blowing transistors

[rtill]

I built
figure 1a
http://sound.westhost.com/project03.htm

it worked great for about 3 hours with perfect sound, low distortion, great response and then I heard a fizzle sound and the bd139 and bd140 transistors blew. I replaced them and they blow again as soon as I turn the amp on.

What could cause this? and why did it work for so long?

[Tube_Dude]

Originally posted by rtill


it worked great for about 3 hours with perfect sound, low distortion, great response and then I heard a fizzle sound and the bd139 and bd140 transistors blew. I replaced them and they blow again as soon as I turn the amp on.

Hi

Inspect the 220 Ohms resistors and if the output transistors 3055/2955 are not burned.

If they are burned, when you put new drivers BD139/140, they immediately burn...

PS: Inspect also the 0,22 Ohms resistors

Edit: PS...

[Andy L. Francis]

Possibly thermal runaway

[rtill]

looking at it closer, after replacing the bd139 and bd140 and firing it up, the .47 5 watt resistors cracked as well. So does this sound like the output transistors (mje3055 and mje2955) are blown?

[Tube_Dude]

Originally posted by rtill
looking at it closer, after replacing the bd139 and bd140 and firing it up, the .47 5 watt resistors cracked as well. So does this sound like the output transistors (mje3055 and mje2955) are blown?

Yes, strong possibility, inspect them...

[jaycee]

Mount the 1N4001 diodes on the heatsink. Better still replace them with a VBE multiplier circuit as shown in figure 1a - use a BD139 or MJE340 for the VBE transistor and make sure it's tab is bolted to the heatsink

[sakis]

is almost impossible to fail ..... whats the rail voltage ???

and also have you got original pcb or home made ???? is the caps values correct or you made something that creates some ocilation together with a bad pcb ????

this is made it hink for rails 30+30 volts max ...higher rail voltage will not possible for this amp

[rtill]

rail voltages are 34.7 + and -
I made it on perf board and used a heatsink from a computer processor.

all 100 pf caps that it calls for, I used 150 pf, otherwise the circuit is exact. It worked great for 3 straight hours. So I would be confused of why it could only handle 30 Volts per rail.

[sakis]

34.5 is not much but cooler of a procesor from a pc might not be enough ....

meaning that if you have just a bit toomuch idle .....then all thing gets too warm and then boom .... put a decent heatsink allow proper insulation be tween transistors and then all will be fine ....

you might consider going to project 3 a that has ajustable idle

will work like hell and also plays very fine

[sakis]

on vas amplifier bd139 there is capacitor that should remain 100 pf ..... 150pf will have alot of effect in the quality of sound .....

about the second driver bd140 this change of 50pf more i dont know what really effects

also in a version i made 100mf decoupling caps are located very next to the power transistors conected to rails and ground ..... very nice improovement also