engineering question

so what is happening is the fans are running in parallel to the 555 and off a rectified 12v step down transformer. all this seems to work fine and the 555 oscillates somewhere around a kHz. i dont know the exact number.

the LEDs are running off a rectified AC mains. this is the same plug that is powering the fans.

the current is limited to the LEDs by means of MOSFET pulsing.

the entire circuit worked when i used a variable DC power supply but when i replaced the power supply with s transformer rated for 12v, it seems to have thrown off my available power to the LEDs. i cannot get them to draw more than .1a

i may have to use two separate outlets to deal with this. i did try putiing the transformer in series with the 120v rectifier. with the transformer hot going to ac main hot, the transformer neutral going to the recifier and the ac neutral going to the other part of the 120v rectifer.
this caused the circuit to all together not work.

I'll measure the current coming out of the transformer to find the inductance and see if that's the problem. the reactance could be too low and in turn, making the 120v 4-6a side (which is 30-20 ohms after the mosfet) match the ohms of the transformer cuz theyre in parallel....

anyone agree?, disagree?

every light operates correctly when individually tested.... just checked.

gonna try an AC capacitor across the transformer... or maybe in series.... h/o... math time

is there a recommended pulse frequency? ir is duty cycle all i really need to worry about? i pumped a frequency generator through thia circuit and it only suggested the higher the frequency then the lower the instensity. the duty cycle percentage controlled the current really well. 90-50% duty was the most useable range. and this controlls the intensity as well, just better.

my mosfet is getting a bit hot. only good for 200v 25a and im running it @ 120vdc 5.4amps with a bad ripple and a peak voltage of 170 vdc so i ordered a 400v 25a to swap it out. i might have to do that whole seperate outlet scenario...

120v X 5.4a= 648watts as of today. pretty cool! should handle 1000w maxed

6 red (3 diff nm), 3 blue(2 diff nm), and one warm white cuz i burned out a blue by over powering the light a bit ago

oops

Wouldn't you need an inductor and a diode in the output for it to be a working buck converter? (as well as another capacitor at the output for more smoothing)



Also, that would make a voltage source without feedback. How do you intend to regulate current with your design?

iHearAll said:

is there a recommended pulse frequency?

Click to expand...

The higher the frequency, the smaller L can be. This is the main reason switch mode power supplies are so much smaller than transformers at 60hz.

im regulating current with the MOSFET by switching it on and off. I originally designed it for pwm but when i tested it with the square wave from a lab frequency generator and noticed it limited current as the dc% went down. The MOSFET puts off a fair amount of heat like one would imagine from a power resistor.

i think the MOSFET isn't taking up the current but rather controlling the foward bias cut on/off in the LEDs characteristic curve. so the resistance of the LED increases and decreases.

Does the inductor and cap need to be calculated per pulse frequency? like (.5)X(frequency) to find the resonanant freq.? i gotta wrap my head around that schematic.

ok yea you're right. that makes sense. I'll put a pwm in instead. tl494cn has dual outputs Q and Qnot if i recall.... or do you know any better IC for the job? flip flop to get the delay aspect?

yea, caps were unnecessary in this instance. all that i needed was the inductor right in front of the LEDs as well as the diode across the circuit. caps parallel to the LEDs restored too much current.

seeing 129v 5.5a and no flickering of the LEDs. adding inductance to the LED side allowed storage of energy in the same vector direction as the fans and 555.

cool cool