PWM with CTRA, variable on time, off time?

T Chap

I need to control something with PWM, and will need to experiment with the on time and off time.

I have been testing the CTRA but can only find methods to adjust the duty using 001100, making variable on and off times not possible. Maybe I have missed something.

The PWM.spin object does the job, but requires a cog.

Is there any solution for making a counter accomplish this goal without the assembly object and additional cog?

Thanks

Mike Green

You can't do variable on/off times with just a counter. You need to have a cog involved.

Phil Pilgrim (PhiPi)

Originator,

No, not really. For a counter to generate true PWM on its own, it needs either to reload PHSx from two different sources or reload from one source and trigger a second counter for the alternate phase. The Propeller counters don't have either capability. Consequently, unless DUTY mode is adequate, PWM needs to be mediated in software. For even moderate PWM rates, this entails assembly code.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!

T Chap

Thanks guys.

Tracy Allen

I think yo can enlist both CTRA and CTRB together to the task, both addressing the same pin. Both run at the same frequency but with offset phases. The counter outputs are OR'd, so the pin is high when either is high and low when both are low. The range without further tricks is either 50% to 100%, or 0% to 50%, depending on how you look at it. A choice of either with differential output.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

T Chap

That is a good idea. I will need smooth ramping from 0% duty to 100%, 100% - 0%, adjustable Freq, which seems to mean either use the PWM obj or build a custom loop that just controls the pins.

Post Edited (Originator) : 9/25/2008 5:43:44 PM GMT

Phil Pilgrim (PhiPi)

Tracy, that's brilliant!

-Phil

Addendum: The tricky part will be setting the relative phases in Spin. It can probably be done by doing a WAITCNT, clearing PHSA, then doing another WAITCNT before clearing PHSB. The latencies between the WAITCNTs and the clearing of PHSx should be equal. The relative phases will depend on the programmed delay of the second WAITCNT.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!

Post Edited (Phil Pilgrim (PhiPi)) : 9/25/2008 6:05:40 PM GMT

Tracy Allen

Thanks Phil, but I must add that that PWM solution has come up here before. To return the complement, it is nothing compared to the six counters you enlisted for in the homodyne radio receiver!

Originator, with an external XOR gate and two propeller pins, it could do the whole 0-100% range.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

TJHJ

On this note, Does anyone know what the maxium frequency the PWM object is capable of?

Thanks TJ