Printable View
Hi im jenifer,
Im doing a light dimmer with a PIC16f877.
In my code i need to detect the rising edge of the scare signal thats enterring in my PIC. I don't know how to control the intensity of my light. I know that i need to change the delays to change the intensity.
Any ideas on how to do all that?
thanks alot :)
So, I take it you are using pulse-width modulation?
What's a "scare-signal"?
--
Mats
to detect the rising edge can be done by using external interrupt pin...
you can set it to positive(rising up) or negative(going down) triggering...
thanks auralius!! i can detect my rising edge now!!
Now what do I have to write in my code if I want to change the intensity in my light? I have one output connected to my light
So, what I think you are referring to with your "add delays" is something called pulse-width modulation - it is a digital method of controlling "voltage" by simply giving full voltage for some time, and then no voltage for some time. Depending on the proportion of on-time and off-time, it will appear as if you have a proportion of the full voltage. E.g. we want to give 25% of the full voltage, we have an on-time of 0.25 seconds, and an off-time of 0.75 seconds. Overall that will be 25% of the voltage. [You may want to use shorter times than that - probably best to stay under 20 milliseconds / 50 Hz to avoid flicker effects].
Basically, what I'm saying is that if you flash (blink) the light quickly enough, you get the effect of a dimmer.
So, let's say we have 20 milliseconds as our 'time-base', then we could make 25% light by having a 5 ms on-time, and 15 ms off-time. 75% of full light would be achieved by 15ms on, 5 ms off.
--
Mats
yes you are right mats. thanks!
I have to put some On, Off and delays it should work!
something like that would work?
..Code:#INT_EXT
interruption_EXT()
{
delay_ms(200);
output_high(PIN_D1); //on
delay_us(10);
output_low(PIN_D1); //off
}
Ｙｅｓ， ｔｈａｔ ｗｏｕｌｄ ｇｉｖｅ ｙｏｕ ｓｏｍｅ ９５％ ｄｕｔｙ ｃｙｃｌｅ 。
－－
Ｍａｔｓ
Of course, you'll need to put that in a loop. And, don't mix-up ms and us. (I'm not familiar with the PIC compiler or these special commands, but the concept looks good.)Quote:
Originally Posted by jenifer85
I'm not sure what you're doing with the interrupt input... Turning the dimmed light on/off?
I assume you have an LED connected? You're not trying to control an AC lamp, right? With an AC lamp, you do need to sync to the 50Hz or 60Hz line frequency, and the interrupt can be used for that purpose. But, then there are hardware considerations that would affect your software design.
yeah I have an AC lamp connected. thats why im using the interrupt input!
mmm i don't understand it's not working..
Im not able to make it work after multiple attemps..
Its only flashing the lamp always at the same intensity.. the intensity never change.
someone can help plz??
thanks
If you can see it flash, then the timing is far too slow. You need to use timing in the ten-twenty or so millisecond range to ensure that the eye doesn't see the blinking - and if I'm right that you are using pulse-width-modulation, then the idea really is to have it blink on at full effect and then go off again, but done so quickly that the integration circuit that is our eyes will think that it is "not full brightness, but not off either".
If the idea is not pulse-width-modulation, then I have no idea what you are doing, or how you are supposed to solve it.
--
Mats
yeah Im using the pulse-width-modulation. But when I put faster timing the lamp just don't do anything, it stays off.
what other delay values have you tried? try 10ms off and 10ms on and see what happens.Quote:
Originally Posted by jenifer85
You said you are using an AC lamp? I.e., an incandescent bulb? Such bulbs have an inductance, as well as a heat-up time, which will prevent them from turning on unless the current is held steady long enough. You can't "flash" an AC lamp at extreme frequencies.Quote:
Originally Posted by jenifer85
If the lamp is incandescent, the proper way to control its intensity is by varying the voltage, not pulse-width modulation.
the point is not to flash the bulb though. the idea is to have an average voltage that is less than the maximum by varying the duty cycle of the power supply. Also, a light bulb as an inductor is a pretty low-value one. it's essentially an air-core unit with many turns on a very small diameter. I can't imagine the inductance is much more than a few microhenries - an insignificant amount at 50-60Hz. In fact, assuming 1uH (1 microhenry), the reactance of the inductor would be about 3.77e-4 ohms at 60Hz, which is negligible. Electric motors are far more inductive than a light bulb, and very often have their speed controlled by means of pulse-width modulation as a means to vary the average voltage.Quote:
Originally Posted by brewbuck
pulse width modulation has the effect of varying the voltage by providing maximum voltage for a controlled duty cycle. assuming 120V, you would have 108V at 90% duty cycle, and only 12V at 10%. pulse width modulation should work just fine for dimming a light. your average wall-mounted light dimmer uses a form of pulse width modulation to accomplish its task, so there's no reason why it wouldn't work in this case.Quote:
If the lamp is incandescent, the proper way to control its intensity is by varying the voltage, not pulse-width modulation.
the point is to make the pulses short enough and often enough to avoid flashing and flickering.