saving RCREG to word

[BrianT]

Hi Magman,
I did not optimise the code since I had plenty of time in the way I did it. It may turn out that your pauseus approach may save space and time. If I needed code space or clock cycles I would have checked that method out.

I reckon a single checksum across all channels uses less time and code than the 100% redundancy approach you are taking with the true/inverted signalling but I agree yours is a safe and simple approach. If I get a bad packet or a bad channel within a packet (out of range or delta too big) I use the last good value for that channel. Given that real data does not jump much from frame to frame this method seems to fly quite well.

My uplinks are 36 and 2400 MHz using the transmitter in PPM.

What are you building. Is it an aircraft?

Cheers
BrianT

[Macgman2000]

Looking over your code closely, I realized something. Do you pause in the TX output for 16 ~ 20ms before starting up your next frame?

I am not doing that, it may be why I am seeing the results I am seeing. using a continuous back to back output from the TX @ 4800bps I am able to get 4 servo + 2 hardware PWM control. BTW I figured out why my servo is a bit coarse. in your code you multiply with a 2 constant. I am not moving in 2us increments. I am doing the following: 2* (4*byte + 500), So for every crank on the potentiometer I am moving in chunks of 8us, so I am not loosing packets, its my scaling factor....ggeeeezzz.


yes, it is an aircraft. Actually VTOL.