Sony amp bias issue

[ITAM805]

hi folks

I have a Sony TA-F442E amplifier with one dead channel due to a shorted output pair. Before replacing them I tested the rest of the semiconductors and all appeared good. I noticed that 'R521' 0.1R emitter resistor had been replaced, a witness to there being a fault on this channel before the o/p pair went pop

I replaced the transistors with genuine Toshiba's (at least they look and test so) and tested the amp with no load. No magic smoke happily and the o/p protection light turns green, but when trying to set the bias current, to 50mA (5mV at the test point) as per the service data, it wont go below 10mV. But more alarming is a slight turn of the trimmer RV500 anti-clockwise the voltage shoots up to between 30-56mV - 560mA !

The base voltages on Q510/511 is around 0.65V and not +/-1.1V as per the diagram, unhelpfully so are they on the 'good' channel, who's test point does measure 5mV


Any suggestion what might be happening here?

[Andrew2]

Although we were Sony agents, I never had one of these on my bench. Mysterious burnups in big-ish amps are often caused by ultrasonic oscillation due to faulty caps, bad repairs or zobel networks gone o/c. The funny behaviour of that bias pot means it is either faulty or the amp is bursting into oscillation. Have you scoped the output?

[ITAM805]

Thanks Andrew, I fitted a new trimmer but it made no change. Thanks for the oscillation tip, I'll have a look

[Radio Wrangler]

Oscillation is more common than everyone thinks. It's why the scope is my first instrument, even if I'm only intending a DC test.

If not oscillation. Check the power transistor emitter resistors for value and dry joints. Check resistors and that diode around the vbe multiplier. Also check voltages across R150 /R610 There could be too much current in the VAS.

David

[ITAM805]

Thanks guys,

well Andy got it in one, the scope shows it's singing away and so no great surprise it burnt the output pair. Time to go through the caps now ...

[Diabolical Artificer]

"Oscillation is more common than everyone thinks" It's usually one of the first things folks suggest when I have a query, mostly regarding odd behaviour of a valve or amplification stage. Had it recently in a valve amp i'm building, music sounded fine, but 27vRMS on a DMM with no input gave the game away. Also I had it on a HH S500D amp, used it for years, sounded fine. Before selling it on I tested it, it functioned better as HF oscillator than it did as an amp.

What fault condition in regards to yon caps that could be causing the osc?

Andy.

[Radio Wrangler]

There are two sorts of instability. Local RF oscillation which can happen in a single device or a small group, and overall feedback loop instability.

Confronted with a honking amplifier, it's difficult to tell the two apart. The crap seems to be everywhere, and it is everywhere.

One approach is to modify the feedback network and put in a really slow pole to dominate alll else. The feedback now only stabilises the DC conditions and you now have an amplifier with an awful lot of gain. This local instabilities will still show up, though you might have to apply a DC signal to move it out of the zero volts DC region to get them singing. You can also now measure the open loop gain and phase response - which will give clues to the causes of overall loop instability. Of course, you need to know what the results should be! So you need to do some analysis, or measure a good one under the same conditions.

For overall instabilities, usually something is introducing excess phase shift before the gain is rolled right off and put to bed.

People are reticent to think of phase. It relates to trigonometry and the fear of that engendered during school days lives on for almost everyone. But it is crucial in these matters. Gird your loins, and treat your protractor as your friend.

In the case of this amplifier, I'd ask if the transistors in it were A) slower than the originals, or B) faster than the originals in terms of Ft at realistic operating currents. I believe the amps originally had Sanken power devices and try finding those which are not fakes!

Random musings

David

[Andrew2]

Several years ago I began to hear some odd interference covering about 30 kHz of Topband (1.8 to 2.0 MHz). It was a strange, rhythmic gurgling noise. It was very weak but it kept bothering me, so I took a bearing with the Wellbrook loop and found it was coming either from the north or south.
It couldn't be heard on my portable as it was too weak a signal, so I set out to the south and after walking about half a mile and not hearing it I turned around and walked to the north.
Lo & behold, after about 20 minutes walking I began to hear it getting stronger and it got very loud as I approached The Stone Jug, a pub about ten minutes away from home.
The signal was strong enough to recognise as music, and it was FM and sounded passable if I tuned to one side to slope detect it.
The pub was open so I called in ostensibly to wet my whistle, but really to play 'spot the amplifier' - and there it was, an H&H 100 watt thing. Making conversation, I said something along the lines of 'oh look at that amp, I used to use one of those when I played in a pub band years ago. Is it providing the music?'
'Yes', said Mine Host, 'but it's nothing but trouble, it keeps popping fuses and has to be sent away...'
I made my excuses and left, as I knew if I stayed any longer I wouldn't be able to resist a repair and I probably would have regretted it!
The music stopped a few days later and never returned.

[Radio Wrangler]

It's a frequent cause of 'things that go bump in the night' problems that never get explained.

Too few people understand stability issues. The maths scares away undergrads on degree courses and they take up other options. Consequently, there's an awful lot of marginal stuff around, even from names you'd expect to be reputable.

David

[Ted Kendall]

I came across a Kenwood L-07C preamplifier from an estate a while back. The guy had evidently had it apart, changed this, changed that and got nowhere. By the time I got hold of it, most of the outer panels were missing, but I found some more and set about it. This was hairy stuff, to me at least - differential pairs, current mirrors, DC to light bandwidth (in an audio preamp, yet!) ...and it honked big time. Eventually I caught on that the little black flag caps supplying the vital few tens of pF on every stage had expired...

[Radio Wrangler]

Good find and a good diagnosis, Ted. A systematic problem with corrosion or degradation of compensation capacitors... That would certainly catch a lot of people out. Worth doing a write up on that one for posterity's sake.

David

[ITAM805]

Thanks for the comments guys, good stuff

David, the o/p pair I replaced were Tosh as are those on the working channel, but whether Sony originally fitted Sanken in the early 90's I cant say?

There are four electrolytics on the o/p stage and they all test within spec, that said, the ESR is a little higher than new ones I have here, I'll swap them out and try it

Ted that's interesting story as the pF's caps are something I always assume are ok, but that's not the case by your experience ! If new caps dont fix it then maybe I should try temporarily bodging some in strategic places?

[Ted Kendall]

Could be worth checking, although usually I regard small polysomething caps as being pretty reliable, but it appears that these Japanese black flat polystyrenes often go bad, either low in value or completely o/c - I had both. Unfortunately, some of the Japanese super-heavyweight power amplifiers used them as well, with results which may be imagined. I suppose a write-up would be fun - I'll get my notes out...

[ITAM805]

Finally got round to having another go at this amp... Fitted the new caps but unsurprisingly, seeing the originals were measuring good, it was still oscillating?

Looking round the board I spied two unpopulated through holes near the drivers, with capacitor screen prints on them. I wondered, could they have been used to tame the amp on the original design but found to be unnecessary in production amps?

So I stuck a couple of 100pF (1kV - it's what was to hand) in said holes and turned it on, oscillation gone and the bias trimmer now working at normal !

Still have to put the amp through its paces but it's hopefully now fixed

Cheers everyone for the help

[Ted Kendall]

Value engineering forever! Good show.

[Radio Wrangler]

There was an old joke that Max Grundig designed a radio then started taking parts out until it stopped working. He put the last part to be removed back, and that is what went into production.

David

[Ted Kendall]

This may have started for reasons other than economy - when Grundig started, the occupying powers forbade the sale of working radio sets, so he sold them minus one bit. The customer then had to buy said part separately and fit it...

[ITAM805]

Well you've gotta laugh, while testing the amp at power @ 10kHz for just a few seconds the "good" channel decided to divest itself of the o/p pair - pop

The drivers check out ok so I dont know what happened, I was monitoring the signal on the scope when the protection relay clicked off and R617 started to smoke. Oh well

I happen to have a matched pair of Toshiba 2SC5200/2SA1943 which are higher rated, although the case style is different they will fit so I'll stick them in and see what happens ...