what is the heatsink size needed for some lm4780 ICs. my planned useage is for a 100 watt amplifer on one IC, and 2 50 watt amps on the other two. the thing is that the woofer amp will be powered off of a seperate power supply. ther will be an acitve crossover with filters at 60hz, and 2khz (or values near there).
the woofer IC will probably be strained during heavy useage though. i like to listen to music with impactful bassline whenever i listen to loud music.
the mid and tweeter ICs would probably not be driven as hard. i could power them off a smaller supply if needed.
heatsink -- http://www.apexjr.com/new.htm
Heat Sink size 6" X 8" X 1/2" great for use with the Alesis Modules $3.95
i have a total of 3, but would kinda like to have a neat looking project and don't really want to use 3, and there so big that 2 is still pretty obstrusive. (and i lost one of the extra ones and i have an amp another already...)
in anycase, how much cooling should i be looking for? i plan to use fan cooling as it is.
what about water cooling using copper tubing pressed against the heatsink?
the woofer IC will probably be strained during heavy useage though. i like to listen to music with impactful bassline whenever i listen to loud music.
the mid and tweeter ICs would probably not be driven as hard. i could power them off a smaller supply if needed.
heatsink -- http://www.apexjr.com/new.htm
Heat Sink size 6" X 8" X 1/2" great for use with the Alesis Modules $3.95
i have a total of 3, but would kinda like to have a neat looking project and don't really want to use 3, and there so big that 2 is still pretty obstrusive. (and i lost one of the extra ones and i have an amp another already...)
in anycase, how much cooling should i be looking for? i plan to use fan cooling as it is.
what about water cooling using copper tubing pressed against the heatsink?
theChris said:
Whats happenin dude? These goin in your car?
What is the thermal dissipation of those IC's? And ApexJr doesn't list the thermal resistance of their sinks... They do look like they'd work for that amount of power though. But fins do look kinda tiny. I'd suggest setting up the circuit and thermally testing the ICs on the sink at full power into a power resistor BEFORE making a chassis to make sure the devices don't overheat on that sink. For $4, you haven't got much to lose.
Water cooling would be pretty hype! If you were going to use tubes pressed against the sink, I'd run them back and forth like a rad with LOTS of thermal compound. I don't know if you can get the tubing with a square x-section but that would be preferable.
on a similar note, i assume it is a good idea to have airflow over the bridge rectifiers and transformer?
The transformer should be ok without much airflow. Usually they're >90% efficient with small losses attributed to hysteresis and eddy currents so it may get warm, but I seriously doubt it would smoke before something else does. The rectifiers though.... being semiconductors I'd at least put them on a small sink and airflow couldn't hurt. You're best bet would be to do some thermal calculations on all heat dissipating semis. I've got Sloane's High Power Amplifier Construction Manual and he gives a good breakdown of the formula and variables used but I don't have it here at school. I'll post it when I get home.
i was assume maby 2V of drop to be safe, and 7A of current for a value of 14 watt. i'll probably put a small sink on em and get forced airflow over them.
Ok heres the breakdown:
Rthja = Rthjc + Rthcs + Rthins + Rthsa
th - thermal
ja - junction to air
jc - junction to case
cs - case to sink
ins - insulator
sa - sink to air
All these variables are thermal resistances. Sloane mentions typical values of Rthcs = 0.1*C/W, and Rthins = 0.4*C/W (for a TO-3 type silpad). He also notes that you should multiply the number of similar devices that are mounted on the same sink by Rthsa for calculating the final Rthja
To calculate the individual junction temperature, multiply Rthja by the individual device power dissipation and add the ambient room temp (~25*C).
To find the max temp of the heatsink, multiply Rthsa by the total power dissipation on that sink and add 25*C.
Rthsa should be given by the heatsink manufacturer and will be different depending whether or not you use a fan for convection. Good manufacturers should provide numerous values associated with different amounts airflow.
Rthjc should be given by the device manufacturer.
Whether you want to go through this is up to you. 14 watts of dissipation isn't much to sweat over. But when you get into making class A's or higher powered class B's, I wouldn't be guessing sink size.
Make sure you post some pictures of your project! I'd love to see what a fellow CAF'er is doin to push the hobby to the extreme!
Cheers
Dan
Rthja = Rthjc + Rthcs + Rthins + Rthsa
th - thermal
ja - junction to air
jc - junction to case
cs - case to sink
ins - insulator
sa - sink to air
All these variables are thermal resistances. Sloane mentions typical values of Rthcs = 0.1*C/W, and Rthins = 0.4*C/W (for a TO-3 type silpad). He also notes that you should multiply the number of similar devices that are mounted on the same sink by Rthsa for calculating the final Rthja
To calculate the individual junction temperature, multiply Rthja by the individual device power dissipation and add the ambient room temp (~25*C).
To find the max temp of the heatsink, multiply Rthsa by the total power dissipation on that sink and add 25*C.
Rthsa should be given by the heatsink manufacturer and will be different depending whether or not you use a fan for convection. Good manufacturers should provide numerous values associated with different amounts airflow.
Rthjc should be given by the device manufacturer.
Whether you want to go through this is up to you. 14 watts of dissipation isn't much to sweat over. But when you get into making class A's or higher powered class B's, I wouldn't be guessing sink size.
Make sure you post some pictures of your project! I'd love to see what a fellow CAF'er is doin to push the hobby to the extreme!
Cheers
Dan
- Status
- Not open for further replies.