Greg Mayman wrote in a message to Roy J. Tellason:

 >>> Part 1 of 2...

 -=> Roy J. Tellason said to Greg Mayman
 -=> about "POWER SUPPLY" on 01-08-04  04:06.....

 RJT> I don't even *try* to troubleshoot surface mount stuff, it's just not
 RJT> economical...

 GM> When it is a free modem, and you just *KNOW* that it should be able
 GM> to work properly... and you don't have a replacement... 

Yeah,  sometimes you just want to fiddle with stuff.

 RJT> Heh.  Too bad what used to be an industry leader of sorts (as in
 RJT> "Hayes Command Set"!) is now gone...

 GM> Yes indeed :-((((
 GM>  
 GM> *  a 350w audio power amplifier

 RJT> There was a time when I'd have enjoyed that much power,  but having

 GM> That sort of thing is absolutely no use to me either.

 RJT> Motorized?  No kidding.  I'd like to have a decent _simple_ circuit
 RJT> for an electronic volume control of some sort that didn't involve the
 RJT> use of special chips,  and still haven't found it.

 GM> The advantage with a motorized one is that it is still at the
 GM> setting you left it at, next time you switch on.

I saw some fancy studio-type equipment like that,  they pushed some keys on
a computer-style keyboard (there was a monitor over the top of the
"mixing console" and this software made sliders move around and
such.  I also recall something of the sort with Yamaha organs,  some years
back,  where the presets would to the same sort of thing,  move switches
and sliders around so you could get a pretty good visual indication of
where things were. 

 RJT> About the closest I've come to that would be an abortive experiment I
 RJT> started once that involved CMOS switches and pulse-width modulating
 RJT> chopping the signal.  It distorted horribly,  I'm pretty sure because I
 RJT> should've been using a bipolar supply and watching my signal levels
 RJT> more carefully,  but perhaps I'll get back to that somewhere along the
 RJT> line.

 GM> Yes, it could have been. I've seen some very low distortion
 GM> circuits, but most of them used specialized chips or modules.

There are a lot of very high-spec analog switch chips out there,  some of
which will operate from 15V bipolar supplies (and why did regular
4000-series CMOS have to stop where it did voltage-wise anyway?),  but
that's not the sort of thing I'd care to experiment with.  And I have a
whole *mess* of 4066 chips on hand...

 GM> But one that worked well used a junction-gate FET to shunt the 
 GM> signal. Normally the FET is non-linear due to the amount of signal 
 GM> between drain and source, but this circuit fed a selected amount 
 GM> of signal back to the gate -- approximately 50% of the amount on 
 GM> the drain -- which reduced the distortion by a factor of up to 100 
 GM> in some cases.

Hmm.  I think there was *one* 4007 chip in that can...   :-)

 GM> I saw this done in various ways, but the easiest was to take the
 GM> signal from the drain to an emitter follower or a unity gain op-amp
 GM> and take the feedback to the gate from the output of that stage.

 GM> *  a 0-80A LED ammeter for cars

 RJT> No zero in the middle?  :-)  I've seen an article or two like that,

 GM> Actually it used 4 LEDs on one side and 5 on the other with the
 GM> "middle" one as the zero. The heart was a LM3914 chip, fed from an
 GM> op-amp amplifier measuring the drop on the battery earth cable. I 
 GM> bought this unit as a kit from Dick Smith Electronics, but I 
 GM> haven't made it up yet... doesn't that sound familaiar? 

Could be.  I have some stuff on paper around here someplace that details a
couple of variants on this idea.

 RJT> I've yet to get my hands on any of those "white" ones so far.

 GM> Our three main "hoppyist" suppliers in Aus all have them at 
 GM> reasonable prices. And they all do mail order around the world.
 GM> Look for jaycar.com.au, altronics.com.au and 
 GM> dicksmithelectronics.com.au on the internet. AFAIK they all have
 GM> on-line catalogues.

I'm sure I could find one,  but that's okay,  I can wait a while yet...

 GM> *  an efficient power supply for a 1W ultrabright LED "star"

 RJT> Yes?  What would that be all about?

 GM> The "star" is a 1 watt white LED. The original project in Silicon
 GM> Chip drove one with a linear constant current suppl from a 9-18v
 GM> source. This later article uses a flyback supply from two "D"
 GM> cells, and claims it will give near normal output down to 1v per
 GM> cell, if I remember correctly. The total current on new cells is
 GM> about 300-400mA which is about the same as an incandescent torch
 GM> bulb, but this thing gives a LOT more light!

Sounds good!

 RJT> Sounds like you've got a fairly decent magazine there.  The ones here
 RJT> have all gone to hell,  pretty much,  with little of interest to me in
 RJT> them in recent years.

 GM> Silicon Chip was the "poor" relation back about 10 years or so ago.
 GM> The leaders were Electronics Australia and Electronics Today 
 GM> International.

 GM> Then EA absorbed ETI. then a couple of years back, EA turned itself
 GM> into a yuppie cocktail table magazine -- "40 PAGES OF MUST HAVE
 GM> TECHNO TOYS" and the like -- and died after about 2 or 3 issues!

Euw.  I get really tired of that sort of thing really quickly...

 GM> So SC is the only one left...

These days about the only thing I get into is Nuts & Volts,  and even
there I'm only interested in some of what they're publishing about. 
Robotics isn't one of those things,  at least not at this point in time
though I guess some of the technology involved in it is a bit interesting. 
And this one series of articles on a laser setup was quite something else
-- liquid cooled!  The guy who wrote this stuff spent *way* more money on
his setup than I'd ever consider spending on something just to fool around
with,  though he did give excellent examples of how an industrial-type
control system with all sorts of safety interlocks should be designed.

 GM> Check out their internet site at siliconchip.com.au and you will
 GM> find a table of past articles.

I'll put it on my list.  

 RJT> I guess he wanted to build some serious machinery,  or something.  Me,
 RJT> I'll be happy if I can get something of a "tabletop"
nature going --
 RJT> something that'll mill circuit boards for me,  so I don't have to mess
 RJT> around with applying resist and etching with chemicals.

 GM> Yup, me too!

It shouldn't take much,  particularly if I can use the mechanical aspects
of some old printers.  Got this big old daisywheel sitting down in storage
that I have absolutely no use for,  but it oughta make one heck of a
starting point,  for at least one axis...

 GM> Bronze is usually considered as a better bearing material, and
 GM> sintered bronze bushes aren't all that hard to buy off the shelf!

 RJT> I haven't the foggiest idea where I'd go for something like that
 RJT> offhand.  I do have a bunch of these nicely smooth steel rods that came
 RJT> out of old printers,  and they did use bronze or something pretty close
 RJT> to it to slide on those,  but I don't know that I kept any of those
 RJT> bits.  It'd sure save me some trouble if I could do that sort of
 RJT> thing...

 GM> Yep. The bits that come out of photocopiers are even more 
 GM> substantual than the printer bits, especially if you can find one 
 GM> of the really old machines where the master page was slid back and
 GM> forth on a pretty substantual glass table!

There's a thought.  I don't know where I'd find such a thing,  or where I'd
put it while I dismantled it,  but I have some ideas now that you mention
it...  At least as far as the obtaining it part,  there's no way in hell I
could get one in here.  OTOH it might make a worthwhile project for a bit
later on...

 RJT> I think mine was around $60,  probably fairly comparable?  I got it at

 >>> Continued to next message...

Ok.

---