The past few days, I have been occupied with other matters of more immediate importance. Consequently, I have been unable to devote the time needed for troublshooting this amp. I should have those other matters sorted out soon, possibly even today. When this bumpy patch is behind me, I will continue with following the most recent suggestions re this amplifier.
just out of interest, why are you running it at potential full current, you are much better throttling it back a bit while testing and less likely to break something else in circuit while doing so.
Everyone has their own methods
and I do stand by the methods that I've outlined. I've also tried to keep it as simple as possible to try and get a handle on what is going on.The thought process is that if the vbe multiplier is linked out it is taken out of the equation. It can be discounted. With no load attached no current should be able to flow anywhere in the output stage in this state no matter what the front end does. A check of DC offset (if zero which it is) shows that the front end DC conditions are correct. If there is a DC fault causing a sudden offset to a rail then the output stage should still draw no current.
Something else seems to be going on here and a really important question is the failure mode of the output transistors, and that is whether they really are failing short circuit from C to E? hence:
If @62vauxhall could measure them on the diode range and just see if they read very low from C to E that would be informative
Hopefully we'll get to the bottom of this one.
the present shorting between C and E - at least on one of both output devices (maybe of both) is extremely obvious to me, additional even on the driver stage of output buffer. I conclude this from the observed description in post #1 - quote
As per the service manual, the target is 25mv and the left channel bias was pretty close. The right channel was nearly 60mv so I attempted to adjust that channel's bias pot.
As I rotated the pot, both 1.6A fuses blew and I noticed the telltale odor of an overheated component. Even now, I fail to see anything that shows heat damage.
As I mentioned before, this happens due to the deficiency concerning the connecting kind of the pot for BIAS adjustment.
Because not all semiconductors were replaced after this event (output buffer, VA-Stage and Vbe multiplier), the CE shorting of Vbe transistor don't help to find out remaining defective transistors.
As per the service manual, the target is 25mv and the left channel bias was pretty close. The right channel was nearly 60mv so I attempted to adjust that channel's bias pot.
As I rotated the pot, both 1.6A fuses blew and I noticed the telltale odor of an overheated component. Even now, I fail to see anything that shows heat damage.
As I mentioned before, this happens due to the deficiency concerning the connecting kind of the pot for BIAS adjustment.
Because not all semiconductors were replaced after this event (output buffer, VA-Stage and Vbe multiplier), the CE shorting of Vbe transistor don't help to find out remaining defective transistors.
Last edited:
Wouldnt't the emitter resistors of those be seriously heating up and smoking in that case?
The 0,7V measured some posts above are a bit far from rail voltage.
The problem is that while measuring voltage over the emitter resistors the rail voltages were forgotten...
@tiefbassuebertr we will have to agree to differ in our methods
Confirmation (or otherwise) that the output transistors that appeared to fail when the DBT was in use are in fact reading short circuit could provide a clue. What I'm trying to get at with that is whether the DBT lighting is really caused by a real transistor failure or whether it just appears to be so and the that cause of the bulb lighting is actually something else, maybe even something physical such as the mounting and insulation of the transistor and the act of removing and replacing it seems to provide a (temporary) fix.
Confirmation (or otherwise) that the output transistors that appeared to fail when the DBT was in use are in fact reading short circuit could provide a clue. What I'm trying to get at with that is whether the DBT lighting is really caused by a real transistor failure or whether it just appears to be so and the that cause of the bulb lighting is actually something else, maybe even something physical such as the mounting and insulation of the transistor and the act of removing and replacing it seems to provide a (temporary) fix.
I will confess to my sin of using that component tester to judge if a transistor is good or bad. A bad transistor causes the display to show: "resistor", "diode", "unknown or damaged part" or it triggers the component tester to enter a "self test" mode. But it does not explain to me in terms I understand about what specifically is wrong with bad transistors.
So just now, I removed the 3 remaining right channel output transistors. Q410 is still out of circuit as it has been since I jump wired the bases of Q405 and Q411. Q405 and Q411 are still jump wired.
So the transistors I just removed ar Q407, Q408 and Q409.
I dispensed with the component tester to usie a DMM in diode mode and followed 5 steps....B to E, B to C, E to B, C to B and C to E.
What the DMM displayed made me question if it was functiong properly which was 0.0 readings in all 5 steps for Q407, 408. & 409, To see if the DMM was functional, I checked the transistor from Q410 and saw voltage readings where they were supposed to readings and OL where it was supposed to be.
With all right channel output transistors removed, I used full power from the variac and DBT stayed dark. I left it powered up for 20 minutes and it remained dark. I thought that might be significant because with 3 right channel output transistors in circuit, DBT was a dull orange glow then suddenly went full on bright after a few minutes.
Presumably that transistion from dull orange to full bright was the result of an output transistor failing? Those were fresh new transistors not long ago.
And should I continue with the jumper wire from Q405 base to Q411 base?
If you put your meter on the Diode range and just do one measurement from Collector to Emitter of the removed transistors then that will give a good idea of whether they have failed. If the meter reads very low (such as 0.01 or 0.1) then they are zapped.
To take that one step further you should place the red meter lead on the collector and the black on the emitter for the NPN transistors and reverse the connection for the PNP's. A good transistor should show open circuit if you do that (same reading as with the leads not connected to anything).
Lets look logically and critically at that.
DBT dark shows minimal current flow which is what we want. So all good there.
The DBT glowing with three of the output transistors removed needs investigating. If we look at the circuit we can see removing any three of the four output devices removes either an upper or lower pair completely and leaves either a single NPN or PNP in place.
With just one transistor fitted there is no low impedance path current could flow in. A short in the output line (the line from CB1 or CB2 current breakers) to the speaker socket would be one possible route.
I thought I should clarify. The DBT glow I referred to was with 3 of the 4 output transistors in place, not removed. At that time, there was one output transistor removed - Q410.
That was done as instructed in post #109.
The "one transistor" I referred to was removed, not left in place. I have not attempted turning on power with just one only output transistor.
I can certainly look closely around those CB devices and the speaker terminals. By appearance, a neon lamp and a diode(?), stuffed inside a piece of clear nylon tubing, those look "custom made" for this amplifier.
It would help to know if they really are failing though. The one you removed and that stopped the DBT lighting... does it read short circuit or very leaky from C to E?
The CB devices will likely be some kind of auto resetting overload device and they will normally read short circuit like a good fuse.
All I can offer for thr one transistor I removed (an NPN) is that C to E reads OL. If helpful, reversing the leads, to E to C also reads OL.
And the rundown of readings I got from that transistor are as follows:
B to E .553
B to C .548
E to B OL
C to B OL
C to E OL
Reversing leads produced these readings:
E to B OL
C to B OL
B to E .549
B to C .543
E to C OL
First thing that stands out is the OL or 'open' from C to E meaning they have not failed short circuit and that is the most common failure mode for any power transistor.
Those look fine. The 0.5 result is the meter displaying the forward volt drop across the junctions B/E and B/C junctions which read like a diode. So all good there. The red lead would be on the base for that with a NPN device. So E to B and C to B with the leads reversed gives OL or open which is correct. No conduction from C to E which is the one that usually throws up as short for any failed power device.
Those look fine. The 0.5 result is the meter displaying the forward volt drop across the junctions B/E and B/C junctions which read like a diode. So all good there. The red lead would be on the base for that with a NPN device. So E to B and C to B with the leads reversed gives OL or open which is correct. No conduction from C to E which is the one that usually throws up as short for any failed power device.
Those don't compute if its the same transistor. Black lead on the base and red on the emitter and then on the collector should both read open. Maybe just a measurement error
Yes, you are correct. Black on B and red to E....then black to B and red to C are both OL.
Disregard that 2nd set of measurements I posted. It is the same readings as those in my first list but in a different order. It was not astute thinking on my part to include them.
The way I listed them was with the mistaken assumption that the the first letter was always DMM red and the 2nd letter was always DMM black lead.
I did a quick re-check of those other three output transistors that displayed 0.0 volts all the way around and used the DMM's audio continuity function. Not surprisingly, the audio alert sounded all the way around for all three transistors. So dead shorts on all combinations of BCE.
It seems to me that those transistors were damaged when the amp was under full variac power and I was using the probes to make voltage measurements across the emitter resistors. It was at sometime during that process when the DBT unexpectidly went from dull orange to full bright.
I insulated the probes metal sections except for the extreme tip. So I can say with some degree of confidence that I caused nothing to short. In fact, one incidence of the DBT's brightening, occurred when my hands were elsewhere - not near the board.
Starts looking like probe invoked oscillation.
What is peculiar about this schematics is that there are no base (grid) stopper resistors in use and with today's fast devices (in parallel) they may be necessary to tame the output stage.
Can you check for oscillation in any way - during next time DBT becomes bright?
Doesn't that require an oscilloscope? I do not have one of those.
Last year I learned that in the area where I live, there are no oscilloscopes to borrow or rent. From time to time I have seen online "for sale"ads for them, Typically in the $400 and up range and being on fixed income, my finances are limited. Condition is a gamble and drive time to the larger cities where they are usually located is 4 to 6 hours in each direction.
Peridically I do check for them locally on Facebook Marketplace.
An oscilloscope (or a spectrum analyser) with a bandwidth large enough to see the oscillation is the most convenient way to check for oscillations, but if I remember well, @indianajo uses a fast diode detector and a multimeter for it. Some signal tracers have a diode detector and a meter, others don't.
Bias points that make no sense at all and that change when you touch or just approach a node with a small insulated screwdriver are usually signs of oscillations.
Bias points that make no sense at all and that change when you touch or just approach a node with a small insulated screwdriver are usually signs of oscillations.
Last edited:
Interesting as is the rest of the others reading short. So they are zapped.......
That is a possibility.
Can you remind up what actual replacement types you are fitting for the outputs as this is a very old amp now. We're so many posts in I can't remember all the twists and turns
Have the driver transistors been replaced and if so with what?(I've intermittent internet access at the moment so may not get to look in as much for a while)
For possible ease of reference, I am including what transistors this amlifier was initially equipped with:
Original output transistors PNP were 2SA670.
Original output transistors NPN were 2SC1060.
Replacement output transistors PNP are MJE15033G.
Replacement output transistors NPN are MJE15032G.
____________________________________________________________________________________
Original driver transistors PNP were 2SA706.
Original driver transistors NPN were 2SC1124.
Replacement driver transistors PNP are 2SA1220A - accommodation was made for the different pinout.
Original NPN driver transistors are still 2SC1124 and not replaced - the original 2SC1124 transistors were maintained.
_____________________________________________________________________________________
Original pre-driver transistors 2SA678 were not replaced - the originals were maintained.
Original output transistors PNP were 2SA670.
Original output transistors NPN were 2SC1060.
Replacement output transistors PNP are MJE15033G.
Replacement output transistors NPN are MJE15032G.
____________________________________________________________________________________
Original driver transistors PNP were 2SA706.
Original driver transistors NPN were 2SC1124.
Replacement driver transistors PNP are 2SA1220A - accommodation was made for the different pinout.
Original NPN driver transistors are still 2SC1124 and not replaced - the original 2SC1124 transistors were maintained.
_____________________________________________________________________________________
Original pre-driver transistors 2SA678 were not replaced - the originals were maintained.
My old Towers transistor data book doesn't give the Ft (speed) of those original outputs but I suspect they are in TIP41/TIP42 territory which means they are many times slower than the replacements. Instability is a possibility.
Is there anything lost in trying some TIP41C and TIP42C outputs and BD139 and BD140 as drivers? I think you should replace both the drivers as a matter of course.
Is there anything lost in trying some TIP41C and TIP42C outputs and BD139 and BD140 as drivers? I think you should replace both the drivers as a matter of course.