Idling current in Rega Mira 1, 2 or 3

[needsp]

Hi 2N6658

Failure of old output devices, as here, is a major cause of amp failure. The amp is 17 years old, new output transistors all round is, IMO, good practice. You clearly disagree

My soldering is good, and the devices are inserted correctly. But we all think that- I'll check again!

What do you mean by "old 2SC3519 and 2SA1386"? Not the originals I assume, as the 2SA1386 was shorted through? I bought the replacements from Mouser or Farnell, so they are genuine. Some of the other devices are obsolete- I bought these from smaller but reputable sources, but its more likely- but still unlikely- that these might be a fault I suppose.

I'll post test voltages for both channels soon. As I said, it's trickier than it might be, as the two channels are laid out very differently, and the marked component names on the pcb don't correspond to the circuit diagram.

Thanks

Paul N

 

[2N6658]

2SA1386/2SC3519 are not available at Mouser !!!
I wanted you just to put the new 2SA1386/2SB3519 on the good channel to check if they are good.
"The amp is 17 years old, new output transistors all round is, IMO, good practice. You clearly disagree". Yes I completely disagree: transistors are not like lamps, good transistors (not very old ones like AD149, germanium SC...) don't get old : they can work during very very long time with no trouble : for small transistors (low current) they can work million hours...

As I see your PCB (down right) : it's very burned, others things than transistors must have died (resistors). Did you check the trimmer ?

 

[needsp]

I ordered them from Farnell.

The burnt area contained the remnants of the most serious casualties. TR19 had exploded, TR16 was incinerated, together with R31.
I had to repair the pcb too.

I'll check the trimmer, though it seems fine.

Thanks

 

[2N6658]

Did you change Tr15 and Tr17 and D13 D14 D18 ?
How do you test the transistors and diodes ? What kind of tester ?
Check R25, R26, R7

 

[needsp]

The following had failed and I replaced them-

TR8, 12, 16, 19, 15, 17, 18, D13. Since then I've replaced all the remaining semiconductors too, including D14. Without effect- as I expected really.

I tested them initially cold in circuit with multimeter's diode tester, and again after removal.

I'll recheck the resistors, and do the other measurments

 

[needsp]

Hi

Fixed! Measurements showed TR16 had failed, despite earlier replacement. Presumably when the incandescent lamp limiter lit originally during repair.... I'd swear it measured ok recently, but maybe not

When I found it tonight I was pessimistic as a probe slipped causing a short while I was measuring it! I assumed I'd wrecked it again. (I was again using the incandescent light limiter- so simple, yet so protective ) But bias now adjustable.

I'll tidy it up and give a good testing after resetting the bias. Thanks 2N6658. A useful discipline writing everything down. As a reward, I won't replace the output transistors in the other channel

Many thanks

Paul N

 

[Ian Finch]

Good work. That's quite a list of failed semis but then, some designs are really prone to failure in a chain of events that often go right back to the voltage amplifier or even the input stage. This all usually happens simultaneously - in the flash of an explosion in this case.

It's difficult for DIYs to prove the point but semis can also fail due to "second breakdown" which is microscopic, localised damage to the chip that accumulates with repeated peak current events that may only marginally exceed the transistor's SOA. Eventually, the semi will fail just the same and power semis will likely take out their drivers too.

I have also been tricked by triple output stages like this one where the driver seemed to have failed after it checked OK in the circuit. Since the driver had failed, the Vbe I was measuring was actually that of the pre-driver. Ah well...