Hello all, I have been working on this amp for some time and am running into an issue where one channel will not bias. The Output transistors have the 80V +/- collector voltages, the Emitter voltage correctly sets very close to 0mV and the Base voltage is adjustable just like the good channel at .3 to over 1.3V.
I have tested all of the resistors in the output circuit as well as many other components on this channel. The only outliers in terms of voltage differences I have found is that there is no voltage drop across the emitters nor across resistors R39 and R40 which lead to the +/- base voltage supplies.
I could go into more detail regarding the work that has already been done but I wanted to make my question as concise as possible.
I have tested all of the resistors in the output circuit as well as many other components on this channel. The only outliers in terms of voltage differences I have found is that there is no voltage drop across the emitters nor across resistors R39 and R40 which lead to the +/- base voltage supplies.
I could go into more detail regarding the work that has already been done but I wanted to make my question as concise as possible.
Try measuring the voltage between R35 and R36 as you turn the bias up. You need enough voltage to overcome six base/emitter junction forward volt drops which means you should see approx 0.6 volt times six or 3.6 volts in total.
Compare that voltage with the good channel.
(If semiconductors have been replaced then its possible modern devices have a slightly higher forward voltage and that is putting the bias out of range)
Compare that voltage with the good channel.
(If semiconductors have been replaced then its possible modern devices have a slightly higher forward voltage and that is putting the bias out of range)
A few things I've found during my work with this amp, at one point the bias pot failed and went to very high resistance 5k+ ohm. It did show bias at that point however it was unstable and a resistor on the amp driver board smoked. At that point I shut it down and replaced the pot with the correct 0 - 200 ohm.
While I had two meters set up and doing a resistance check from ground to pin 37 the unit showed 33mV bias, measuring ground to pin 36 -33mV and if I perform a resistance test between 37 and 36 it shows 66 mV but it is not adjustable under these conditions. (Pin 36 and Pin 37 lead to the bias test points and the emitters).
When adjusting the offset to far outside 0V the unit will also start to show a few mV bias.
I doubt any of that helps but just in case..
While I had two meters set up and doing a resistance check from ground to pin 37 the unit showed 33mV bias, measuring ground to pin 36 -33mV and if I perform a resistance test between 37 and 36 it shows 66 mV but it is not adjustable under these conditions. (Pin 36 and Pin 37 lead to the bias test points and the emitters).
When adjusting the offset to far outside 0V the unit will also start to show a few mV bias.
I doubt any of that helps but just in case..
A range of 3.3 to 4.9 volts should be plenty.
Put your meter negative lead on the main output node. That is the junction of all the 0.33 ohm emitter resistors. Point 92 on your diagrams.
Now measure forwards from that point to points 94 and 95 which is the base of the output transistors. You should see no more than around +0.65 volts on point 94 and - 0.65 on point 95.
Now move to the emitters of TR14 and 15. You should see the same values.
Next the emitters of TR12 and 13. You should see another +0.65 and -0.65v added to those values. So no more than -1.3 volts and +1.3 volts on those emitters.
Now go the base of TR12 and 13. Again another 0.65v added to the totals. So you should see about +1.95v on TR12 base and -1.95v on TR13 base.
Finally measure on the left hand of R35 and R36 that connect to the diodes.... and somewhere along the line you should find the problem. If you had 4.9 volts between those resistors then you have over 0.8 volts 'available' for the six base/emitter junctions and the amp should be melting.
(a failed driver is likely. Any of the those drivers and pre drivers with more than 0.65 volts or so across base emitter is going to be faulty)
Put your meter negative lead on the main output node. That is the junction of all the 0.33 ohm emitter resistors. Point 92 on your diagrams.
Now measure forwards from that point to points 94 and 95 which is the base of the output transistors. You should see no more than around +0.65 volts on point 94 and - 0.65 on point 95.
Now move to the emitters of TR14 and 15. You should see the same values.
Next the emitters of TR12 and 13. You should see another +0.65 and -0.65v added to those values. So no more than -1.3 volts and +1.3 volts on those emitters.
Now go the base of TR12 and 13. Again another 0.65v added to the totals. So you should see about +1.95v on TR12 base and -1.95v on TR13 base.
Finally measure on the left hand of R35 and R36 that connect to the diodes.... and somewhere along the line you should find the problem. If you had 4.9 volts between those resistors then you have over 0.8 volts 'available' for the six base/emitter junctions and the amp should be melting.
(a failed driver is likely. Any of the those drivers and pre drivers with more than 0.65 volts or so across base emitter is going to be faulty)
Lets look a bit closer at the results 
That is where things fall down when we factor in the earlier measurements.
Lets looks at the 1.82v and -1.818v you measured. Those voltages are OK but would suggest being just at the point the output stage should start to conduct. If we add those together we get 3.63 volts/ Divide that by six (the six B/E junctions we must forward bias) and we have just 0.6 volts per transistor as an average. Its not quite enough.
The crucial bit is this:
If you could get up to 4.9 volts then we would have 0.81 volts per device. Plenty
So we need to know where the error is occurring. Begin by seeing if you can still reach 4.9 volts between R35 and R36 and if you can then the six voltage measurements should all add up to that 4.9 volts.
That needs rechecking before we look any further.
That is where things fall down when we factor in the earlier measurements.
Lets looks at the 1.82v and -1.818v you measured. Those voltages are OK but would suggest being just at the point the output stage should start to conduct. If we add those together we get 3.63 volts/ Divide that by six (the six B/E junctions we must forward bias) and we have just 0.6 volts per transistor as an average. Its not quite enough.
The crucial bit is this:
If you could get up to 4.9 volts then we would have 0.81 volts per device. Plenty
So we need to know where the error is occurring. Begin by seeing if you can still reach 4.9 volts between R35 and R36 and if you can then the six voltage measurements should all add up to that 4.9 volts.
That needs rechecking before we look any further.
I'll look later but that is a massive clue.
If you have 4.9v between the left hand sides of those resistors then you should see essentially the same between the other ends. The current gain of the three transistors (three per upper and lower half) is massive and so those resistors should drop virtually no voltage.
I think one or more of the drivers or predrivers is faulty.
So I may be confusing things. The 4.9v is being measured from the base of tr12/ 13 in reference to each other. When testing on the left side of the r35/36 I was referencing the common node.
To keep things simple I measured again at the right side of 35/36 across those two points i show 4.9v the left side of the resistors across 35/36 also show 4.9v.
I went ahead and swapped the drivers (tr14/15) with like for like new parts. Both channels are running on semi mje15032/33. Changing them in the problem channel made no change.
To keep things simple I measured again at the right side of 35/36 across those two points i show 4.9v the left side of the resistors across 35/36 also show 4.9v.
I went ahead and swapped the drivers (tr14/15) with like for like new parts. Both channels are running on semi mje15032/33. Changing them in the problem channel made no change.
Last edited:
That's correct. So red and black meter leads as marked here and you measure 4.9 volts. That should be melting the outputs.
Does this make sense. Here is a simplified diagram (simulation) of the output stage with 4.9 volts applied. The bias current is over 400ma. All those base/emitter drops add up to 4.9 volts.
Its just simplified. First pair are TR12/13, second pair TR14/15 and the final pair is one set of outputs. No point adding parallel pairs for a quick simulation.
I looked a bit closer at the circuit when you said that. Are they 4.7 or 47?
Have you compared the value with the good channel?
If they were high or open then that suggests an overload has happened.
Typing this in real time so I'm not altering it
I see there are two sets of R39/40. One set are those I have on the sim and the others are the series base stopper resistor to the outputs which are 4.7 ohm.
The theory should hold up. If you have 4.9 volts across the bias generator then you should see all the base/emitter voltages add up to that total.
Make sure that you have continuity from the actual leads on all the transistors to the appropriate print. Its easy to crack a pad/print node and find for example the collector is actually floating and not connected to the supply.
The r39/40 on the right of the schematic ( on the ps board) show 4.7 ohm and the r39/40 on the left (amp driver card) are 47. Both of these values match the good channel. There was an overload at one point as when I recieved the amp it had shorted driver transistors.
The 4.7 ohm ones show 1.2v and -1.2v and show no voltage drop.
The 47 ohm ones show 1.2v and -1.2v on one side and .04 on the other side. With the meter across these it shows a ~1.12v drop.
(These numbers aren't perfect as they seem to fluctuate a tiny bit when turning off and on to reorient my probes).
The 4.7 ohm ones show 1.2v and -1.2v and show no voltage drop.
The 47 ohm ones show 1.2v and -1.2v on one side and .04 on the other side. With the meter across these it shows a ~1.12v drop.
(These numbers aren't perfect as they seem to fluctuate a tiny bit when turning off and on to reorient my probes).
Those voltages do seem to show a problem.
Look at the circuit
The 0.04 volts is the common point of the two resistors and is actually the output line. The other end of each resistor has around 1.2 volts on it. That 1.2 volts is supposed to appear across the base/emitter junction of each output transistor.
Follow that 1.2 volts and it goes into the 4.7 ohm and from there into another 4.7 connected to each base.
So we have +1.2 volts on R39 on the driver board. That means you have 1.2 volts on the 47 ohm (also R39 on the PS board). That goes to terminal '94' which goes to the three 4.7 ohm base resistors.
If you have +1.2 volts on those three base resistors then something is wrong because the base/emitter junction of the output transistors should be limiting that to around 0.7 volts. The output node is at 0.04 volts that you measured. Something must be open or incorrectly wired... something basic is going wrong here
Work through it and check where the voltages fall down. If you have +1.2 volts on each side of those three 4.7 ohms then measure the the emitter volts and see what that is.
Look at the circuit
The 0.04 volts is the common point of the two resistors and is actually the output line. The other end of each resistor has around 1.2 volts on it. That 1.2 volts is supposed to appear across the base/emitter junction of each output transistor.
Follow that 1.2 volts and it goes into the 4.7 ohm and from there into another 4.7 connected to each base.
So we have +1.2 volts on R39 on the driver board. That means you have 1.2 volts on the 47 ohm (also R39 on the PS board). That goes to terminal '94' which goes to the three 4.7 ohm base resistors.
If you have +1.2 volts on those three base resistors then something is wrong because the base/emitter junction of the output transistors should be limiting that to around 0.7 volts. The output node is at 0.04 volts that you measured. Something must be open or incorrectly wired... something basic is going wrong here
Work through it and check where the voltages fall down. If you have +1.2 volts on each side of those three 4.7 ohms then measure the the emitter volts and see what that is.
Measuring at the 6 (4.7 ohm) resistors in the output I am seeing ~1.2v and -1.2v on both sides of these resistors.
The emitter is showing ~.035v on both sides of the emitter resistors.
I guess thats the whole question. At these voltages why wouldn't these outputs be turning on? (I have swapped the output transistors from the good channel to the bad without change)
The emitter is showing ~.035v on both sides of the emitter resistors.
I guess thats the whole question. At these voltages why wouldn't these outputs be turning on? (I have swapped the output transistors from the good channel to the bad without change)
Lets be absolutely clear on this. You see that voltage if you:
1/ Use ground as the reference for the meter. (We assume the DC offset is close to zero)
2/ Use the output node (junction of the 0.33 ohm resistors) as a reference.
3/ As a final check measure directly across the base and emitter pins (be careful) of any one output transistor and confirm there is 1.2 volts present.
If you have 1.2 volts across the base/emitter junction as measured on the pins then something very suspect is up with those transistors. Could they have been replaced with vertical FET in a past life?
The theory won't fail us. If those transistors have 1.2 volts across B and E then they are not what they seem.
Don't assume anything
check check check. that there isn't something like a cracked bit of print across the board and that the transistors are in some way isolated from the components around them.Also... if you measure that 1.2 volts then I would expect that voltage to alter with the bias preset. Check that. The preset should be able to take that voltage much lower.
Testing B-E across each of the 6 outputs show the ~1.2V. In reference to ground and the common emitter node it shows the same. (That 1.2 does adjust with the bias pot being adjusted. ~.3V - ~1.3V)
I went ahead and installed six new MJ21193/94 and it made no change, old outputs are back in now. The outputs that are in place currently were working on the good channel so I don't necessary suspect they are the issue.
Could something downstream on the emitter circuit cause something like this? When turning the unit off or on I do see the Bias spike up momentarily.
I went ahead and installed six new MJ21193/94 and it made no change, old outputs are back in now. The outputs that are in place currently were working on the good channel so I don't necessary suspect they are the issue.
Could something downstream on the emitter circuit cause something like this? When turning the unit off or on I do see the Bias spike up momentarily.