Zen Lite and Heatsinks

[skidave]

I just finished building a Zen-Lite based off of a Zen Version 4 schematic and I have a question on heat sink size and heat of the actual mosfets? I used heatsinks that are .25 inch aluminum 12 inches long, 2.5 inches high and they have 5 big fins per heatsink. I used IRFP044Ns for the regulator and the output devices.

How hot can the mosfet get before failures? I am running it with 36 volts (regulated voltage) and 500 watt bulb combination. The mosfets are too hot to touch for more than two seconds (after about 10 minutes), and after a half hour of use, the heat sink is almost too hot to touch also. I have heat sink compound between the isolation pad and the mosfet and also between the pad and the heat sink. I have read all of the specs from the device data sheet, but what is everyone's experience with this type of design?

If I turn the gate voltage down, the drive signal will be less to the mosfet (and the bulbs are not as bright) so this would mean less current through the mosfet, correct? I am going to borrow an infrared thermometer and test the amp tonight with 500 watt bulb combo on one channel and then 400 watt bulb combo on the other with the gate drive lowered.

Any suggestions / corrections are welcome. The unit has no oscillations and the freq respons is great and sounds great with no noise or hum. I'm just concerned about the heat.

Thanks,

Dave

[GRollins]

I'm more conservative than Nelson, true, but I like being able to keep my hand on the heatsink indefinitely. He'll tell you something like 5 to 10 seconds. Either way, if you can't touch it, it's too hot.

Grey

[flg]

Well, how about this... Do you have a source resister on your gain FET? Measure the voltage across it... V/R= I. Also measure the voltage across the drain to source of the FET... V*I= Watts. Some people will give you different answers but I think over 40 Watts is getting close to to much

[Nelson Pass]

Originally posted by skidave
after a half hour of use, the heat sink is almost too hot to touch also

If I turn the gate voltage down, the drive signal will be less to the mosfet (and the bulbs are not as bright) so this would mean less current through the mosfet, correct

"almost too hot" approximately right.

As you adjust the Gate voltage, you will go through a region
of maximum dissipation on the Mosfet. As the Gate voltage
goes up from there and the current increases, the Mosfet sees
less dissipation as the lamp limits the voltage.

As the Gate voltage goes down from there, the voltage across
the Mosfet increases but the current decreases.

You can adjust this bias setting to any value that pleases you,
and this is likely to be in the "middle ground".

[skidave]

Thank you Nelson and Flq. I do not have source resistors, so I put my meter in between the lamp the the mosfet. With a gate drive of 3.7Volts, I am dropping 24 volts across the lamps, there is 10 volts across the mosfet and the current is 1.01 amps. If I drop the gate voltage, the current is less and the voltage across the lamps drop. At 1 amp, the lamps are well lit and the resistance measurement is very close to the scale in the Zen Lite article. Unfortunately, with my 10K pot, I run out of adjustment to drive the gate voltage any higher. Nelson, with your reply, should I be able to be at the balance point of the current Vs the voltage, or would I need more gate drive?

Nelson, Thanks for all of your efforts!

Dave

[Vix]

Dave,

In your case, I would adjust the voltage on the Drain to be 1/3 rd of the supply voltage. If you have 36 volt regulated, then just adjust so you have around 24v across the bulb and 12v or so across a mosfet.

If the heatsinks are still too hot, decrease the current by using a lower wattage bulb. The downside of this is the increased distortion. So, it looks that there are two practical solutions:

1. Use a bigger heatsink, or
2. Use a fan

I use computer 12v PSU fans run at half the voltage, they are quiet and still manage to cool enough.

Regards,

Vix