class A power amps and how they work?

[the cracken]

i want to amplify 500mv to power of 500W single channel. i want to use solid state but im not sure how to go about it. i get how a simple single class a amp would work but i dont get how to make multiple stages to increase power. do you increase the voltage and then have a single current ampifier?? thanks for any help

 

[ronv]

Depends on what you are trying to do. Is it audio, rf, or ???

 

[the cracken]

Depends on what you are trying to do. Is it audio, rf, or ???

im thinking of making a audio amp!
and what voltage do they usually run on!!

 

[Diver300]

Audio amplifiers are far cheaper to buy than to make.

The voltage that power amplifiers run on depends on the power needed and the speaker resistance. Speaker resistance is usually 4 or 8 ohms. A 4 ohm speaker needs about 45 V RMS to get 500 W, which is about 63 V peak, so power supplies would be +/- 70 V for that.

 

[ben7]

Class A amplifiers use the most power, but reward you with high quality music!

 

[audioguru]

Class-AB amplifiers are the most popular and have excellent specs. They are cool when they do nothing. Class-A amplifiers waste a lot of power making enough heat to warm your entire neighbourhood, even when they do nothing.

 

[the cracken]

most things are cheaper to buy but that's no fun
im thinking of making a guitar amp. so what voltage do you think 8 ohm speakers need?? 140v?
so i would have to get the input up to 130 p to p and then stick a single? current stage in there??

 

[the cracken]

i don't think i have heard a class a amp. but i think i would want it to be even though cross over distortion is pretty low in class AB!!

 

[audioguru]

i don't think i have heard a class-A amp. but i think i would want it to be even though cross over distortion is pretty low in class AB!!

Crossover distortion in class-AB is so low that it cannot be heard and cannot even be measured.
Most guitar amps have as much distortion as is possible. For lots of FUZZZZZZ.

The voltage for a speaker depends on a simple arithmatic calculation involving how much power is needed. Power output equals the RMS voltage squared divided by the speaker's impedance. So 500W into 8 ohms is 63.3V RMS. It is 179V p-p. It needs a plus and minus 93V power supply.

The peak-to-peak voltage is the RMS voltage times 2.828. An amplifier is not perfect so it needs a power supply voltage that is about 6V more than the p-p voltage.

Most speakers blow up when fed 500W.

 

[the cracken]

iv got a small 18inch celestion speaker i like to keep everything clean as possible and let the distortion pedal do the distorting. your saying plus minus 93v but thats for a AB class amp? the other thing that went through my head was it would be slightly easier to build a class a amp is that right?
thanks for the info

 

[Nigel Goodwin]

iv got a small 18inch celestion speaker i like to keep everything clean as possible and let the distortion pedal do the distorting. your saying plus minus 93v but thats for a AB class amp? the other thing that went through my head was it would be slightly easier to build a class a amp is that right?
thanks for the info

No, a class A amp (particularly of that power) would be difficult, VERY expensive, and VERY, VERY heavy - you don't need class A, it would give no advantage (and doesn't for any audio power amp really).

As others have suggested - just buy one, it's cheaper, better - and likely to be a LOT more reliable.

You also don't need 500W for a guitar amp, 100W is plenty - for filling a large room you have PA.

 

[the cracken]

so you've convinced me how would i go about making a ab class amp?? im guessing an opamp for the first stage!

 

[Nigel Goodwin]

so you've convinced me how would i go about making a ab class amp?? im guessing an opamp for the first stage!

Even a class AB is going to cost a LOT more to make than to buy - it's fairly unusual to use an opamp, but the entire power amp is basically a big opamp.

Here's a fairly typical 100W design for a guitar amp:

**broken link removed**

For a 'kit' you might look at:

http://www.class-d.com/Mosfets/Modules/NX400-Amplifier-Heatsink-Fan-HiFi-PA-DJ-400w/prod_21010.html

Which is a 450W power amplifier, the PSU is extra to that price.

 

[audioguru]

A class-AB amplifier has an excellent and commonly used circuit. With a full output power of 500W RMS it will idle at only about 50W (of heat).
It will produce about 300W of heat at full output.

A class-A amplifier is rare. Its circuit has about the same complexity as a class-AB amp but it produces a lot of heat all the time so it needs many output transistors and huge heatsinks, maybe cooled with fans. The power supply must be extremely powerful and very well regulated. With a full output power of 500W RMS it will continuously heat at about 1250W (even when it is idling!). Here is a link to an article about extremely inefficient class-A amplifiers: **broken link removed**

Most modern high power amplifiers use class-D (a switching circuit) which is very efficient and produces only a small amount of heat.

 

[the cracken]

class d bleh i have a bad taste in m mouth! do you know of a good power mosfet to use??

 

[the cracken]

any simple push pull class A/B low power amps to get me going?

 

[audioguru]

any simple push pull class A/B low power amps to get me going?

"Simple" for a high power amplifier means very poor quality. An LM386 litttle power amplifier IC is very simple but its output is only 0.45W at clipping into 8 ohms with a 9V supply.

Here is a typical high power class-AB amplifier project (1500W into 4 ohms or 800W into 8 ohms) that nobody ever made:
**broken link removed**

 

[the cracken]

i though you couldn't parallel transistors like that??? is it d9 and d10 that drops a voltage on the base to make it class a/b? so i just want to make it clear i need to up the voltage from the input to something (depends on power output) higher. and i do that with a transistor that varies the voltage (e.g output on the collector) and then stick that into a load of transistors that vary the current (eg speaker in series with the transistor) is that right???
thanks

 

[Nigel Goodwin]

i though you couldn't parallel transistors like that??? is it d9 and d10 that drops a voltage on the base to make it class a/b?

No, Q5 and VR2 etc. set the bias conditions.

Those two diodes are part of the overload protection.

There's no problems paralleling transistors as long as it's done correctly.

 

[Diver300]

You can parallel transistors, as long as you add an emmitter resistor on each one like the circuit shows. The resistors basically make sure that all transistors share the current.

The resistor value has to be large enough to stop themal runaway of any one transistor. If one transistor is hotter, its base-emmitter voltage will be less, so there will be more voltage across its emmitter resistor, so there will be more current and more power in that transistor. That is a recipie for a thermal runaway.

How well the transistors are coupled to the heat sink and the resistor value has to be worked out so that it doesn't go unstable.

D9 and D10 are there to stop spikes from the speakers doing any damage. They aren't connected to the base of any of the transistors.

It is R9, VR2, R10 and Q5 that set the bias current above zero so that it is class A/B.

I've just seen that Nigel beat me to the answer....