7 Jun
2010
7 Jun
'10
1:45 p.m.
On 05/06/10 10:55, Tony Duell wrote:
I suspect a short in the transformer. Whatever it was, it made a loud
bang and dumped a lot of Magic Smoke...
Hearing the fan (briefly) rev up to WAY beyond full speed pretty much
cemented my (correct) opinion that the machine wasn't going to be doing
anything useful any time soon.
Argh!. I can remember when Maplin sold some
reasonable-quality tools.
Like 'Elora' spanners (They did a few very small sizes that were very
difficult to find). And Weller soldering irons (I bought at least one bit
for my TCP from them). Not any more :-(.
They used to stock the Antex 660TC soldering station. I got one as
"discontinued stock" a while ago for a good bit less than the RRP. That
thing has had one new soldering iron (was ?2 cheaper to replace the
whole thing than just replace the dead heating element), a couple of new
bits, and I've had to open the back to tighten up the screws on the
power transformer. Other than that, no issues. Solidly built
British-made kit.
And they don't sell tin/lead solder any more.
But Farnell still do :)
The Multicore stuff is fiercely expensive, but the stuff they stock as
"Multicomp" (i.e. "whatever was cheapest at the time") is pretty
decent.
At one point the Multicomp stuff was badge-engineered Multicore "562"
60/40 SnPb or something very close to it.
* PC power
supply. Blew up and took an expensive motherboard and CPU=20
with it. Even put a PSU tester on it, which showed the +12 rail at=20
almost 20V! -- "We'll replace the power supply but you'll have to RMA=20
That's an odd fault... With many SMPSUs the 'extra' outputs will be low
if the main output is not sufficiently loaded, but having such ouptus go
(very) high is uncommon. I wonder what the fault was. Incidentally, my common comment 'Check the PSU on
dummy load' applies to
brand new PSUs too :-).
"Dummy load / PSU tester" is right on the top of my "stuff to build"
list :)
I just have to figure out how to do the "constant power" regulation. I
want to go with opamps for the regulation (faster response time to V/I
transients) and a D/A to set the current/voltage/power 'setpoint'.
Constant-voltage is easy (drain enough current to bring V_in down to the
setpoint).
Constant-current is just as easy (drain a fixed current).
Constant power involves a bit of maths (a multiply and some division if
you're doing PID control or some lesser variant thereof). A
microcontroller-based implementation would work, but respond slowly to
load transients (which is hardly ideal).
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/