Check this

[hantto]

Hello, I decided to clean up this shematic because it didn't look very nice originally. I have double checked that i copied it right but as i'm only human it's good that someone else also checks it. So if you have the time please have a look at these.

Greets: Hantto

 

[Russlk]

This is a class A/B amplifier, it probably works good, but it is tossing 180 watts down the drain all the time. Adding another diode in series would improve the efficiency and most likely not cause enuf distortion to notice.

 

[Nigel Goodwin]

This is a class A/B amplifier, it probably works good, but it is tossing 180 watts down the drain all the time. Adding another diode in series would improve the efficiency and most likely not cause enuf distortion to notice.

Personally I would replace the two diodes with a VBE multiplier, then you could adjust the current as you wish. Schemes using fixed components are often too inaccurate - if only slightly too high it can easily cause thermal runaway.

 

[hantto]

Thanks for the replies.

So, what's a VBE multipier?

 

[Nigel Goodwin]

Thanks for the replies.

So, what's a VBE multipier?

It's a transistor with a resistor from base to collector, and another from base to emitter - giving effectively a two terminal device. If the two resistors are the same value it works like a zener diode with a voltage of about 1.4V - two times the Vbe switch on voltage. By making one of these two resistors variable you get a variable voltage zener diode. By replacing the two diodes with this circuit it allows you to adjust the quiescent current - it's commonly used in audio amplifiers for this purpose.

BTW, the original circuit won't be dissipating loads of heat as suggested, two diodes are not enough to turn the output transistors on - they are darlingtons and require roughly four diodes for that - adding a third one as suggested could only increase the current, not reduce it (although three may still be too low).

 

[hantto]

Hi again.
Because i'm no eletronics vet, I just built the original circuit with 2 diodes. And it worked perfeclty for about 30secs. Then it a strange popp sound "pop piuuu-plim, tr pöööööö" and then the element was all the way in and hummed. It sounded like that it had gone on overdrive or something. Then I powered it off and waited for a while and then I put on again. Same thing happened. I tried to put more diodes as you suggested, but the same happened again. Then I accidently tripped over the cables that went to the element, and they got lose and then it blue the fuses. I guess I destroyed the output transistors, didn't I?

SO, do you know what went wrong?

Thanks for your time!

 

[Nigel Goodwin]

Hi again.
Because i'm no eletronics vet, I just built the original circuit with 2 diodes. And it worked perfeclty for about 30secs. Then it a strange popp sound "pop piuuu-plim, tr pöööööö" and then the element was all the way in and hummed. It sounded like that it had gone on overdrive or something. Then I powered it off and waited for a while and then I put on again. Same thing happened. I tried to put more diodes as you suggested, but the same happened again. Then I accidently tripped over the cables that went to the element, and they got lose and then it blue the fuses. I guess I destroyed the output transistors, didn't I?

SO, do you know what went wrong?

Thanks for your time!

I would say that it most probably suffered from thermal runaway, taking more and more current until the output transistors blew. A nice feature of a vbe multiplier is that you can initially set the quiescent current to zero (with the transistor turned fully on) and make sure it all works OK - it will have cross-over distortion at low levels, but you can test it and even run it at high power. Then all you need to do is turn the pot to give it a little quiescent current, you can use a scope and low level sinewave to adjust until the cross-over distortion disappears.

With a fixed circuit, like the one you showed, the quiescent current is set at a value determined by the particular components you have used - you should also clamp the two diodes to the heatsink the output transistors are mounted on, this will reduce the quiescent current as the transistors get hotter.