PWM Hardware problem???!!!

Started by power_overwhelming_4 October 13, 2004

So what basically is the solution? We tested the BX24
over and over with the LEDexample and PWM example
program and it registered inconsistent results.
LEDexample shows pulses from pin26 whenever it is set
to high, so the pin can output. But PWMexample says
otherwise, no pulse trains appearing on the pins. We
concluded that the pwm functionality of the chiip is
in question. Is there any diagnostic program anyone
could suggest that could determine the problem?

ALso i ran the PWMexample program, then i tried to
short pin26 to pin27, the green led began to blink. So
i think there is no problem with the connection
between pin 26 and the green led. Am I still making
sense =)?

I've tried both suggestions in this dual PWM thing. I did what Tom
said which is to initialize pin 26 in the BX IDE (Project menu ->
Chip) as low output, but nothing changed. Only pin 27 has pulse
trains from PWM and pin 26 still has none.

I also tried Chuck's suggestion to "hack" the chip by cutting pin 26
from the green led, anyway, i only disconnected the green led from
the pin. I expected p26 will work now afterwards, but still it

I also tried to tap from the AT908535 pins directly. I also
discovered that the Atmel datasheet also fed us with wrong
information on the pinouts of OC1A and OC1B. Anyways, i eventually
found the correct OC1A and OC1B pins from the Atmel chip and its
position is very different from the specified in the datasheet.
Using ohmmeter, i tested the OC1A and OC1B pins if they are indeed
connected to the header pins, and they are. The OC1B pin (in 27) in
Atmel chip is working, but OC1A is not working. It's strange since
just last month the PWM pin 26 is working. Are there any more
suggestions on how to remedy this problem? Do i really need to buy
a new BX24 chip? The dual PWM is really critical. Help!!!