JDM Troubles (ICSP, clock pin 39)

[Jidis]

I know you guys hate this thing, but I'll give it a shot anyway. ;)

I built a core/jdm hybrid, on which I've installed a two way switch, using a jumper block. The jumpers just flip between the normal core connections and what you would get if you connected it to a JDM with the ICSP connector (I've checked all the connections on the switch block for proper continuity).

It doesn't appear to be reading chips properly (can't verify the received buffer against what's really in the PIC). On my usual JDM laptop running Win95, the voltage checks come out as:

Empty Socket

-------------

MCLR: 13.01V  /-1.81 when off

Vcc (pin 32): 5.09V /0.00 off

Clock (pin 39):  -0.31V  /0.68 off (?)

Data Out (pin 40):  4.60V  /0.34 off

While reading the  PIC

---------------------

from pin 12 to 32: 4.85V

from pin 12 to MCLR(1): 12-12.5V

Does it seem like the clock is my problem? Everything else is there, just a bit low I guess. The clock pin makes it to the anode and cathode of those two diodes (D1&2 in the schematic), and to pin 7 of the sub, and I checked the two diodes in circuit with a DVM (OK according to that). What else would you suspect or check for a "pin 39 problem"?

Also- That 10k resistor (R1) from Vss to pin 36(RB3) in the schematic puzzled me. Do you just leave that out for ICSP?

BTW- I've got the sub shell tied to pin 5 as recommended. The two zeners were pulled from a JDM, which I've used many times with this laptop, although it was never used for programming over the ICSP block.

Much thanks and let me know if I left anything out,

George

[Jidis]

Good news last night with this thing. :)

I had gotten sick of it and left it for a few days, and got stuck waiting for something at the studio last night, so I went back and re-traced all the JDM parts of the board. Turns out there was a pretty serious error in the Vss area. The back of diode3 was tied to the negative of the two caps, and the backs of diodes 6&1 were tied to each other and to the Vss lines, but the two groups of parts never made it to each other. :-[

Everything seemed good after a short jumper wire was mounted. I didn't get a good verify of the PIC's current (unknown) contents on the 486 laptop I used to use, or after moving it to my home desktop (XPsp2), but on that same machine trying out WinPIC800 for the first time, I got a perfect read/verify on the first try. Looks like it should do OK.  --> oh, and the voltage checks were the first thing I went for after the mod (all good now)

Take Care

George

PS- Not as ashamed as usual on the missing connection. The jumper block made the layout sort of a PITA. Because of the "5-pole double throw switch" function that the jumper block is serving, a bunch of those connections aren't initially linked without the jumpers installed, and they change destinations if it's in JDM mode.

Considering the area that the error was in, I guess I still screwed up nonetheless. ;D