Simple Symetrical Amplifier

[PauloPT]

Hi shaan.

I got your PM. But I'm sorry to say that right now I am typing from my cellphone and am far away from home, so can't help you with printing problems. Please don't mind.

No worries. Thank you for your kind help.

[brianco]

Originally Posted by shaan

Hi paulo.

.

.................................................. ..... You can safely run the amp from 35V rails into 4ohm with a single pair of FETs.

Happy listening.

shaan

Hi Shan,

I too know little about solid state design and especially heat-sink requirements. Run as shown in red in the quote above, what is a safe minimal heat-sink requirement per channel?

I can wait until you return for a response!

Many thanks.

[shaan]

Originally Posted by brianco

Hi Shan,

I too know little about solid state design and especially heat-sink requirements. Run as shown in red in the quote above, what is a safe minimal heat-sink requirement per channel?

I can wait until you return for a response!

Many thanks.

Hi brianco.

My rough guess is a 0.5degC/W heatsink per channel. You can see I am no expert in heatsink designing. However, I use a p4 CPU heatsink with the 12v fan running at 6v, per channel. And this works well at full power output. If you try this I doubt you'll be disappointed.

shaan

[brianco]

Thank you Shan...I am sure your temperatures 'near the Ganges' are a lot higher than ours here in Scotland!

That gives me a good clue that the heatsinks I already have will do - at least for testing.

[shaan]

My pleasure Brian. But, do you really need heatsinks in Scotland? Well London may need sinks though, in mid summer.

[Esperado]

Originally Posted by shaan

You can see I am no expert in heatsink designing.

It is very simple. If V is the power supply voltage ( "V+" + "V-" ), and Rl the impedance load, the max dissipated power will be: Wdiss = V²/(19.75 x Rl). for a class B amp.
You add the thermal resistance of the device, isolator if any, and radiator and you get the total thermal resistance. Now, it is easy to calculate the max temperature.

[Esperado]

If you allow the temperature junction to reach 150° when atmospheric air is at 50°, you need a temperature delta of 100°.
You have to know the power of your transistor at 25°. Now, you divide 25 by this power to know the internal thermal resistance of your transistor. lets call this "K"
You add the thermal resistance between the case of your power device and the radiator for a mounting with thermal grease. Lets call this 'D"
Multiply this D value by two if you add a isolation material, and add the thermal resistance of the isolator itself.
Now, you have the total thermal resistance of the mounted transistor by adding "K"+"D" = "R".

Divide this value by the number of power device you will use: "S".
P is the power value you need, according to the last post.
We need "T"= 100/P of total thermal resistance.
T-S will give-you the maximum desired thermal resistance of your radiator. Any value less than that will feat.
Yes, i know, it is boring to calculate.

[PauloPT]

Hi shaan,

I've been doing simulations with LTSpice. I get a lot of harmonics. See the pic below. Is this a concern or just a simulation glitch?

Thanks

[shaan]

Hi paulo. We need to see the schematic you are simulating. At what output voltage and load resistance was this FFT taken?

[Esperado]

Originally Posted by PauloPT

I've been doing simulations with LTSpice. I get a lot of harmonics.

H2 is at -75dB from the signal level, (<0.02%) what the hell ?