> Thanks guys, you both got some valid points, I agree ICE 2000 price is
> ridiculous, especially for a small business man like me. I am a UV pipe
> liner ("HI! Quentin!") and are about to step onto Flash with the 16F87x,
> thus my query about the ICD. After the blurb by the Microchip rep., I
> just wanted to get some 'real' feedback from users before I buy.
>
> I am placing my order on Monday. Thanks.

We have found the ICD works well, barely a hiccup EXCEPT that the
software complains about comms errors that dont seem to be errors at
all. We have got used to just ignoring these error messages and all
works fine. Our first project development using it (16F877) has been
at Auckland Zoo feeding Sealions for 3 months now without a hiccup.
A really nice feature of an ICD is that once everything seems to be
working you can unplug the header comms link and the 16F877 carries
on working, great test of stand-alone operation. The software
complains horribly on the PC and MPLAB really needs to be rebooted
when you reattach the lead but thats no big deal.

The ICD is a pain when it came to serial comms debugging, we
had a bug where the RX was writing over the buffer and we had to use
our ICEPIC running 16C74 emulation on to catch the problem. That was
further complicated by the bug in our ICEPIC that put low level
digital clock noise on the I/O lines.
The delay was hideous finding that bug and we decided to go with the
MPLAB-ICE 2000 but the 16F87x module isnt out yet, so we can only use
the 12C67x module or the noisy old ICEPIC (16C7x and 16C/F84) for
now.


_____________________________

Lance Allen
Technical Officer
Uni of Auckland
Psych Dept
New Zealand

www.psych.auckland.ac.nz

_____________________________