On 8/3/2019 8:46 AM, Scott Dorsey wrote:
In article , Paul wrote:
On 8/2/2019 5:46 PM, Scott Dorsey wrote:
Now, he does have a good point that the original poster likely doesn't
have the right equipment or skills to diagnose the problem. But....
everybody has to start somewhere and if you don't try you'll never learn.
I've got a Tektronix 465M o'scope, a digital multimeter, and an ESR
meter.
Get a load bank and a storage scope and you'll probably be set. The storage
scope is REALLY handy to look at the startup of the pwm ic because it's only
a few milliseconds long. If it's not repeatedly shutting down, it can be
sometimes shutting down on the first try, and then the storage tube tells you
what is going on. BUT... the good news is that with these all-in-one PWM ICs,
there are only a few static voltages that can cause it to turn on or off, so
usually by looking at static voltages you can see what is going on. With
discrete transistor PWM circuits you need the storage tube.
Good point. A storage scope would be a nice addition.
If you haven't used the ESR meter on every electrolytic, now is the time.
Even if they aren't related to the problem, you may pre-empt future failures.
And having been a microwave/Radio frequency engineer for nearly 3
decades, I'm not likely to electrocute myself on the bench!
This puts you in the category of the "goddamn white hat engineers" and makes
it that much less likely that you should touch a soldering iron. I'm in
that category myself, but at least I don't have a PhD so I am reasonably safe.
I've done plenty of bench engineering, and REAL engineers do
at least SOME of their own soldering!
But I'll admit I don't trouble-shoot modern switching power supplies
very often. But it can't be denied that these are NOT trivial circuits
to diagnose, especially WITHOUT a proper schematic!
I apprenticed at a TV repair shop in the days of sweep tubes locked up in
high voltage cages and all the set voltages being derived from the flyback,
and it was amazing seeing some of the tricks. Guys used to run a TV set in
series with a 500W light bulb as a current limiter so they could keep them
going long enough to diagnose oscillation issues. Sometimes with schematics,
but mostly without.
And even your article mentions the when-in-doubt-replace-it, brute
force approach to repairing these tough-dog problems. I've certainly
successfully done this in the past, but I'm not sure I wanna do that
on this unit, because my stock of spare parts is not what it used to
be.
That's easy to fix with a digikey order. Pick up ten of every common PWM
chip too, because sooner or later you'll need them.
This unit might end up on Ebay.....
Haha!
See, that is giving up. That is not right.
The "shotgun" approach of replacing components,
until the unit works, often does work, but I have to balance
the cost of the parts, and what I want to spend my time doing,
versus what I expect to gain at the end.
But you are right: I'm kinda lazy, and I'm more into the
money, than the knowledge, to be honest, especially at my middle age.
I prefer to call it: Selective Cut and Run!
I mean, sometimes you can make more money, and save
yourself some aggravation, by selling the parts of a unit,
instead of repairing it yourself.
--scott