Hey all,
im building a small circuit with an ATmega32a to control a 1amp 12v LED with PWM. This is my current circuit:

Does it look ok? I added the capacitor from the BC547 base to ground since i was getting a little bit of voltage though the emitter when the micro-controller was off, making the LED slightly dim instead of off. Should I be doing that?

The circuit does work fine, just want to make sure it wont die soon or if there's a better way to do it.

Cheers

 

You could eliminate the 5V supply if you supplied Q2 with 12v also.

I believe C1 is taking some noise to ground for you. Reducing R1 would probably eliminate the noise also, unless it's coming from the ATmega.

Oh, I see you have the noise only when the ATmega is off. You could consider a 1M resistor or such to replace C1, to act as a pull-down when the controller is off.

 

are there any benefits of using a cap vs a resistor?

Im only using 5v into the first transistor because when im writing the code for the microcontroller, everything is being run off 5v (from usb, not using the high current LED's yet)

 

A resistor from the base of the TIP31 to ground would aid in turning off the LED drive transistor. A 10K would do the trick, I think.

hgmjr

 

Using a resistor with the LED is also required.

http://forum.allaboutcircuits.com/blog.php?b=378

 

are there any benefits of using a cap vs a resistor?

Using a pulldown resistor to turn off the output when a signal is absent is a standard approach, something you should always have in your mind and look for ways to apply. Your cap may be working in this case, and that's fine, but it's specific to pulling down very high frequency noise. Any low frequency or DC float would still be there. A resistor will pull everything down. Be sure to note Bill's comment above. If not, bye bye LED

 

I replaced C1 with a 1M resistor but now the LED lights up when it should be off. Put C1 back in, LED is now off

 

Are you sure that the ATMEGA output being used to drive the transistor is configured as an output?

hgmjr

 

    tccr1a |= 1<<wgm11 | 1<<com1a1 | 1<<com1b1;
    tccr1b |= 1<<wgm12 | 1<<wgm13 | 1<<cs10;
    tccr2 = (0<<foc2)|(1<<wgm21)|(1<<wgm20)|(1<<com21)|(0<<com20)|(0<<cs20)|(1<<cs21)|(0<<cs22); 

    icr1 = 255;

 

I replaced C1 with a 1M resistor but now the LED lights up when it should be off. Put C1 back in, LED is now off

Something's wrong. Are the two grounds really the same? And I assume the LED goes off if the ATmega is disconnected (with the 1M resistor still in place)?

 

It would help to have a schematic posted.

hgmjr

 

Something's wrong. Are the two grounds really the same? And I assume the LED goes off if the ATmega is disconnected (with the 1M resistor still in place)?

Ground is the same one all off the USB programmer. Also still the same result if I use a different power source.

If i remove the connection from the atmega pin the LED turns off.

When the pin should be off, its reading 0.22v at the atmega pin. At the BC547 base pin its reading 0.09v with the 1M ohm resistor in.

 

Hmmm, maybe you need to pull down the ATmega output directly, not after the 100k resistor. Try moving the 1M to the other side of the 100K.

 

No difference :/

 

You can attach the full schematic and the source code here.

hgmjr