Use of PCNT

Alan,

The PCNT instruction sets the 'Z' flag whenever an edge is detected.  

So lets assume you might do something like this:

L22 PCNT {next=L23,hr_lr=HIGH,reqnum=0,request=NOREQ,type=FALL2FALL,pin=5,period=0,data=0,hr_data=0}

LC1 BR { event=NZ, cond_addr=L23 }                            ; Skip to L23 unless PCNT detected edge

LC2 DJNZ {reg=NONE, data=5, cond_addr=L23}                    ; Count 5, 4, 3, 2, 1, 0

LC3 MOV32 { type=IMTOREG&REM, reg=NONE, remote=LC2, data=5, hr_data=0}; Reset DJNZ to '5'

LC4 ADD {src1=REM, src2=ZERO, dest=IMM, data=0, hr_data=0, remote=L22}; Copy data from PCNT instruction 

LC5 ... Do you want to notify the CPU that data is available in LC4? 

L23 ... your code continues here

The "ADD" at LC4 which is really a 'MOV' of sorts, is there to make a copy of the PCNT result.  If your CPU is responsive & fast compared to the time between each pole, you might skip this and have the CPU pick directly from the PCNT.  Otherwise the CPU can read the data from LC4 instead of L22 and it'll be valid for much longer.  

Another option could be to average the last 6 values captured - if you want to do that instead let me know.

Note that PCNT is not valid for the initial measurement, because on startup you assume you're starting your PCNT asynchronously so the first edge's results are just a partial period measurement (depending on where you started).

Didn't show anything in the above code to handle that case intentionally, however counting every 6th pulse and grabbing pulse 5 not pulse 0 does have the nice side-effect of handling this case in the N2HET rather than in the CPU.