PicForum

You didn't enabled the IOCIE bit before the sleep command.

drh
The IOCIE bit is set in the Initialization. Does it need to be set just before sleep?
Jay

Nope. You configured the polarity but not the IOCIE. Unless the code you posted is not current. If you can monitor the OSC2 pin with a scope, you can verify the clock stops when it should be sleeping.

drh
INTCON = 0b11001000; //set GIE and PEIE so TMR2 can interrupt IOCIE set
Jay

Okay, I would have done it differently. You also need to clear the IOCBF4 flag after you wake up. I would also clear the enable bit also, and re-enable just before "attempting" to sleep. Based on a previous post, you may not be entering sleep.

drh
I may have found something. When I reset GIE things stop. This is my testing using the RedSig being added at different points.
Jay

Code: Select all

void start_train(void) {
    RedSig = 0; // Red off                                                         
    GrnSig = 1; // Set Green
    __delay_ms(2000); // for about 2 sec.
    Sound = 0; // Turn sound on
    __delay_ms(100); // for about 0.1 sec. short pulse required
    Sound = 1; // Sound off
    INTCONbits.GIE = 0; // prevent branches to the interrupt address (0004h).
    INTCONbits.IOCIE = 1;   //Set IOC enable
    RedSig = 1; // This happens [b]Test 1[/b]
    SLEEP();       // Put it to sleep wakeup on busy pin going high RB4
    RedSig = 0; // This happens [b]Test 2[/b]
    //INTCONbits.GIE = 1; Test 5 commented out, program steps thru but rest will not work without GIE
    __delay_ms(1000); [b]added just in case[/b]
    RedSig = 1; //This does not happen [b]Test 4[/b]
    INTCONbits.IOCIF = 0;// Clear the flag
    INTCONbits.IOCIE = 0;// Clear enable
    RedSig = 1; //This does not happen [b]Test 3[/b]
    START = 1;

drh
I moved the INTCONbits.GIE = 1; to after the call to start the train and it appears to work. But does not cure the static on the sound. May have to go to a separate supply. The sleep is acting as a delay until the recorded sound is finished. That a good thing. I will leave it in.
Jay

grandpa jay wrote:drh
I moved the INTCONbits.GIE = 1; to after the call to start the train and it appears to work. But does not cure the static on the sound. May have to go to a separate supply. The sleep is acting as a delay until the recorded sound is finished. That a good thing. I will leave it in.
Jay

As an alternative to a separate supply, you could try decoupling it - say a 47 ohm resistor in the supply line to the MP3 device and a 100uF capacitor directly across the MP3 device supply

Roche
I will give that a try, Thank You. Another issue has appeared. When I program with the PICkit3, The first time that I run the program I get the delay start at the end of the sound as expected. When I run thru the rest of the program and do a restart, I do not get the delay.
Jay

I tried a 33 Ohm resistor, could not find a 47 Ohm. I think that it dropped the voltage to low? It would not play.
If I go to a separate power supply, what would be the best way to trigger the MP3 module? I am thinking a relay but not being electronic minded I am not sure. Could I use an optical isolater?
Jay