[sdiy] Panel and Heat Sink

[James Patchell]

No, the more voltage, genreally more heat....here is a way of thinking 
about it...for a 50 Watt, class A amplifer (well, any amplifier actually) 
into a nominal 8 ohm load, you need to put 28.3 volts across the load 
(speaker).  The peak current will be 3.53 amps.  Now, a class A amplifer, 
by definition will be biased with this same current, this is so no output 
driver will ever completely cut off.  So, if we are talking about 
transistors, and lets assume for the sake of argument a push pull 
configuration (same applies to single ended), the voltage between rails 
will be a minimum of 56.6 volts, but we need to add a few volts to that 
because you can't get zero volts across a transistor, plus there will be 
things like emitter (or source) resistors, so that gets us up to 60 volts, 
now if you multiply that by the bias current, guess what, we now have 200+ 
watts being burned up (radiated away) as heat.  That is a lot of heat to 
have to dispose of...and remember, we are only talking about a 50 watt 
amplifier here.  A 100 watt class A amplifier needs to dispose of 400 
watts, a 200 watt amplifier needs to dispose of 800 watts.  Getting rid of 
800 watts is not an easy task, which is one of the reasons why you don't 
see a lot of high power Class A amplifiers.  By the way, pretty much the 
same number apply to tubes as well.  It is just the nature of the 
beast.  Also why Class B amplifiers are so popular.  With proper design, a 
Class B amplifier can be made with very low distortion without having to 
resort to large amounts of feedback.

         MosFets, while interesting to play with, pretty much turned out to 
be a dead end, in my opinion.  They are not thermally stable and 
predictable, they are non-linear, and just plain cranky... :-) .  If you 
want a challenge, they are the way to go.  Just be prepared to burn a few 
out along the way, and don't hook up any expensive speakers to the amp 
until you are quite sure it is going to remain in a "working" state....take 
this advice from someone who has built quite a few FET amps...(who by the 
way, has switched back to BJTs).

         Oh, yes, on cool thing about MosFets....I built a 75 Watt 
amplifier that had a -3dB bandwidth of about 2.5MHz....I could run a color 
TV signal through the amp and view it on a video screen...you can't do that 
with BJTs....

At 09:29 PM 1/23/2005 -0500, anthony wrote:

>----- Original Message ----- From: "James Patchell" <patchell at cox.net>
>
>
>>Well, here is the panel and heat sink I managed to finish, more or less, 
>>today.  I still need to drill the holes and tap them to fasten the 
>>heatsink to the panel, but, all of the hard work is done....except I have 
>>one more panel I need to do.  I could have drilled those last holes but I 
>>ran out of daylight again...(dang)...I need to install those new light 
>>fixtures out in my garage...then I can DIY 24-7 :-)
>
>Good amp design. I've seen so many mosfet amp projects in electronics 
>magazines in the 80's and 90's. All of them were too complicated I 
>thought. I've seen some single-ended designs recently that look like they 
>might be fun. Very few parts. Almost like making a tube amp (guitar). I 
>have some IRF510's (4). Maybe I could make a tiny amp that ran hot. Would 
>a Class A amp run cooler the more voltage you put through it?
>
>

         -Jim
***************************************************************
http://www.oldcrows.net/~patchell

***************************************************************