Just wondering if anyone has a transistor catalogue from the 1970's and could look these numbers up for me? there's practically nothing on the net about them. as far as I can tell they are low voltage 12v?
there was lots of manufactures of transistors in 1970 era. I do not believe it is in any of the manuals that i have from that era. I am not familiar with that part number. It could be a specific mfg house number for all i know.
Providing information, like where they are used, what they look like, schematics, model numbers, will help a lot in finding a replacement.
What's is the amp's Make and model? That's good place to start.
I noticed that MT numbers make sense if you use MPS instead and these actually are TO92 Motorola types. The board looks to be of Asian manufacture though, so I'm not certain.
The amp is an eagle a6400 and I have been unable to find a schematic, It was sold in the UK in the 1970's but made in Korea (same factory as Sherwood, Goldstar etc.)
So MT 9680 is a PNP TO92 12v 350 mw 100ma HFE 45
MT9630 is the NPN equivalent.
Any suggestions for substitutes in those positions? I have sa56 and sa06 available
Thanks to all for your help it is much appreciated
Without a schematic, It would be risky to recommend equivalents but based on the Motorola types, virtually anything would work for the bias (presumably Vbe multiplier type) transistor. The risk would be excess gain but not likely there. I can't say much about the PNP but presumably there isn't much Vce to worry about. If that checks out, probably fine. Otherwise, play safe and fit MPSA92/42 which should cope with any voltage likely in the amp.
Thank you very much Ian, I'm just installing the drivers at the moment BD239/140 and going by the board markings 2 go emmiter to output base and 2 go collector to output base does this sound right? You can probably tell that my skills are not that well advanced at the moment but I am learning lots with this amp because not having a schematic is making me think more about what does what.
An amplifier of the 1970 period could well have been a Quasi-complementary type, where one half of the output stage is configured as a Szlikai (complementary feed back or CFP pair) and the other as an emitter follower (EF) pair. That enables push-pull operation with same (usually NPN) type upper and lower push-pull output transistors. The CFP half is driven from the collector of a complementary driver pair as you say and the EF half is driven by the emitter.
Here's the basics of the two different output stage types and the "quasi" type compared, with some informative notes. It's a link to a large and popular site with reams of info. on just about any aspect of power amplifier design and construction - well worth a read or several actually. Compound vs Darlington
A question though; these transistors are US/European types and seem odd in a Korean product. It was manufactured for a UK company for sure, but this usually involved the manufacturer's local parts sources. Has someone attempted repairs before and substituted whatever parts were available, do you think? If you know the power supply voltages, you can at least tell what transistors would not be appropriate. BD139/140 for example, can only be used up to around +/- 35V supplies or a single supply of 70V. If the power output is rated much more than about 50W/8 ohms, something is amiss.
Thanks for the info Ian, I think all transistors were original as most had strange alloy covers on them. The supply voltages are about 39 volt while on my dim bulb tester so probably a little higher on direct mains. the amp is around 45/50 wpc
I did get it working today on headphones but something shorted when I cranked the volume up. I did notice though that the volume was considerably louder when switched to mono. would it be feasible to have 3 drivers going emitter to base and only one collector to base?
Hi, good to hear it works - a little at least. It is possible to have headphone level output with as little as one driver transistor and with output transistors both blown too, so don't get too hopeful of a result yet. We (me at least) still need to establish what the output transistors are. As before, if they are all the same type then it's a Quasi-complementary design and Fig.6 in the linked ESP article applies. The terminolgy is different but probably more familiar to hobbyists, in that Darlington is used for EF and compound pair is used for CFP in the article.
Draw at least this much of your amplifier's circuit out for comparison. There will be differences and a few added components (perhaps a Baxandall diode+resistor in the CFP side) but that should at least clarify the pinout question. You will find the ESP articles a goldmine of reliable explanations and rules for what can and can't work. Keep reading to get a grip on this stuff, including the overview article on DIY repairing - very useful and vastly better than over-enthusiastic "You-tubes".
Note the 3 diodes shown at the input are an elementary form of bias generator, for simplicity. As you have a bias transistor, yours will be a more common, adjustable Vbe multiplier circuit. The resistor values shown are only ball-park and yours will be less - around 100 and 0.33 ohms might be closer values to the 270 and 1 ohms shown.
You could check all transistors for Vbe (base - emitter diode voltage drop) in the powered amp, measuring about + or - 0.65V in healthy transistors. As the output transistors are socketed, you could remove them for testing in isolation. They will need cleaning up and replacement of the heatsink grease and probably new mica washers on reassembly. Use the correct transistor type grease- not the sometimes conductive types used for CPU cooling. There should not be any C-E shorts in the unpowered amp or as isolated transistors, which is by far the most common failure mode.
What do you measure at the output terminals? A DC voltage? I assume there is no output protection relay and the amplifier has a limiter circuit to protect against shorts and overload. This is usually reliable (a couple of general purpose TO92s, 2 diodes and a few passives) and many early direct coupled power amplifiers were fitted with them because the implementation was cheap and simple. It doesn't always work in old amps or over long periods and that can be a reason why these things wind up in lofts, garages and Ebay.
I don't understand what you mean in reference to 3 drivers. Each amplifier output stage has 2 driver transistors (BD139/140) and presumably 2 x TO3 output transistors of some type(s). If you refer to all the transistors as drivers, do I assume you mean all the output transistors mounted on the heasink? If so, then you are asking should one channel be of different design to the other, which won't be the case. Recheck the actual driver transistors fitted to the PCB for correct pinout.
You probably don't need the advice but watch your probes don't slip when checking output stages. They have a habit of finding ways to make sparks and melt silicon whilst you turn your attention to the meter display. Use IC clips, hooks, small clip leads or even tack-solder the probes.
On the other hand, MJ802/4502 are about biggest, meanest TO3 audio transistors around but with lower current gain. I wouldn't substitute those, since in a fault, they'll see off anything associated with them in the amp, including the PCB, causing more damage than you may care to think about.
These are still overkill but much better substitutes:
MJ21193G for the 2SB
MJ21194G for the 2SD
Avoid old parts, choose new stock, genuine On-Semi brand. Unfortunately, these parts are not a match to the the Japanese audio transistors used but you probably won't have problems at the power involved here. Refurbishing very old amplifiers generally, is unrewarding unless you can source genuine or better suited parts, if affordable and available. It isn't a matter of buying the highest of every specification listed - its about best match them to the design requirements.
If you look at Fig 7A and B here: Compound vs Darlington you can see two possible arrangements of your output stage. In either arrangement, the driver emitters all connect to the appropriate output transistor bases, so you still have an error there if you have one driver collector connected to its output transitor base either in the PCB markings or your reading of the PCB tracks. It certainly is an unholy mess of parts styles, types and manufacturers in there. Assuming they are original components, it doesn't inspire confidence in the product.
If you want to repair amps you have to study their design because you can't repair by numbers or try this/that when there is no schematic. It may be a bridge too far unless you come to understand the essentials of design. At least you do know that some output transistors are blown (likely shorted E-C) and resistors have burnt due to the high fault currents. The drivers are also likely burnt but that may be all, if you have a little output leaking through, as you say. The "dim bulb tester" should point out any repair mistakes, when you do get the appropriate parts and insulating kits (washers + insulating bushes). First, remove all the remaining output stage semis and before replacing them, test the amplifier for bulb brightness - there should be very little then. You then need to determine an appropriate bias current to set and the fun begins.
As you are already conducting a long series of short posts and pics at AK and making progress, stay with them or not - no point in playing one forum thread and members against another and omitting full replies to questions in either.
Thanks Ian, turned out to be a printing error on the board so have corrected and the amp is now working, just got to set the quiescent for the drivers trying 3mv to start with. I do appreciate all the time and effort you have given me and I have learned a lot with this project. I had no intention of playing one forum against the other just trying to cast the net as far as possible to get answers and I must say there are some great guys on both forums.
For minimum crossover distortion, EF designs require more bias than equates to 3mV measured across the output transistor emitter resistors. It's a simple Ohm's Law relationship where Ibias = Vre/Re. Assuming Re is 0.22 ohms, Ibias would be 14 mA (actually about perfect for a CFP design)
These replacement output transistors are pretty sluggish, since they are similar 1 MHz types to 2N3055 and need as much bias as possible up to 26 mV, about optimum for the usual range of emitter resistors. They will then be dissipating quite a bit of heat. I doubt the design will be stable at that level though so I suggest you start at 10 mV and test for a stable reading at intervals over a listening session. Better still, build yourself a dummy load and test at real power rather the few average watts you and your neighbours may be able to tolerate at home. It will be a compromise result but I think you will already find the sound a tad harsh and unpleasant. I think it is, from your recording but it's unrealistic to expect accuracy with so many dubious extra processes between us, so it could be fine for all I really know.
Keep a close watch on bias and back off current at the first sign of increase without your making an adjustment. 'Best to replace the old trimpots actually, as any deterioration in them is bad news and will probably lead to disaster. Still, so far - so good and congratulations on your tenacity
Thanks Ian, I have set to 10 mv and amp still runs cool after 20 mins if I put a finger on an output transistor it is not warm to the touch.do I keep adjusting upwards until I can feel some proper heat. Amp sounds nice as it is but I want it to be as good as possible. I think all the modern parts agreed with the old girl.