Strange 16f877a behavior

[jacob.zurasky]

I am using the classic TAIT programmer with 7407 and PNP transistors. I recently ordered some 16f877a chips. The programmer used to work perfectly for 16f84 chips. The problem is that my code only downloads correctly and verifies only IF I am holding a probe of my multi-meter at the VSS pin on the PIC. (The meter doesnt have to be turned on for this to work or the probe plugged into the meter either) If the probe is not present, the chip doesnt program or verify. I have PGM tied to ground, and VDD, MCLR, PGD, and PGC are recieving correct signals from the programmer. Any ideas as to correct this? -Jacob

 

[ljcox]

What voltage do you read on the Vpp pin when it is programming? From memory, it should be between 12.5V and 13.5V.

If works when you press on the Vss pin, then you may have a dry joint somewhere or a crack in the PCB.

Can you still programme a 16F84? I know it has an 18 pin socket, but you could solder wires onto an 18 pin socket and insert them into the appropriate pins of the 40 pin socket.

 

[Pommie]

Check Len's suggestion first. If it still doesn't work try downloading WinPic. This implements the various timing requirements better than other software and can sort out funny problems like this.

Mike.

 

[evandude]

What kind of decoupling capacitors are you using?

One of the effects of bringing a multimeter probe (or your finger, or anything like that) close to the power pins of a PIC is adding additional capacitance. If you don't have enough capacitance on there already, this can sometimes make the difference. at the very least, you should have a small ceramic cap like 0.1uF directly across the power pins as close to the PIC as possible, and even better, have one on each side of the PIC, and a small electrolytic across the power rails elsewhere.

Probably 3/4 of the times that I have experienced inconsistent/unstable programming like this, it has been due to forgetting sufficient decoupling caps, most importantly the small ones directly next to the PIC.

 

[jacob.zurasky]

Thanks guys for all the suggestions. Vpp was at 13.2v and the other lines were at 5v when they were supposed to be. I had a 0.1uf cap on each side of the chip. This problem seemed to remedy itself when I added a LCD display and a few other components. I figured it must be a capacitance problem since adding the probe helped....weird though. Thanks guys! -Jacob