16F88 internal clock configuration

[Tom Gonser]

I am using the _INTRC-IO, and also tweaked the serin2 stuff, so I have 'data' coming across all com lines now. Problem is that it is just 'off'..

1. ) It is resetting. I have one routine at the front which only happens on power up. The data from this is showing up off and on.

2.) the math is off. Where a value should sa 4.9volts it reads 3.3. Were a GPS lat should be 47.xxx, it reads 65,511, or some such. Same with all the other values, so I am sure the loops are off.

I made a simpler program that just spewed data to the different com ports - out on RB2, RA0, RA1. This just said "text", #counter - and looped from 1-100.. that worked fine. When I put the full program in, it gives bad data.

I think the original program for the 16F876/a was a 4Mhz program (??).. Does running this at 8 Mhz mess that up? Is there some setting that is needed? Does not seem like there should be....

Thanks a bunch for your assistance - I've never 'ported' from one PIC to another. This program works great on the other one... but we need to be on the F88.. did not know it was going to be so hard.

Tom

[mister_e]

o.k now it seems more a hardware problem than a software problem. be sure your supply line is neat an un-noisy. 10uF tantalum+0.1uF must be place close to your PIC... AS close as you can.

AND be sure your MCLR pin is tie to VCC via resistor.

Enable power-up timer can be great. Also as Melanie previously said it can be interesting to test the 'oscillator frequency is stable bit' before proceed to any of your initialisation process.

look at the a/d register setting(ADCON and others)... i'm sure that they're different from the PIC you used before.

[Tom Gonser]

The hardware is pretty skimpy.. I have the F88 a 5.1v supply, and a resistor from MCLR to power. I have only power and ground supplied, and then serial lines out from the appropriate port pins. No caps no nothing.

Whats odd is that with small program.. like only a portion of the one I am porting over, it works. It wakes up, it does its routine, then it starts doing a simple read from data input, put data into array, and read data from array out to the three serial outputs. This works fine.

Whats odd is when I put my WHOLE program in place, it chokes.. It periodically shows the data, but when it does the math is always wrong, and it looks to be resetting as I see the initialize routine results now and then.

I don't know about ADCON as I am only doing serial data in on Port B1, and serial data out on B2 and also A0,A1. I am setting the serial data out on A1 to 4800b by using SERIN2, and the rest is 9600b. - But again, this seems to work fine with a smaller program...

Must be some timing setting beyond what I've had to set before.

TG

[mister_e]

so there's no analog to digital conversion on a.0 and a.1 ???

you must disable those analog pins to make them digital.. at least

ANSEL=3 'set a.0 and a.1 as digital input

and capacitor should be consider.

[Tom Gonser]

Set the .. ANSEL=3 'set a.0 and a.1 as digital input, thx. Probably a good idea, but did not have any impact.

Oddly as I test, I find that my setting for OSCAL 8 mhz or OSCAL 4 Mhz does not have any impact on the system..

I have tried:

OSCCON=%01101000 ' 4 mhz
DEFINE OSC 4

----

and in another test...

DEFINE OSC 8
OSCCON=%01111000 '8 Mhz


Something must not be getting written somewhere. I'd expect that these settings would have a big impact on overall system operation.

Is there some prescaler setting that I am missing?

tg

[mister_e]

prescaller have nothing to do with your code if you're not using TIMERs. Can you post your code here or it's a 'military secret'

[Tom Gonser]

This program reads a data sentence sent at 9600b 8,1.n. The sentence has two start bytes, then 34 data bytes. It works on the 16F876A on the test board from Microchip, but the plain old F88 on my breadboard is acting strange as the thread indicates..

TG