Normally if a fuse blows on an amp circuit I would normally assume all the transistors are suspect, and replace them because of the possibility of "hidden" damage. By running the amp with a shorted Tr5 other devices may have been partially damaged. It is possible that the pnp input device is suffering due to a reverse base-emitter bias and has a gain loss.
The only way I would consider re-using a device is if it has passed all checks compared with a new component. That usually takes time and by the end of the exercise, finding it doesn't is a worse waste of time than just plugging in a new one. In your case, I would be curious to know if the pNP input transistor - or the NPN driver - have a gain loss. Does your DMM have a transistor gain test socket? You may not have the original info but you might see if the gain is lower than you would expect.
To check the gain of the individual channels you need to make sure that the input signals are the same. It may be that the transistor in front of the main amplifier (TR1?) is not behaving, but that may be due to the switches not working as there are quite a few to select options. With the mono/stereo on mono do you measure the same voltage at the inputs to the two channels (preferably with both fuses intact)? - that is on the collectors of the Tr1's. (You might want to make a temporary AC probe using (say) a 1uF capacitor and 100k resistor attached to the collectors to isolate the DC before measuring the signal on your DMM. Connect this temporarily (without shorting anything else - you've had enough of that!) before turning on and allow a few seconds for the capacitor to charge., then measure the volts acros the 100k. If you've managed to create a simple sinewave generator, that is, to provide the input signal.
The only way I would consider re-using a device is if it has passed all checks compared with a new component. That usually takes time and by the end of the exercise, finding it doesn't is a worse waste of time than just plugging in a new one. In your case, I would be curious to know if the pNP input transistor - or the NPN driver - have a gain loss. Does your DMM have a transistor gain test socket? You may not have the original info but you might see if the gain is lower than you would expect.
To check the gain of the individual channels you need to make sure that the input signals are the same. It may be that the transistor in front of the main amplifier (TR1?) is not behaving, but that may be due to the switches not working as there are quite a few to select options. With the mono/stereo on mono do you measure the same voltage at the inputs to the two channels (preferably with both fuses intact)? - that is on the collectors of the Tr1's. (You might want to make a temporary AC probe using (say) a 1uF capacitor and 100k resistor attached to the collectors to isolate the DC before measuring the signal on your DMM. Connect this temporarily (without shorting anything else - you've had enough of that!) before turning on and allow a few seconds for the capacitor to charge., then measure the volts acros the 100k. If you've managed to create a simple sinewave generator, that is, to provide the input signal.
I replaced both TR1s and TR3s with BC237Bs and also replaced the TIP41Cs on the left channel just because it was easy while I was at it, and they only cost pence.
Switched on and what a difference! Both channels much louder at less than 2 on the faders than the left channel had been at maximum. So much so that I swapped R20 back to 1K2 from 470R, resulting in very firm and dynamic sound at level 3 - far better and much louder than before, and still not getting too hot. I haven't had time to take more measurements as I have had to come away to help my parents with hip operation stuff, but I have kept the old transistors for whatever analysis I can carry out. I had to do a slight adjustment to bias on the left channel to avoid distortion, and I have the feeling that the right channel trimmer doesn't alter anything, but I will take measurements when I can get back to it and report back.
Once again, many thanks to all for the help, advice, thought and hard work - I'm sure guiding me through this has been like pushing a double decker bus with flat tyres across a ploughed field, but it is really very much appreciated!
Switched on and what a difference! Both channels much louder at less than 2 on the faders than the left channel had been at maximum. So much so that I swapped R20 back to 1K2 from 470R, resulting in very firm and dynamic sound at level 3 - far better and much louder than before, and still not getting too hot. I haven't had time to take more measurements as I have had to come away to help my parents with hip operation stuff, but I have kept the old transistors for whatever analysis I can carry out. I had to do a slight adjustment to bias on the left channel to avoid distortion, and I have the feeling that the right channel trimmer doesn't alter anything, but I will take measurements when I can get back to it and report back.
Once again, many thanks to all for the help, advice, thought and hard work - I'm sure guiding me through this has been like pushing a double decker bus with flat tyres across a ploughed field, but it is really very much appreciated!
Hope your parents OK.
And that your amp still works.
If you still see the quiescent current drifting I stand by my earlier advice to use a BD139 (properly insulated) mounted next to the output transistors for the bias stabiliser. Evidently your transistors were blown, but that does not mean the circuit could not be improved.
There is the old adage "if it ain't broke don't fix it" ... but if the quiescent current moves too much, it needs a fix.
And that your amp still works.
If you still see the quiescent current drifting I stand by my earlier advice to use a BD139 (properly insulated) mounted next to the output transistors for the bias stabiliser. Evidently your transistors were blown, but that does not mean the circuit could not be improved.
There is the old adage "if it ain't broke don't fix it" ... but if the quiescent current moves too much, it needs a fix.
Thank you - just got my mum back from hospital after eight days, and all is well, although I'm still not sure when I will be back home again to attend to the amp. I will definitely re-install the BD139s back on the heat sink, or perhaps even directly on one of the output TIP41Cs? Reading through some old threads, it sounds like that might be a good solution, and easy enough to achieve. What do you think? I'm certainly not fixed on total authenticity, and if I could have improved the reliability or longevity of my original amp many years ago, I'm sure I would have done it (and I'm also sure that the original designer was working to a tight budget and couldn't necessarily implement the ideal - what engineers have to do after all is come up with the right solution at the right price in the right time frame) I'm looking forward to getting back to it with the luxury of a bit more time, and I can't emphasise enough how grateful I am for all the help. I hope to be able to report back with some measurements maybe next week...
I managed to get home briefly and brought the amp back with me, so I can persevere, albeit with limited tools and facilities.
I re-fitted the BD139s, attaching them to one of the output transistors
The difficulty with setting bias continues, however. If I set bias to be the same on both channels, the left channel is very distorted until I increase it.
This is the right channel (sounding good)
And here is the left channel (sounding good)
There is a slight rustling background noise on the left channel which doesn't vary with volume.
There is also some mains hum which goes if I touch the chassis, but I assume that this is some grounding issue - I'm not too concerned, as the only place I can plug it in here at my parents' house is on a multi point extension lead which also supplies pc, router, monitor etc. I can investigate later when I know I have a cleaner mains supply.
Any ideas about the right/left discrepancy? I didn't replace the left channel output TIP41Cs after I blew up the right channel, so maybe I should?(when I can get some - Cricklewood Electronics is a long way from here)
I re-fitted the BD139s, attaching them to one of the output transistors
The difficulty with setting bias continues, however. If I set bias to be the same on both channels, the left channel is very distorted until I increase it.
This is the right channel (sounding good)
And here is the left channel (sounding good)
There is a slight rustling background noise on the left channel which doesn't vary with volume.
There is also some mains hum which goes if I touch the chassis, but I assume that this is some grounding issue - I'm not too concerned, as the only place I can plug it in here at my parents' house is on a multi point extension lead which also supplies pc, router, monitor etc. I can investigate later when I know I have a cleaner mains supply.
Any ideas about the right/left discrepancy? I didn't replace the left channel output TIP41Cs after I blew up the right channel, so maybe I should?(when I can get some - Cricklewood Electronics is a long way from here)
Quasi Comp ?
you wouldn't put drivers on the heatsink
bias million miles apart.
DC offset is way out.
Something still wrong with the circuit.
over bias just compensates for one side doing nothing.
likely why the channel blew. all current going one device.
you wouldn't put drivers on the heatsink
bias million miles apart.
DC offset is way out.
Something still wrong with the circuit.
over bias just compensates for one side doing nothing.
likely why the channel blew. all current going one device.
The BD139s added onto the heat sink are TR4 on the diagram - Vbe multiplier.
I have taken some measurements - on the above diagram at TR1 the emitter is marked 4.1 V and the collector 9.5 V
I have 2.47 V left channel and 2.36 V right channel at emitter and 8.55 V left channel and 8.91 V right channel at collector.
Base of TR2 is marked 12.5 V I have 13.04 V left channel and 13.3 V right channel.
I'm afraid I don't understand the significance of these....
I have taken some measurements - on the above diagram at TR1 the emitter is marked 4.1 V and the collector 9.5 V
I have 2.47 V left channel and 2.36 V right channel at emitter and 8.55 V left channel and 8.91 V right channel at collector.
Base of TR2 is marked 12.5 V I have 13.04 V left channel and 13.3 V right channel.
I'm afraid I don't understand the significance of these....
I'm a little concerned that you didn't start at the beginning again before moving your BD139 to the heatsink.
So,
The voltages on Tr1 are a concern to check because the diagram voltages usually indicate typical voltages. While the operating voltages are not outrageous and may depend on the actual supply voltage, the differences may indicate trouble like oscillation. That comes back to whether you have an earth loop in your set-up, or improperly earthed or screened cables, perhaps.
Can you repost the whole diagram as if there is an input to Tr1 signal it might be possible to connect an earthed input to see if your noise goes away, to skip the preamplifier section.
So,
- did you have this instability before the move? Reading your earlier posts, you did, but I thought that had been cured when you put everything back together as it was in the beginning.
- if you had the noise you reported, then moving the BD139 is not the solution to that. Moving the BD139 was to stabilise the quiescent current. The question I asked was "do you see the queiscent current moving or unstable when everything was back as it originally was?" If the answer to that was "yes", then the BD139 move might help.
- if you had the noise before, and after replacing things into the original place, then we still need to understand where that is coming from.
The voltages on Tr1 are a concern to check because the diagram voltages usually indicate typical voltages. While the operating voltages are not outrageous and may depend on the actual supply voltage, the differences may indicate trouble like oscillation. That comes back to whether you have an earth loop in your set-up, or improperly earthed or screened cables, perhaps.
Can you repost the whole diagram as if there is an input to Tr1 signal it might be possible to connect an earthed input to see if your noise goes away, to skip the preamplifier section.
Sorry for the delay - my mum is making progress, but there is a lot to do, so I don't have much time at the moment.
I hope this is relatively clear - I don't have the paper copy with me, so I can't get another scan for a while.
Before I swapped back TR4 to BD139 again the bias symptom was the same - when set at the same level for both channels the right channel sounded good, but the left channel was distorted. I adjusted the left channel bias by ear until the sound was clear, and played music for several hours. It remained stable and didn't get hot, although I wasn't playing it at a very high volume. This has stayed the same after changing to the BD139 mounted on the heat sink.
I had also changed all the transistors for new ones after the shorting incident, but not the capacitors - these had all been replaced before - and the resistors measured within tolerance. I now cannot remember if I replaced all the TIP41C output transistors or just the right channel - my feeling is that I did, since they cost pence and it doesn't take long, and I have kept all the components I have replaced in a separate container, but don't have them here with me to check. I have also checked the PCB tracks and not found any breaks or shorts.
I hope this is relatively clear - I don't have the paper copy with me, so I can't get another scan for a while.
Before I swapped back TR4 to BD139 again the bias symptom was the same - when set at the same level for both channels the right channel sounded good, but the left channel was distorted. I adjusted the left channel bias by ear until the sound was clear, and played music for several hours. It remained stable and didn't get hot, although I wasn't playing it at a very high volume. This has stayed the same after changing to the BD139 mounted on the heat sink.
I had also changed all the transistors for new ones after the shorting incident, but not the capacitors - these had all been replaced before - and the resistors measured within tolerance. I now cannot remember if I replaced all the TIP41C output transistors or just the right channel - my feeling is that I did, since they cost pence and it doesn't take long, and I have kept all the components I have replaced in a separate container, but don't have them here with me to check. I have also checked the PCB tracks and not found any breaks or shorts.
I'm not sure where you are up to with this either.
1/ What voltage do you measure between the two arrows (red). That voltage allows the bias current to be calculated. Measure on both channels.
And:
What voltage is needed between those points to get the left channel to sound clear?
2/ Now measuring from ground, what voltage do you see on the green arrow point. This should be approximately one half of the total supply voltage.
1/ What voltage do you measure between the two arrows (red). That voltage allows the bias current to be calculated. Measure on both channels.
And:
What voltage is needed between those points to get the left channel to sound clear?
2/ Now measuring from ground, what voltage do you see on the green arrow point. This should be approximately one half of the total supply voltage.
Can I suggest that you measure the voltages across both of the 0.5 ohm (the parallelled 1 ohms) as a further check (i.e. between the red and green arrows)?
It may indicate if there is a significant imbalance in the two halves as they would be expected to be the same. They may be, but it may offer a clue if not.
It may indicate if there is a significant imbalance in the two halves as they would be expected to be the same. They may be, but it may offer a clue if not.
For what it is worth the simulation I just ran indicates 2.3V on the emitter of Tr1 and 8.5V on the collector. The voltages you reported are in line with that which makes the indicated values suspicious. (My simulation was with 24V supply on the preamp power line rather than a zener diode and with 40V supply rail rather than 39).
It does not look as though Tr1 is the problem at this point.
Previously, on this channel, the last episode suggested that Tr5 might have been shorted. So that is important to measure Mooly's green arrow.
It does not look as though Tr1 is the problem at this point.
Previously, on this channel, the last episode suggested that Tr5 might have been shorted. So that is important to measure Mooly's green arrow.
My board doesn't actually have two pairs of 1 ohm emitter resistors, but two single 1/2 ohms, but the readings are:
Right channel - 9.2 mV across both, 4.5 mV across each individually.
Left channel - 128.6 mV across both, 63.8 mV and 64.9 mV individually. This is the point at which the sound is no longer obviously distorted, but if the bass tone control is raised it distorts - this is not the case on the right channel.
Right channel centre to ground 18.25 V
Left channel centre to ground 18.02 V
This is roughly half the supply voltage.
The background rustling noise is constant on the left channel, and does not increase with raising the volume control.
Right channel - 9.2 mV across both, 4.5 mV across each individually.
Left channel - 128.6 mV across both, 63.8 mV and 64.9 mV individually. This is the point at which the sound is no longer obviously distorted, but if the bass tone control is raised it distorts - this is not the case on the right channel.
Right channel centre to ground 18.25 V
Left channel centre to ground 18.02 V
This is roughly half the supply voltage.
The background rustling noise is constant on the left channel, and does not increase with raising the volume control.
One of the notes added to the circuit diagram reads:
'If components are changed in the power amplifier the quiescent current shall be reset to 15 mA with Rv5. The current is measured through the fuse with the volume control at minimum & the amplifier cold'
'If components are changed in the power amplifier the quiescent current shall be reset to 15 mA with Rv5. The current is measured through the fuse with the volume control at minimum & the amplifier cold'
I see your meter leads have the common leads with too much exposed metal. If you can I would advise obtaining a short length of red and black sleeving or a stripping from a red/black wire (or even brown/blue these days as we've changed our mains colours) and push them over the probe tips to leave perhaps only 1 or 2mm of metal at the very tip? Less chance of shorting tracks you didn't want to. You need the thinnest sleeve which fits fairly tightly so it does not get in the way of probing.
Reason is that if you set the quiescent current on this problem channel to 8 mA (4-5mV) are you able to measure the voltages across the bias transistor, and the individual driver and output transistor bases?
If you can do this safely without blowing anything else up it might indicate if the transistors are working properly or not. At this point, if you cannot do that easily don't try- it may cause more damage. I'd rather you got the aid of someone with an oscilloscope and oscillator because that's what I'd do next. One simple measurement with a scope might reveal the problem.
Are you sure your TIP41's were new and working? The excess noise you report indicates a component is still malfunctioning, if the right channel is quiet. That's why I suggested making sure you fitted known good components when rebuilding. If your meter has a transistor gain measurement that might be a useful check.
That the bass control causes distortion may or may not mean much because it increases the drive to the amplifier. AS it seems to be struggling to get a decent signal out that just makes it worse. But does increasing the volume also have the same effect?
Reason is that if you set the quiescent current on this problem channel to 8 mA (4-5mV) are you able to measure the voltages across the bias transistor, and the individual driver and output transistor bases?
If you can do this safely without blowing anything else up it might indicate if the transistors are working properly or not. At this point, if you cannot do that easily don't try- it may cause more damage. I'd rather you got the aid of someone with an oscilloscope and oscillator because that's what I'd do next. One simple measurement with a scope might reveal the problem.
Are you sure your TIP41's were new and working? The excess noise you report indicates a component is still malfunctioning, if the right channel is quiet. That's why I suggested making sure you fitted known good components when rebuilding. If your meter has a transistor gain measurement that might be a useful check.
That the bass control causes distortion may or may not mean much because it increases the drive to the amplifier. AS it seems to be struggling to get a decent signal out that just makes it worse. But does increasing the volume also have the same effect?
Interesting...
It should have got hot (well warm anyway) with that level of current flow.
Are both 0.5 ohm resistors OK and not high in value?
good question, it would obviously limit the maximum output power, and may contribute to the distortion!
Remember when measuring low resistances to check the meter lead resistance by shorting them first, then subtracting the lead resistance from the measured value. If you can't get a stable low reading when checking, the contacts to the leads and/or battery (batteries) in you meter may need to be replaced. Usually we have to use four terminal measurements for an accurate low resistance reading. That is tricky to do in situ, but it could be done. The difference between 0.5 and say 5 ohms, say, would be easily discerned with the ordinary 2-terminal measurement.
(edited to change my opinion on whether the resistors would cause the distortion. They may, but I suspect some sort of imbalance between the transistors).
You could set the bias current to 8mA with the resistors as they are, measure the voltage (approx 4-5mV) across them as though they were correct then measure the current across the fuse terminals.
This is also a tricky measurement. You do not want to blow your meter but at the same time you may need to powerup and down. Connecting your meter across the fuse terminals when the other channel is powered will see a surge which you don't want. Best to try to connect the meter leads across the fuse then carefully lift one end of the fuse when things are stable. Reconnect the fuse after measuring before removing meter leads. Tricky if you don't have meter cables wtih clips on the end.
Remember when measuring low resistances to check the meter lead resistance by shorting them first, then subtracting the lead resistance from the measured value. If you can't get a stable low reading when checking, the contacts to the leads and/or battery (batteries) in you meter may need to be replaced. Usually we have to use four terminal measurements for an accurate low resistance reading. That is tricky to do in situ, but it could be done. The difference between 0.5 and say 5 ohms, say, would be easily discerned with the ordinary 2-terminal measurement.
(edited to change my opinion on whether the resistors would cause the distortion. They may, but I suspect some sort of imbalance between the transistors).
You could set the bias current to 8mA with the resistors as they are, measure the voltage (approx 4-5mV) across them as though they were correct then measure the current across the fuse terminals.
This is also a tricky measurement. You do not want to blow your meter but at the same time you may need to powerup and down. Connecting your meter across the fuse terminals when the other channel is powered will see a surge which you don't want. Best to try to connect the meter leads across the fuse then carefully lift one end of the fuse when things are stable. Reconnect the fuse after measuring before removing meter leads. Tricky if you don't have meter cables wtih clips on the end.
Last edited:
I've managed to get a few more measurements, but I may not be able to get back to it for a few days due to care duties, so if I go silent I'm not being rude, just busy elsewhere!
Emitter resistors all measure 0.5 ohm
Measurements at base of transistors is as follows:
TR2 L - 12.9 V R - 13.2 V
TR3 L - 0.6 V R - 0.6 V
TR4 L - 17.4V R - 18.26 V
TR5 L - 17.5 V R - 18 V
TR6 L - 19.5 V R - 19.5 V
TR7 L - 0.6 V R - 0.6 V
TR8 L - 18.8 V R - 19.4 V
many, many thanks!
Emitter resistors all measure 0.5 ohm
Measurements at base of transistors is as follows:
TR2 L - 12.9 V R - 13.2 V
TR3 L - 0.6 V R - 0.6 V
TR4 L - 17.4V R - 18.26 V
TR5 L - 17.5 V R - 18 V
TR6 L - 19.5 V R - 19.5 V
TR7 L - 0.6 V R - 0.6 V
TR8 L - 18.8 V R - 19.4 V
many, many thanks!
TR6 and TR8 voltages might show a problem.
The base of TR6 should be around 0.6 volts higher than the base of TR8 and your measurement shows just 0.1 volts difference.
First suspect would be either a solder blob between B and E of TR6 (measure on ohms to see if there is a short) or a faulty/incorrectly fitted TR6
The base of TR6 should be around 0.6 volts higher than the base of TR8 and your measurement shows just 0.1 volts difference.
First suspect would be either a solder blob between B and E of TR6 (measure on ohms to see if there is a short) or a faulty/incorrectly fitted TR6
Those incorrect voltages ... are they really on the right channel?
Left one looks OK by comparison!
Left one looks OK by comparison!