would recomend learning a fair amount of electronics before attempting
> to repair such a unit. Or at least making sure you know the safety
> precautions.
>
> > Even more important is basic power line safety. Knowing what the rules are
> > for hot vs. neutral vs. ground (and when to recognize that a previous
> > repair might have been done by someone totally clueless) can save your LIFE!
>
> The first rule is to _never_ assume that either the manufacturers or the
> previous owner knew what they were doing.
>
> Don't connect anything to the mains before you've pulled the case and
> checked inside. At least a visual inspection, and preferably also an
> insulation test and an earth continuity test (if you have the instruments
> to do that). Of course the visual test depends on knowing what to look
> for. Bare wires (and strands of wire coming off terminals), evidence of
> modifications (which may well have been done incorrectly, especially if
> they look to have been done poorly), burnt components, trapped wires.
> damaged insulation, missing earth wires, loose wires, and so on.
>
> -tony
>
>


This reminds me of repairing a Perq power supply many years ago. The
switching supply had the switching transistor drive circuitry on the
isolated side, driving the switch through a transformer. Seemed like it
couldn't possibly work when I first looked at it, since how did the power
supply ever start up if there was no low voltage to begin with. Turns out
that there was a simple 3 component relaxation oscillator (diac? +
resistor + capacitor) that tickles the main transformer, generationg
enough voltage on the isolated side to start the main switching
transistor...

The resistor of that circuit was open which is why the power supply
failed...


Peter Wallace
Received on Tue Nov 13 2001 - 20:49:09 GMT