That was the way I had it at first, but I decided to try it connected.
Not sure that it would fry the chip, but I'm using a fresh chip anyway.

New results...

I've simplified the ckt and narrowed the problem down to one of...
(1) poor thru-hole connections (home-built board w/o eyelets),
(2) some component fried/dead,
(3) incorrect transformer design/usage.

For (1), I've re-built the PCB as a single-sided design (with a minimal
top layer actually), and using 1 "explicit" wire to connect the sides.
I'm VERY confident that this is not the problem now.

For (2), when I re-built the circuit, I tested almost every component
(caps, resistors, and the schottky diode) outside of the circuit.  Also
using a very new LM2588T chip.

For (3), I've re-verified everything -- using a T130-3 (gray) core with
23 turns on the primary and 22 on the secondary.  I'm winding with the
same orientation (start over the core, going down thru the middle, and
moving counter-clockwise as I wind the turns), with the start of each
winding marked as the "dot".

Latest results get me 1.3V at the output, with the LM2588 getting super
hot till the point that it burned my finger.  I should be getting a cool
12V.

I'm desperately trying to get my hands on an oscilloscope,
but outside of that, I'm stumped.

Cheers,
-Neil.





> -----Original Message-----
> From: pic microcontroller discussion list
> [mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Larry Reynolds
> Sent: Saturday, March 01, 2003 9:10 AM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: Re: [EE]: DC-DC converter woes?
>
>
> Pic Dude,
>
>         The data sheet says very clearly on Page 23 to leave Pin
> 1 OPEN!  Do
> not ground it, nor tie it to supply.  It CANNOT source or sink any
> current!! You may need a new device....  :-((
>
> Larry
>

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.