PA power & heat sink

Hi, I just put together amplifier using a pair of TDA2009 configured in bridge mode which is claimed to deliver 18W each channel. I found the heat has a huge role in this. The amp would turn really hot and output signal becomes ugly after puting out some 16W peak for a few minutes. I used 1Khz sine wave source. An amp channel was loaded with a 4 ohm resistor.

My question is is it true that it would be a waste if I have to make the heat sink that can help the amp to put out 9W RMS (18W peak sine wave) over a long period since in real life, songs's RMS power should be much less than that of sine waves. If it is true what is the rms power should I target? Thanks.

Reply to
Seeker
Loading thread data ...

Data sheet indicates that 4 ohm is wrong R for this amp in bridge mode. What supply V are you using? Do you know what the rating of your heatsink is? (C/W)

The chip's rated to 12.5W pc = 25w P_out and you think a heatsink goof for 9w rms is too big?

Have you got the anti-oscillation bits on it? Show us the cct diag youre using would help

NT

Reply to
meow2222

Check max Z load. Average music is typically one tenth peak. Rough guide! Max heat load power must be checked. This is usually something less than full power.

greg

Reply to
GregS

Those parts *MUST* have a heat sink.

Where did you get the idea the use of a heat sink was optional ?

Without a heat sink, the chip will get incredibly hot internally and even if it doesn't die on you straight away it may very well be damaged, leading to probable early field failure.

A common technique at these low powers is to bolt the heatsink tab to the metal enclosure (using an electically insulating washer if required ) and also use thermal transfer compound to keep the thermal resistance low.

To establish the size of heatsink required I like to test with real music (use rock and roll or 'DJ music', not symphonic for a more demanding test) and the lowest impedance load that's likely to be connected *and play it quite loud* !

Graham

Reply to
Eeyore

Define average ! And what's your source for that info ?

The heatsink has to work for the *worst* case btw.

Graham

Reply to
Eeyore

I guess from memory. Trouble is, todays music has much less variation. Average power = average voltage squared, devided by average Z. Z is frequency dependant so its a bit of a guess.

I used to watch the power meters on my Tigersaurus amps, which was just a voltmeter, and much of the music averaged about 1/4 or less in average voltage with the scale in power.. That might indicate a 6 db variation, but I said or less. If the amp was driving a bass speaker with active crossover, it mught be more like 50% average power or more with todays music.

greg

Reply to
GregS

}Hi, }I just put together amplifier using a pair of TDA2009 configured in }bridge mode which is claimed to deliver 18W each channel. }I found the heat has a huge role in this. The amp would turn really }hot and output signal becomes ugly after puting out some 16W peak for }a few minutes. I used 1Khz sine wave source. An amp channel was loaded }with a 4 ohm resistor.

Do what the other posters suggest, then put a scope across the output and look for a signal above 20KHZ (like a feedback signal way above the hearing range). This would cause excessive heat dissipation (and cause dogs in the vicinty to act up).

Stan.

Reply to
Stan

I might be repeating someone here, but if the amps are in bridge mode, I think 4 ohms is too low, I think 8 ohms is the minumum. and you need lots and lots of cooling. You say 'an' amp channel was loaded with 4 ohms, but surely in bridge mode, there IS only one output channel, or am I missing something?

IMO

Ron(UK)

Reply to
Ron(UK)

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.