Pulsin Stabilization

[Bobbo_ZA]

Hi All,

Below is my code for turning on and off the landing lights of an RC model, however, I find the lights turning on and off randomly.... ??? ... confused...

I'm using a 12F675, the pin connections as follows:

GPIO.0, 1, 4 and 5 have LEDs connected (landing LEDs on gpio.4) all via the respective Resistors.

GPIO.3 is connected to the RC input through a 1k Res (i've tried it direct)

The program responds as expected on the RC input however on either side of the switch (on or off) the lights randomly turn on (when they are sposed to be off) and off (when they are sposed to be on.)

There is nothing connected the GPIO.2, however I am not sure if tis "open" pin could be causing it.

I checked with the scope and the input signal seems to be stable, just dont know why the signal seems to be jumping within the pic, any ideas?

Many thanks
Rob

@ DEVICE INTRC_OSC, MCLR_off, PROTECT_OFF, WDT_ON, PWRT_ON, BOD_ON
CMCON = 7 ' Comparators OFF
ANSEL = 0 ' A/D OFF -- Port pins all digital
TRISIO = %001000 ' All output but GPIO3 = input
GPIO = %000000 ' All 0 on boot 
ADCON0  = 0
WPU = 0
DATA @1,$FF, $FF
'DEFINE PULSIN_MAX 3000

wing_led2   Var GPIO.5  
land_led1   VAR GPIO.4
tail_Led   VAR GPIO.0
strob_led   VAR GPIO.1   
servoin     VAR GPIO.3
p var word
'---------------------------------------------------------
start:
'---------------------------------------------------------
Pause 200 ' Stabilisation at startup
p = 0
High wing_led2      
High tail_led

PulsIn servoin,1,P
SELECT CASE p
   CASE is >160
      HIGH land_led1
   Case is <130
      LOW land_led1
End select
Low strob_led
pause 250
High strob_led         
pause 100
low strob_led  
GOTO start
End

[cncmachineguy]

Your Rx may be glitching, maybe set it up to wait for 3 vaild pulses before changing states of the lights. Now if the are staying on when they should be off, and staying off when they should be on, maybe you need to hit the rev for that channel in your radio

[Bobbo_ZA]

Thought about that too, tried it on a 2.4ghz set (which should be a more solid signal) and yet it still does it, also I would expect it to remain in the last state when the Rx is off yet they still randomly turns on and off again, :-(. I tried a 1k res to ground from GPIO.3 but then it does not work at all, :-(

Its ironic that something this "simple" is proving such an issue, probably spent 24 hours solid on this project and hundreds of rewrites but still not etting it, :-(

I have another circuit that i had done for another application where i used the same principles that works flawlessly, yet this just doesnt, only this one is on vero board.

[mackrackit]

An open pin is not a problem as long as you code it as an output near the beginning of the program.

You might try a debounce type of routine to hide false signals.

[Charles Linquis]

It might be helpful to take your unused GPIO.2 and connect it to the Rx of a RS-232 port on your PC, then use DEBUGOUT or SEROUT2 to send the value of PULSIN to the PC. It is even better if you have one of those little TTL-RS232 converters lying around and use that.

Serial ports make wonderful debugging tools.

[mister_e]

Motion seconded.