Graeme Geldenhuys wrote:
On 30 January 2012 12:05, Mark Morgan Lloyd wrote:
mainframe switchmode PSUs, so with apologies to the list owner I'm probably
in a position where I have to comment on this for safety reasons.
My advice: don't.
Thanks for the warning. :)
any number of stories about people who've done something that they thought
was safe which has gone on to cause damage or injury.
Same could be said for ridding a motorcycle (which I do often) or even
crossing the road, but I get what you are saying.
Except that riding a motorcycle or crossing the road you have some idea
of what the conditions are, and have spent a few decades honing your
skills. Think of riding a motorbike through the Bushveld, and then
finding yourself on black ice.
Think of the nastiest bit of software you've ever come across, spaghetti
code with computed gotos (which I guess implies FORTRAN). When it comes
to maintainability, that's roughly on a par with the best switchmode
PSU, since you've got power and feedback moving in different directions
through the same wound component. One of my colleagues spent a while
designing them, and even he blanches at the idea of a repair.
SMPSUs are the last thing you want to tinker with, not the first. If you
want to get into electronics fairly painlessly then start off with a
BASIC Stamp or Arduino (I'm not aware of anything of that scale which
will run FPC) and some project cookbooks. Also I believe that there are
maker groups in most cities, although I'm not in touch with my local one.
Things like the output wire on low voltage PSUs are fair game for repair.
You can get spare concentric connectors from RS or Maplin in the UK (Graeme-
I thought you were abroad?)
I lived in the UK for 4 years, and moving back there in 2 months time.
Yell if you're in the Southeast.
unopenable by hand i.e. you /had/ to use screwdrivers etc.) then I'd suggest
looking around for something like a CG-accredited electronic technician
course- which full-time would take years.
When I finished high school (many many years ago), it was a coin flip
between studying pure programming or electrical engineering (light
current) and hardware programming. I opted for the pure software
programming, which has been my career path ever since. But I have
always had the urge to start some part time studies to get an
Electrical Engineering degree - those little colorful bits of
electronics just fascinate the hell out of me (even 20+ years later).
:-)
Thanks for the URL, I'll keep that handy in case I ever need some
electronics repaired.
Bering in mind that he is specifically PSU, and won't touch some types
since they require proprietary components.
From elsewhere in the other thread, Lars said:
-8-
To keep it on topic, there is programmable hardware available where you
can change the hardware using a hardware programming language. Niklaus
Wirth is interested in such technology. Instead of soldering in capacitors
and resistors, you program in something that emulates a resistor or
capacitor. This makes prototyping circuits much easier because instead of
soldering, you program in the devices you would have otherwise soldered.
The devices are called Field-programmable gate array's I think, and from
what I remember Niklaus Wirth was programming a remote control helicopter
with it, or maybe oberon, it's been so long that I cannot remember the
details. Just trying to keep it a bit on topic.
-8-
There's two things here: the first is Digital Signal Processors (DSPs)
where you program them to compute a series of equations in real time
that implement one or more filters. Things like telecoms systems
(including ADSL modems) are full of them.
Then there's FPGAs, although they're usually programmed in a specialised
hardware description language. They aren't so much set up to emulate
resistors etc. as to implement an application-specific CPU and
associated logic, in some cases the hardware contains a CPU core (e.g.
Michael Schnell's earlier
-8-
IMHO, the MIPS port is interesting, too, as the NIOS CPU that is
available as soft core for Altera FPGA uses a very similar instruction
set and supposedly can be handled as a sub-arch.
There is a very active community for NIOS-Linux that now provides a
quite decent MMU-enabled Linux port (including the necessary
Distribution generating Make-based tool chain).
For quite some time I played with NIOS using a NEEK Dev-Kit from
Altera. Nice stuff.
-8-
I can't remember for certain but I think it was Nicoud who was involved
in the helicopters rather than Wirth. Nicoud is notable since one of his
ETHZ projects was a portable assembler notation, I've wondered in the
past whether this could be used for compiler implementation but I think
it's probably less useful than some intermediate notation akin to p-code.
--
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk
[Opinions above are the author's, not
