How instant are Instant Interupts?

[jmgelba]

OK, then let me throw some fuel on the fire...

<img src="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=3293&stc=1&d=123853748 1" /><!-- Attachment 3293 -->

This is the "Full control" version.
I've connected the pulse input to the INT pin of the PIC, for the delayed pulses.

The second input of the first NAND goes to another PIC pin (Zero Delay Enable).
When that pin is HIGH, Zero Delay is selected and the output of the flip-flop will go high immediately on an incoming pulse.
R1 keeps it disabled till the PIC takes over.

Another pin is used to "Set from CPU".
This is used to Set the output Flip-Flop manually from the PIC so it can respond to Delayed pulses timed by the interrupts. Taking the pin LOW will set the flip-flop. This should be tri-stated when in Zero Delay mode.

Clearing the flip-flop (end of pulse width) is done with the CCP out pin. Whether it's actually the CCP doing it, or timed interrupts.

Again, Purely theoretical.
<br>

Hmm, how would one use this for falling edge trigger? I'll be honest, I havent had time to study what you are doing here, but what I've looked at seems logical.
I just ordered a couple of quad nands and a few 40MHz resonators, and over the next week or so will have a chance to work with this.

[Darrel Taylor]

Hmm, how would one use this for falling edge trigger?

Several possibilities there.

If using Option #1, the 4th NAND gate is available and can be used to invert the incoming signal. Selection of Normal/Inverted modes would then be with a Jumper or SPDT switch.

If you wanted it to be selectable by the PIC, you could add a 74HC86 quad XOR. With the extra gates, you could put an XOR on the output too if you wanted the option to invert it from the PIC.
<br>

[Darrel Taylor]

Here's an update for the last post.
Two 74HC86 XOR gates have been added to be able to invert the input and output polarities.

<img src="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=3300&stc=1&d=123889536 2" /><!-- -->

I see that you specified "Trigger input is TTL 5V rising or falling edge, selectable." in the first post.
I must have missed that part.

I could be diverging from what you want to do.
But the problem is intriguing, so I'm continuing onward ... thru the Fog.

I think the circuit above gives you all the capabilities you're looking for.

I did a quick pseudo-program to see what it would take for a simple test without interrupts, and it was surprisingly easy.

If you can settle on a PIC to use ... preferably 2-CCP's (1 for the FOSC4).
Could probably come up with something that's close.

I love shooting things ... with camera's.

[jmgelba]

Looks like 18F4220 at the moment. I only have v2.43 so I am limited in which devices I can use. I cant find my reciepts or proof of purchase anywhere to get the upgrade.

I need a decent amount of I/O's. 2x8 lcd, at least 4 or 5 switches, status led's, logic gates, fet drive, lcd back light, adc for capacitor charge level/status.

I'm not a coder or really much of a hardware guy, but I am a power supply and LED guy. If only I could show you some of the stuff going on here.....

[Bruce]

I cant find my reciepts or proof of purchase anywhere to get the upgrade

If you bought PBP from us, we can look it up for you, and send you a copy of your original
receipt.

Most other distributors should be able to do something similar to help you get the upgrade.

[jmgelba]

Part of this project is obviously timing.

Say I have a pulse coming in at between 50 and 200uS and I need to match this incoming pulse to an output. Not really a problem.

IF PORTA.0 = 1 THEN PORTA.1 = 1

But what if that incoming pulse is longer than the 200uS and the output must be shut off after a predetermined maximum, lets say 1000uS?

IF PORTA.0 = 1 THEN PORTA.1 = 1 no longer works. TMR0 has to be triggered at the same time as PORTA.1 goes high, then needs to count to a maximum of 1000uS and make PORTA.1 = 0.
PORTA.1 then must be kept low until PORTA.0 has gone low.
When PORTA.0 = 0 TMR0's flag can be reset and wait for PORTA.0 to go high again. If PORTA.0 has gone low before TMR0 has set the flag, it must be reset and ready to count to 1000uS.

Does this make sense?

[Darrel Taylor]

If you're doing it with software polling, timers don't add much value unless you plan on doing other things at the same time. But all that really makes things complicated.

So here's a simple example that does the same thing with pauseus instead of a timer.

Delay = 1000

Wait4Pulse: 
    IF !PORTA.0 THEN Wait4Pulse     ; Wait for incoming pulse
    PORTA.1 = 1                     ; Start Output pulse
    PAUSEUS Delay                   ;   pause for pulse width
    PORTA.1 = 0                     ; End of Output pulse

Wait4PulseEnd:                      ; Make sure input pulse is finished
    IF PORTA.0 THEN Wait4PulseEnd

GOTO Wait4Pulse