Output Transistor blowing - B&O Beomaster 901

[HatOfTheCat]

Originally Posted by samjmann

2c131 is shown as across 2R181 a 47k. I don't have a layout to confirm that in practice this is the way it was fitted.

Its definitely across B/E of TR27. The pcb layout from the service manual I have shows it exactly as it is on mine, see reply #114

Originally Posted by Radio Wrangler

The quiescent current setting pot is an unexploded bomb. If the slider skips a little, the Vbe multiplier transistor turns off without base current.

I have my suspicions of the pot/s myself. The originals were a little crusty (actually the LEFT was far worse than the RIGHT) so after the last failure I replaced both with ones from "service kit" of parts from Beoparts in Denmark. I did have a poke at both the old ones with meter and they seemed to give about same resistance at given settings but were far from "buttery smooth". Even with the new ones was like trying balance a grain of sand on a pin head to get the setting when checked their operation under dim bulb. Even slight pressure on the screwdriver, not actual movement, is enough to change the measured voltage across the 0.39 Ohm resistor. After this first successful mains power on, knowing this, I gently wound them up to about 2/3rds of the way round the original ones were positioned then it was just micro nudge, then more just the slightest of pressure with the screwdriver until it got to where I wanted. With them set to give as close to 10mV as I could manage, on power on the voltage seemed to go to about 14mV before coming back to setting in about 2 or 3 secs. Remeasuring after a few hours of operation they seemed to have settled okay and they've not drifted off. Unfortunately this was just one of the half dozen "fixes" I did between last blow up and successful full power on so its just on the "possibles" list and not a "confirmed kill".

[Studio263]

Ive mended loads of these over the years and have never considered the design marginal. They used the same circuit in a whole group of similar models all of which seem to be pretty sturdy and tolerant of abuse - the only fragile bit seems to be the front end MOSFET in the FM tuner which does give a fair amount of trouble.

When the output transistors do fail (normally as a result of long term use after the bias preset has changed value) I use originals from old B&O stock to replace them - there is a lot of gain in this part of the circuit so it needs to be right.

As far as I know the complete group of models which use this output stage circuit are:

Beolab 1700
Beomaster 800, 901, 1100
Beocenter 1400, 1500, 1600, 1800, 3600

Had there been something fundamentally wrong I'm sure they'd have worked it out after designing all that lot!

[Radio Wrangler]

It might be a good idea to fit a nice reliable fixed resistor from the pot slider to the pot's positive end. It'll change the optimum running position a little, but if the slider jumps, the Vbe multiplier transistor will turn hard on bringing the quiescent current to zero, which is the safe condition. At a guess somewhere between 22k and 47k ought to be good.

David

[HatOfTheCat]

Originally Posted by Studio263

Ive mended loads of these over the years

I'll leave the "fragility" questions alone as they are beyond my electronics skills although its definitely fixed with "new old stock" Texas Instruments output transistors as were originally fitted. However can I pick your B&O specific knowledge brains as my next task is to restring the tuner, the peg on the capstan was broken off (found it rattling in the case) and although string is intact, complete with the spring on end, I'm scratching my head as to how to go about putting the string back on once I've put a replacement pin in place. Is there a way to do it in simple steps ?? All the service manual shows is that its 3 wraps around the capstan but how you get to that point with tuning dial setting right is evading me

[stuarth]

Since the unit now works with NOS transistors, perhaps the original B&O designers showed the same lack of foresight as those at Quad, and designed with the transistors they had available at the time....

The switch on spikes reported by Rich’s simulation do look rather alarming and could well be the cause of failure when switching on with “modern” transistors and no lamp limiter, although the design may not have huge safety margins with the old transistors.

Well done Hat for persevering for so long.

Stuart

[Ted Kendall]

Originally Posted by stuarth

Since the unit now works with NOS transistors, perhaps the original B&O designers showed the same lack of foresight as those at Quad, and designed with the transistors they had available at the time...

You can only design around what you have and what you know, and hindsight knows no myopia. Walker himself said that if an amplifier lasted ten years, it had fairly done its job. It's fun and good practice to keep decent stuff going at several times beyond its design life, but to criticise designers for not anticipating that semiconductors would have ten times the bandwidth they were working with is a bit much.

[dave cox]

They wouldn't need hindsight to see the overshoot, just a 'scope
(it happens with the original transistors)

dc

[Ted Kendall]

Originally Posted by dave cox

They wouldn't need hindsight to see the overshoot, just a 'scope
(it happens with the original transistors)

dc

Granted, contemporary incompetence is less defensible. The 303 worked a treat, though.