zero crossing detector 2VAC

ckielstra · Joined: 18 Mar 2004 Posts: 3680 Location: The Netherlands

Aside from your detector problems:

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

You are correct in that this is a bad approach to checking the inputs...
I do know the difference...

I understand perfectly what you are saying regarding checking the inputs. This was just scrapped together and I obviously did not look at it.

Regarding setting up a timer, I have never had much luck getting this to work correctly (any pointers would be greatly appreciated)...I know how to get a timer to overflow, but the way I am seeing this is that I need to set pin_c2 to low every 1/60 sec. (16666.67 us) for a short period of time, then set it to high for the remaining part of that 1/60 sec.

Am I correct on this?

I was thinking that I would have to invert the edge detect, which is no problem, but working at say 500 kHz clock, how can I get this to overflow at the correct time?

I have removed all of the delays also...

I am thinking the best way to handle proper timing may be to write it up in assembly.

I would rather not use INT_RB, but it is definitely an option. Is there a benefit of using this rather than INT_EXT ?

I had thought I would be able to use INT_EXT and switch the edge detect.

I realize there is alot that I need to learn about programming, and where I fall short is definitely in the area of working with timer interrupts. I never have much luck getting the correct timing. Even trying others examples, I seem to find a much different timing when looking under a scope.

Anyways, Thank you all again for the help. I would really like to find a way to use a PIC and control the speed of an AC motor.

I feel that using a 1Mohm resistor is cheaper and easier for doing a edge detection, and If I could simply get the interrupts to work I can be finished with this part of my project.

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

Here is an updated code:

Ttelmah · Joined: 11 Mar 2010 Posts: 19260

In INT_RB, you _must_ read port B.
The interrupt won't clear till you do.
You disable the interrupt, but then enable it on the next loop in main, which wil _immediately_ jump to the handler, since the interrupt will still be set.
Since you are still delaying in the interrupt, read the port as the last thing in the interrupt handler, so that it is cleared at this point, rather than earlier, and possibly becoming set again....

Best Wishes

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

So you think this would help:

Ttelmah · Joined: 11 Mar 2010 Posts: 19260

Yes.

In the demo code as shown, the delay ought to work, _but_ it is likely to cause problems to 'bite you' later, if the code gets more complex.
If you have a delay operation, in the external code, interrupts _will_ be disabled for this. Potentially a problem.
If you use other interrupts (RS232 etc.), these can't be serviced while you are delaying in the first interrupt. Potentially another problem.

As I said before, far better to use a timer. Program this to 'tick' in some interval that suits you (master clock/8 say). Then in INT_RB, output low your first pin, load the timer with the constant needed to make it run for the period you require, clear it's interrupt, enable it's interrupt, and exit immediately without a delay.
When the timer interrupts, turn off it's interrupt, and set the second output pin. Exit immediately.

This way the hardware timer, handles your timing, rather than you sitting waiting, and your code outside, can be doing other things...

Best Wishes

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

Thank you!

I will try this. I have never had much luck getting the timers to interrupt at the correct (calculated) time however, and since this needs to be consistent I am not confident I can get it to work, however I will try.
That approach would be much better.

I do not plan on adding any additional code, and this PIC would be dedicated to only AC speed control.

I have tried the code and still cannot get it to work well however.

Thank you again.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

I'm curious about how you get 500 KHz as your oscillator speed.
What's your crystal frequency ? (and your compiler version)

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

the 500 kHz seems to be fast enough for me...

I tried a 4Mhz, but found that 500k is a good speed.

is there any difficulty with that speed?

I am programming with 4.093 pcwhd compiler

i am not using an external clock...

i do have the AC line wired to t0CKI with a 2Mohm resistor and diode to Vdd, but am not using it in the code at the moment.

is there a better way to just use this input as an oscillator and use T0CKI??

I cant seem to get good interrupts using timers, and i know there is alot I have to learn in this respect, but I would just like to use a PIC for a simple fan speed controller. I dont need to be as accurate as an FOC, and i thought about going to an IGBT driver (if anyone has any recommended P/Ns i would appreciate that).

I tried using an optocoupler for my zero-crossing, and may come back to that soon, but for now I am just connecting the AC straight to the PIC using a 1Mohm resistor.

I feel instead of going low for a short period of time, then delaying, then going high... i should switch this and go high, then delay, then go low (??)

i currently have this PIC connected to a Crydom 2410 SSR.
is that a bad choice? i know it is more expensive they making a circuit, but i can afford it for now and dont want to spend alot of time on this part of the design.

Thanks again.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

deperkin · Joined: 04 Feb 2009 Posts: 83 Location: PA

So from what I understand now, which i did not know before, is that I should remove the XT ?

would I then need to set the other bits?

Thank you btw, this is helpful.