Homicidal 877 user needs help

[BobK]

Hi David,

Maybe we are getting somewhere afterall. You stated that you put in a 10 pin header to allow ICSP. Did you also add the diode to block the voltage from going to the rest of your circuitry? I use ICSP to program a product I make and I do not have any +5vdc connection between the programmer and the pcb. I only use the mclr, ground, RB6, and RB7 pins. When I did another project several years ago with an 877A I also had problems doing ICSP so I gave up on the idea. I have since worked out the problems by just using the connections I gave above. I have read in several places that the PGM pin should be grounded during the programming process.

Do you have a simple breadboard that you can setup a very basic 877A with the mclr to +v thru a 1k resistor, put a led/resistor on RB0, add the 4mhz resonator and then test out the blinky program? This could help isolate whether its the PIC or your boards. You should be able to create the blinky program without adding any config settings. Just use the one that's in the MCS program.

HTH,

BobK

[Jumper]

Hi,

Could you post a picture of your HW and maybe a simple schematic. There must be a reson for the PIC not to work properly and we all wants to help.

A simple schematic will help to provide us an understanding of what levels you have where and if we compare these with what we expect them to be we might be able to spot the problem.

What does RB6 and RB7 do after you have ISCP-programmed the PIC and disconnected the programmer? I have seen of the 877A that if RB6 and RB7 are held at an unfourtunate level/timing the PIC will refuse to start up properly. I suspect this to be related to the VDD rise and then the PIC will try to enter programming mode even when a programmer is not connected.

I have 2 877A:s connecting with each other using a cable on RB6 and RB7. And if I set the pins from one PIC in a special way the other PIC will refuse to power up.

Are there any power applied to any pin of the PIC before VDD is applied?


/me

[mister_e]

and a .HEX file of your PIC program, so we can confirm your configuration fuses are set correctly. If you set them in your programmer software, you should be able to save the .HEX file, so the next time you open it, those should be set as you did.

also check if your software is set to program the configuration fuses.. yeah it happened.

[mackrackit]

The ICSP might be it.
I use a PICSTART PLUS for ICSP. Just tried a 16F877A.
All works well when the programmer is powered. If I leave the programmer connected with the programmer power off the PIC will not start.

HTH

[DavidFMarks]

Thanks all, I am assimilating your comments and will respondwith code details etc. asap. However we are shortly seeting off for a pre Christmas long weekend in Brussels and senior management is going to get difficult if I am caught at the keyboard much more. Thanks again will be back on line asap
Regards
David

[b1arrk5]

David,

I went through almost the exact same scenario almost two years ago, and after losing a little too much hair I discovered that the problem lay in my epic programmer, the one that uses the printer port. Windows XP was playing havoc with the printer port on occasion. I upgraded to the newer USB programmer from MELABS, and have not had a hardware problems since. I still get the occassional software glitch, but as those are all self inflicted............. Anyway, perhaps someone a little closer to you might offer to program a 16f877 for you, and see if that works. I would be awfully suprised to learn that you have 6 defective chips and three defective development boards, my lab-x boards run continuously, and have never been a problem. The programmer seems to be the only thing that is common to all your problems, as you said you should be able to blink an led without any difficulty.

Jerry.

[DavidFMarks]

Few quick words before heading for the channel tunnel. Built a quick breadboard just the four power supply connection, resonator 1Y pullup on pin 1 and and led on RB7. With the software recommended earlier in this thread ... nothing... except, as once before when using the crystal if I touch pin 13 then the led flashes, trouble is its flash rate doesnt seem to have much to do with the program !! Jerry you could be right about the epic, I seem to recall having problems with it in the past I will look into it when I get back from the Brussels trip.
Cheers everyone
David M

[Darrel Taylor]

<table><tr><td valign="top">Probably just another stupid question on my part.

But, are you sure the Power Rails are connected across the entire breadboard?</td><td>
<img src="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=2204&stc=1&d=119758760 2" /></td></tr></table>

[nomad]

There's a fix for printer port issues and epic.
from melabs download section:

Stop Windows XP from polling printer port
Windows XP may periodically access printer ports and interrupt the EPIC™ programming process. This can be stopped with a registry entry.

Download the registry entry file below and merge it into your XP registry.

Download XP_stop_polling.reg (1K)

http://www.melabs.com/support/patches.htm

it programmed and verified ok, there's something wrong with the code or hardware, be it config fuses or tris or icsp left connected.

As far as programming and verifying ok, lol, I just had a hell of a time with my led project. thought I fried the pic, programmed and verified perfectly. spent 3 hours only to find I goofed and set the trisa and trisb to 255 (all inputs) oops! inputs aren't going to make anything blink. lol don't they default to inputs if tris isn't set?

my programmer, if left connected puts -7volts on all the pins.. definately not a good idea.

[DavidFMarks]

Thought the osc had to be set to intrc to use the internal oscillator, now realise it should be set to XT. Have now written simple program to switch on all port b pins, pause and swithch them off again in an endless loop. The program containsthe line cmcon = 7 can't remember why but I recall having to do this about a year ago when I wrotethe original star program. The osc isset to XT it compiles and produces the attached hex file but fails to flash the ports