Myke Predko <myke@PASSPORT.CA> wrote:
>
>It looks like an interesting circuit - maybe you could give an explanation
>of how it works - According to Microchip's programming specification Iddp
>(PIC Current while Programming) is 20 mA and Ipp (Current sunk at _MCLR
>during Programming) is 50 mA.
>
>I'm surprised that an RS-232 port can source this much current.
>
>Good work, I'm interesting in learning more about it.
>
>>I am working on a RS232 powered Production Quality Programmer, look at
>>http://www.ebar.dtu.dk/~c888600/newpic3.htm. Does anyone like to make /
>>sell PCB for this programmer?
>>
>>Regards,
>>Jens Dyekjfr Madsen

For anyone whose email reader gets a "not found" response on this URL,
delete the final stop to give http://www.ebar.dtu.dk/~c888600/newpic3.htm
Always a problem when quoting Web addresses, as they are often at the end of
sentences (and we get told off for leaving off our full stops, don't we!)

As a purely personal viewpoint, however well this works, I wouldn't want to
use it because "Gnd" isn't system 0V ground.  This means that there are all
sorts of problems if you want to test what's going on (have to use scope in
differential mode etc).  Some years ago I spent ages sorting out why a (high
cost commercial) PLD programmer was failing programmed devices on valid test
vectors.  That programmer drove the "0V/Gnd" IC pin from one of its battery
of DACs, relays and analogue switches, and it was an absolute pain getting
usable signals.  I know I could have floated either the programmer or the
scope, but that introduces other problems (including that the floated unit's
"ground" can end up at halfway between the mains live and neutral
potentials).  Where Jens' design will be fine in this respect is where the
serial port is on something like a battery-powered laptop or Psion.

How about producing positive and negative rails from the three driven input
lines, then running a simple dc-dc converter to get +Vpp from the negative
rail?  I'm sure there's a Maxim or TI part that would suit.

Final point while I'm chatting about RS232.  Lots of people have referred to
Maxim parts.  These ARE very good, but don't forget Linear Technology makes
a similar range.  Pinouts of the LT parts may make PCB layout easier, since
they tend to be "side to side" whereas MAX tend to be interleaved.  There
are differences in other aspects such as Vcc noise generated, RS232 drive
voltage when all channels are running at full speed, etc.

Tim Forcer               tmf@ecs.soton.ac.uk
Department of Electronics & Computer Science
The University of Southampton

The University is not responsible for my opinions