Periodic lamp fader

[ronv]

Yes, the lights, as well as the PCB, are powered by the rectified +12 V from the charger, but it only amounts to about half an amp, for what that's worth Do I understand the suggestion correctly that I should power the transistors' collectors directly from the charger, and everything else on the PCB via a diode?

Yes, I think we are on the same page: 12 volts directly from the charger to the lights then a diode from the charger to the circuit with the cap on the circuit side. (anode to the charger)

 

[jelliott]

I still intend to implement Ronv's diode suggestion to improve the fading behavior, but the circuit has recently stopped working and I'm assuming at this point that the transistors have failed (they've always gotten waaaay hotter than I think they should with the currents they're carrying). So what I'm wondering now is if this apparent overheating of the transistors could be related to the aforementioned ripple on the +12V supply. Would there be any value in putting the transistors (in addition to the rest of the circuit) downstream of the diode to shield them from the ripple as well?

 

[jelliott]

I implemented Ronv's diode suggestion (and replaced the transistors to address the failure noted in post #22), and the flickering does not seem to be improved much, if at all. Additionally, there seems to be a consistent and very pronounced flicker at the beginning and end of the cycle. Is there anything else I can do to clean up this dirty +12V supply?

 

[alec_t]

It could be the flicker is due to the 1000uF cap discharging and recharging because of the lamp load. Where do you have the 1000uF cap connected? It needs to be downstream of the diode; and the connection of the lamps to the +12V line needs to be upstream of the diode. That way the diode provides a degree of isolation of the dimmer circuit supply from the load supply and the dimmer circuit should get a reasonably steady 12V with only slight ripple.

 

[jelliott]

Yes, the capacitor is downstream of the diode and the connection to the transistor collectors is upstream of the diode.

I've managed to largely eliminate the flickering by attaching a 5 W incandescent bulb to the +12V signal (directly to the power supply, not via the fader circuit).

 

[ronv]

I simulated the circuit and it has the putt putts. You can make it better by using a comparitor in place of the 324, but it's still not real good. We can try to come up with a good one if you are willing to start over.

 

[jelliott]

I simulated the circuit and it has the putt putts. You can make it better by using a comparitor in place of the 324, but it's still not real good. We can try to come up with a good one if you are willing to start over.

I'm pretty satisfied with the behavior of the circuit since I added the 5 W load directly to the power supply, but I'm not opposed to starting over at some point in the near future; I'm still bothered by how hot the transistors get (I've gone from two to four of them in parallel since the original pair failed, but they still get very hot despite the total load being less than the rating on a single 2N3053).

I'm also curious what kind of improvement I might see with a comparator in instead of the LM324. What else would need to be changed on the PCB to go that route?

Thanks again.