Alejandro Lavarello wrote:

> Hi, folks!!!
>
> Well, days ago I have total success making a "do-it-yourself" ICD.
> I have followed the schematics from Patrick Touzet, and it works OK!

I built mine using the schematics in the user guide, (which is on the Microchip
CD rom along with the firmware required) on a piece of veroboard. After a bit of
fiddling it worked well.

> Some things that I have encountered:
>
> 1) Communications: The ICD is formed like a DCE equipment, and needs a
> RS232 cable for computer-to-modem (NOT null-modem). I need some try-and-error
> before making a correct cable (this ended vey simple,  a DB9 female to DB9
> male straight cable).
> I change the capacitors attached to the MAX232 to 10 microfarads,
> and works OK (100 nanofarads is too small for some kinds of MAX232).

I also changed the caps to 10uf tant. beads

>
> With no computer attached,the LED blinks
> at 1Hz. The LED is fixed "on" when proper communications are established.
>
> 2)Voltages: I have canged the supply 100 microfarads capacitor for a 1000
> microfarads
> (the power has drops with 100 micros).

I didn't have this problem which cap. are you referring to ?

>
>  The LED must be a green LED. It needs to have a votage drop of 2V,

My led is red ! However I had to try several from the junk box to find one which
worked.

> because
> is used as voltage reference. One time the communication with ICD was
> established,
> one needs to adjust a little R1 and R21 for the correct reading
> of the VDD and VPP voltages.
>  VDD can be measured with a multimeter, then adjust
> R1. VPP can be measured using an oscilloscope; when one touch the button
> "Self Test" in MPLAB, the ICD generates high voltage. The point of measure is
> the union of R21 with R17 and R18 (between 12 to 14 volts).
> Anoter method can be adjust R21 same proportion than R1 (for proper reading
> of VPP).

Initially I had problems getting a high enough Vpp. I found changing Q4 2N2222A
for a proper switching transistor - ZTX650 ( had a few spare - they were used in
ZX spectrums, switching a coil to produce -Vdd for memory IC's they failed often
when people plugged in expansion packs without disconnecting the power!) gave me
a higher voltage and no need to fiddle with resistors.

>
>
> 3) Crystal: The crystal of 3,6864 MHz is very hard to find.

I *think* I bought mine from RS components.

> Yes, I know, it
> allows exact
> baud rate generation, but:  why Microchip do not have used
> a much more common  3,58MHz (or 4MHz) crystal? Finnally, I have build a
> separate
> oscilator using a 7,3728MHz surplus crystal and a 74HC04 plus 74LS74 as a
> divider.
> (ugh!! ugly but works! I was lucking obtaining this rare crystal from an
> old RS232
> PC add-on card).

My ICD has been working reliably for several months now. I now wonder how I
managed without it.

I just received a mailshot from A microchip distributor here in the UK. They are
offering the ICD for 66 UK pounds, including an F873 while stocks last.

So its now hardly worth building one.

>
>
> The Touzet's site is:
>
> http://www.multimania.com/silicium31/
>
> more exactly:
>
> http://www.multimania.com/silicium31/Electronique/PIC/free_icd.htm
>
> I am interested in read experiences of other peoples that have build
> ICDs.
>
> Cheers!!!
>                 Alejandro    alejol@adinet.com.uy
>
> --
> http://www.piclist.com#nomail Going offline? Don't AutoReply us!
> email listserv@mitvma.mit.edu with SET PICList DIGEST in the body

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