Hi
I have a Quad 303 amplifier, the right channel first went silent, then started to roar. I switched it off immediately. I ran it for short periods with nothing connected to it to measure voltages etc. Not one transistor reads the same as the working chan. Point A reads near the supply voltage with no control from the relevant pot.
Reading on the internet, it tells you to short out the base of tr101 then tr 102 to test the circuits but I am a bit reluctant to do so in case I damage anything with the voltage being so far off. Not sure how this circuit works in respect of balancing Point A to half the supply rail.
On You Tube the guy called Old git used equvalent (or better) transistors, so I replaced all (recomended) Transistors apart from Tr 105 and Tr106 on the board. Have checked around for S/Cs but didn't find any.
Not sure on how to proceed apart from keeping replacing components. I have no workshop facilities and only a multimeter, soldering iron and some odd componemts.
I have a Quad 303 amplifier, the right channel first went silent, then started to roar. I switched it off immediately. I ran it for short periods with nothing connected to it to measure voltages etc. Not one transistor reads the same as the working chan. Point A reads near the supply voltage with no control from the relevant pot.
Reading on the internet, it tells you to short out the base of tr101 then tr 102 to test the circuits but I am a bit reluctant to do so in case I damage anything with the voltage being so far off. Not sure how this circuit works in respect of balancing Point A to half the supply rail.
On You Tube the guy called Old git used equvalent (or better) transistors, so I replaced all (recomended) Transistors apart from Tr 105 and Tr106 on the board. Have checked around for S/Cs but didn't find any.
Not sure on how to proceed apart from keeping replacing components. I have no workshop facilities and only a multimeter, soldering iron and some odd componemts.
Last edited by a moderator:
This should be repairable with only a DVM and possibly an analog VOM with a 20 VAC and/or 2 VAC scale. The latter required only if the DC voltages become right and the music doesn't flow through. Some people prefer scopes, but I find them unreliabile after 30 years and the probes are $50 if you step on one.
I find in my 50 year old silicon amps, transistors rarely fail. Sometimes the output transistors TR1L TR2L due to inadequate heat sinks, but Quad didn't have the reputation of cheapo heat sinks. A short on the speaker terminals can take out one side output transistors, also. It is true bar bands short a lot of output transistors with the pulled out phone plug to the speaker. Home Hifi amps don't have the villain as an output so don't blow nearly so many output transistors as techs see that service guitar/bass amps and band PA amps. Also home users don't turn the volume knob up to 11 with a crunch pedal all the time and melt the speaker coil, causing shorts.
The schematic I have doesn't show a point "A". I suppose this might be the junction of R124-125 which should at idle be about 1/2 of 67 volts. Does terminal 1 show 67 volts both sides? Terminal 5 33 v?
Incidently voltages over 25 from one hand to the other can stop your heart. Use an alligator clip lead on the negative probe of DVM to terminal 8 or 9 (analog ground) and only one hand. People that use a pamona grabber for the red probe blow up less parts than those using a meter probe with 1" exposed metal. Some DVM won't allow a banana plug of pamona grabber, IE the klein DVM at the Home Depot.
Wear safety glasses in case of parts explosion; also solder can splash in the eyes, especially desoldering. Wear no jewelry on hands wrists or neck. 1 v at high amps through a ring can burn your flesh to charcoal.
People usually debug small amps with a 60 w incandescent bulb in series with the AC cord input. If lamp stays lit after power up, usually something is drawing too much current. You are not blowing fuses, so probably not a problem with the output transistors and maybe the drivers are okay, too.
Basically, taking DC measurements with the amp at idle are much more useful than blindly swapping parts based on internet lore. In 50 year old products I find the rubber sealed capacitors (electrolytic) more suspect than any transistors. I put 122 of them in my Hammond h182 organ and it sounds really good now. Sounded like a wet kazoo when I bought it.
I find in my 50 year old silicon amps, transistors rarely fail. Sometimes the output transistors TR1L TR2L due to inadequate heat sinks, but Quad didn't have the reputation of cheapo heat sinks. A short on the speaker terminals can take out one side output transistors, also. It is true bar bands short a lot of output transistors with the pulled out phone plug to the speaker. Home Hifi amps don't have the villain as an output so don't blow nearly so many output transistors as techs see that service guitar/bass amps and band PA amps. Also home users don't turn the volume knob up to 11 with a crunch pedal all the time and melt the speaker coil, causing shorts.
The schematic I have doesn't show a point "A". I suppose this might be the junction of R124-125 which should at idle be about 1/2 of 67 volts. Does terminal 1 show 67 volts both sides? Terminal 5 33 v?
Incidently voltages over 25 from one hand to the other can stop your heart. Use an alligator clip lead on the negative probe of DVM to terminal 8 or 9 (analog ground) and only one hand. People that use a pamona grabber for the red probe blow up less parts than those using a meter probe with 1" exposed metal. Some DVM won't allow a banana plug of pamona grabber, IE the klein DVM at the Home Depot.
Wear safety glasses in case of parts explosion; also solder can splash in the eyes, especially desoldering. Wear no jewelry on hands wrists or neck. 1 v at high amps through a ring can burn your flesh to charcoal.
People usually debug small amps with a 60 w incandescent bulb in series with the AC cord input. If lamp stays lit after power up, usually something is drawing too much current. You are not blowing fuses, so probably not a problem with the output transistors and maybe the drivers are okay, too.
Basically, taking DC measurements with the amp at idle are much more useful than blindly swapping parts based on internet lore. In 50 year old products I find the rubber sealed capacitors (electrolytic) more suspect than any transistors. I put 122 of them in my Hammond h182 organ and it sounds really good now. Sounded like a wet kazoo when I bought it.
Last edited:
Thanks for your comments. I should have said in my earlier post that I carried out the Dada 303 revision a while back so none of the caps etc. are the original. Yes you are correct, Point A is as you say. I am OK with the Health and Safety aspects or electronic repair but thanks for your advice anyway. I have attached a table of the voltages found on test but don't really know what it is telling me. The small transistors were easily available and cheap so I thought it would be a fine time to update them anyway and possibly clear the fault at the same time.
In that case most likely fault is a bad solder joint, or a different pinout from the transistor you bought and the one that was installed. Will be working on my roof today, will give a look at voltage in 8 hours. Base-emitter of tr105 looks suspicitos, should be .6 or -.6 volts. BC546/556 come in several pinouts I believe, don't have personal experience with bc560.
A difference in voltage between the leg of the transistor and the pad of the PCB indicates a bad solder joint. Use a clip lead on the leg, not two hands.
A difference in voltage between the leg of the transistor and the pad of the PCB indicates a bad solder joint. Use a clip lead on the leg, not two hands.
Possibly the VAS transistor (the first U17219 I think) has gone OC. That would explain the high centre rail voltage, input transistor base voltage but does not explain the low voltage on the collector of the first BC109. It is dubious when replacing one transistor at a time that if the fault exists on the board the new device might be compromised or blow up too.
I would normally replace all transistors and while they are off the board, check the resistor values especially if any look rather overheated. Capacitors too, especially the electrolytics as they might have gone but the one on the PNP input seems to survive the 10V OK. A crude capacitor test is to see if they charge when measuring resistance (depending on your meter, if it's moving needle analogue the red lead might be negative but a DMM should have red positiev for resistance measurements) and then before it has a chance to discharge measure the voltage (you may need to breifly disconnect the meter to change ranges). Electrolytics should hold charge enough to see a value and other caps of maybe 100nf or so, and to see if charging current reduces which shows it isn't short.
Any voltages which show a reverse or zero bias on a transistor (like the PNP input) may indicate a dead transistor but in this case I suspect the device is just turned off by the what might be faulty VAS device.
I've suggested using a 2N5551/2N5401 for the VAS and first stage driver transistors as the BC546/BC556 are running near their limits, but others seem to use them OK. BD139-BD140 are OK as 38495/6 replacements and should be mounted vertically with a suitable heatsink, but the leads will need jiggling to relocate into the TO-5 pin-outs. You could also consider using 2N5680/2N5682 which Farnell offer (120V TO5) which are a little expensive but saves trouble of retrofitting a BD139/140. I don't immediately suspect these drivers though, but without more info/hands on it is hard to say.
There should not be an issue with the pin-outs of the BC546/BC556 they have the same triangular arrangement (although the base is kinked) as the TO-18 devices like the BC107 etc (also looking at the pins flat down, the triangle is E(left) B (kinked/top) C right. The confusion might arise because many TO-92 devices have different pin outs and pretty much all combinations have been covered. Unless of course pirate copies have been made by unscrupulous manufacturers.
I would normally replace all transistors and while they are off the board, check the resistor values especially if any look rather overheated. Capacitors too, especially the electrolytics as they might have gone but the one on the PNP input seems to survive the 10V OK. A crude capacitor test is to see if they charge when measuring resistance (depending on your meter, if it's moving needle analogue the red lead might be negative but a DMM should have red positiev for resistance measurements) and then before it has a chance to discharge measure the voltage (you may need to breifly disconnect the meter to change ranges). Electrolytics should hold charge enough to see a value and other caps of maybe 100nf or so, and to see if charging current reduces which shows it isn't short.
Any voltages which show a reverse or zero bias on a transistor (like the PNP input) may indicate a dead transistor but in this case I suspect the device is just turned off by the what might be faulty VAS device.
I've suggested using a 2N5551/2N5401 for the VAS and first stage driver transistors as the BC546/BC556 are running near their limits, but others seem to use them OK. BD139-BD140 are OK as 38495/6 replacements and should be mounted vertically with a suitable heatsink, but the leads will need jiggling to relocate into the TO-5 pin-outs. You could also consider using 2N5680/2N5682 which Farnell offer (120V TO5) which are a little expensive but saves trouble of retrofitting a BD139/140. I don't immediately suspect these drivers though, but without more info/hands on it is hard to say.
There should not be an issue with the pin-outs of the BC546/BC556 they have the same triangular arrangement (although the base is kinked) as the TO-18 devices like the BC107 etc (also looking at the pins flat down, the triangle is E(left) B (kinked/top) C right. The confusion might arise because many TO-92 devices have different pin outs and pretty much all combinations have been covered. Unless of course pirate copies have been made by unscrupulous manufacturers.
Thanks again. The table I sent you was from my initial tests with the original components in circuit (Middle column). The voltages on the left of the table were form the working channel and those on the right from the faulty channel.
There was very little difference in the readings when tr 100 -104 and tr107 were fitted. I had adjusted the 67v from the PS prior to that, so that probably varied the voltages slightly. I did check the data sheets for the transistors before fitting. Fitting the new transistors probably rules out any bad joints at least in the transistor area.
My thoughts are that either something in the top circuit is allowing the voltage at point A to remain high or something in the lower half is not conducting or allowing Point A to to go to half the supply rail. I am not sure what can be disconnected without causing any damage to verify the effect or not.
I hope the work on your roof went well. Don't feel under any pressure to respond to this quickly please take you time.
We have an expression here in Scotland " When god made time, he made plenty of i"t.
Gratefully yours,
Adam
PS. I bought this amp back in 1969/70 and this is the first fault. I also did the Dada revision a while back, so no old original caps etc.
There was very little difference in the readings when tr 100 -104 and tr107 were fitted. I had adjusted the 67v from the PS prior to that, so that probably varied the voltages slightly. I did check the data sheets for the transistors before fitting. Fitting the new transistors probably rules out any bad joints at least in the transistor area.
My thoughts are that either something in the top circuit is allowing the voltage at point A to remain high or something in the lower half is not conducting or allowing Point A to to go to half the supply rail. I am not sure what can be disconnected without causing any damage to verify the effect or not.
I hope the work on your roof went well. Don't feel under any pressure to respond to this quickly please take you time.
We have an expression here in Scotland " When god made time, he made plenty of i"t.
Gratefully yours,
Adam
PS. I bought this amp back in 1969/70 and this is the first fault. I also did the Dada revision a while back, so no old original caps etc.
There appears to be a disconnect between TR101 collector & TR102 base which are shown connected by a wire. They are different voltages.
Attached is an updated voltage table since I replaced Tr100, 101, 102, 103, 104 and 107. They were all fitted at the same time to avoid interaction if one was damaging another. The replacement types I used are listed on the table.
You haven't been changing parts willynilly since post 3 without analyzing the data? If you find something that should be impossible, you look for a bad solder joint, broken pcb trace, or a voltage on a component that is impossible before changing one thing, then remeasuring to see if that helped. You change transistors when the emitter collector voltage is zero, or improbably high with Vb-e .6 or -.6 indicating open circuit.
In new table TR104 base is higher in voltage than TR103 emitter. T104 is hard shut off. TR103b is supposed to be 1.8 v lower than TR103e. This is the double diode version, or is it the version with T107 on the heat sink? If early something is wrong in the stack TR103e, R120, MR105, MR103, MR104 MR106 R121, TR104e. Look at voltages down through there and see if there is anything wrong at each end of the PCB land.
In new table TR104 base is higher in voltage than TR103 emitter. T104 is hard shut off. TR103b is supposed to be 1.8 v lower than TR103e. This is the double diode version, or is it the version with T107 on the heat sink? If early something is wrong in the stack TR103e, R120, MR105, MR103, MR104 MR106 R121, TR104e. Look at voltages down through there and see if there is anything wrong at each end of the PCB land.
Last edited:
My boards are M12038 issue 10. Tr107 is not on a heat sink?.
I replaced the transistors before I contacted the Forum, I was going to upgrade them anyway as they are the best part of 50 years old, and was hoping the fault would be in one of them. To be honest I didn't know how the circuit worked. Im ok with the signal path, but not sure how the mid point is achieved. (Point A). Have checked again for broken tracks or bad joints but all appears to be OK. Thank for your input, will investigate further as you suggested.
I replaced the transistors before I contacted the Forum, I was going to upgrade them anyway as they are the best part of 50 years old, and was hoping the fault would be in one of them. To be honest I didn't know how the circuit worked. Im ok with the signal path, but not sure how the mid point is achieved. (Point A). Have checked again for broken tracks or bad joints but all appears to be OK. Thank for your input, will investigate further as you suggested.
Please do check the high value resistors, R112 that I mentioned earlier and also R130 and R104.
An open (or very high) R112 will give the voltages you have actually measured.
An open (or very high) R112 will give the voltages you have actually measured.
Spot on Mooly. R112 was open circuit. It is a physically bigger resistor than the others so I assume it is a higher wattage which I don't have. I have temporally piggy-backed an 82K and a 7 5K and it got 39.1K. Point A now adjustable and sets to 33.5 v easily. This resistor affects nearly every transistor on the board. I think that's why I was chasing my tail, that and my lack of knowledge of the operation of the circuit.
Thank you so much for your help and all the others who contributed with advice and suggestions. I am very very grateful.
Adam
Thank you so much for your help and all the others who contributed with advice and suggestions. I am very very grateful.
Adam
Excellent 👍 Resistors are funny things, they have voltage ratings as well as power ratings although in practice you are well within the usual maximum ratings for these type of parts.
If you look at the supply voltage of around 70 volts then worst case dissipation a 39k can see is (70*70)/39000 which is 0.125 watt. Yet they fail... its a surprisingly common scenario in older gear. High value resistor, constant high (ish) voltage and they are prone to fail.
It would be well worth checking the other channels value as well. It is not critical and 39.1k is absolutely fine. It would work and bias up correctly with considerably higher (and lower) values.
If you look at the supply voltage of around 70 volts then worst case dissipation a 39k can see is (70*70)/39000 which is 0.125 watt. Yet they fail... its a surprisingly common scenario in older gear. High value resistor, constant high (ish) voltage and they are prone to fail.
It would be well worth checking the other channels value as well. It is not critical and 39.1k is absolutely fine. It would work and bias up correctly with considerably higher (and lower) values.
Yes, I will be changing that resistor in the other channel as well. Looking at the images that I downloaded for the board, the R112s are all the same physical size as the others in that part of the circuit. Maybe they were out of that value on the day they built it, and used a higher wattage.
I also plan to replace the small transistors on the left channel to match the ones I replaced on the right channel. Like R112 they have all been working for fifty years.
I have the Quad 33/303/FM3 system and have acquired the same system for my son from a friend, so I know its history. I did the Dada mods for the 33 and 303 on both systems. I have had 5 faults in total between the two 33s and the two FM3s , but this is the first failure for the 303s. That's some going in fifty years.
Thanks again for your help.
Adam
I also plan to replace the small transistors on the left channel to match the ones I replaced on the right channel. Like R112 they have all been working for fifty years.
I have the Quad 33/303/FM3 system and have acquired the same system for my son from a friend, so I know its history. I did the Dada mods for the 33 and 303 on both systems. I have had 5 faults in total between the two 33s and the two FM3s , but this is the first failure for the 303s. That's some going in fifty years.
Thanks again for your help.
Adam
I've always liked the Quad's from a circuitry point of view. Another common failure item if it has any are the 'ROE' electrolytics.
None of the above are on the boards now, I replaced them when I did the Dada revision.
I will be looking at the FM3s in a couple of months time. Both are working, but do you have you any recommendations of any known weaknesses/problems. Obviously I will be replacing any ageing electrolytics.
Adam
I will be looking at the FM3s in a couple of months time. Both are working, but do you have you any recommendations of any known weaknesses/problems. Obviously I will be replacing any ageing electrolytics.
Adam
Recomendations.. in what way ? If asking about the sound qualities of the Venerable 303s' ?
Well then ..lets say that they hype Far better than they sound.
Decent mediocre amps but are/were more about visuals than performance .. by most yardsticks.
😉 Let sleeping dogs lie.. imo
Well then ..lets say that they hype Far better than they sound.
Decent mediocre amps but are/were more about visuals than performance .. by most yardsticks.
😉 Let sleeping dogs lie.. imo
I've never worked on any and so would begin by looking at what's fitted, what brand/s they are and what gets hot and what is near to what gets hot. Also give the board a good visual inspection for dries. If it uses a Supercap or NiCad cell/s for any kind of system memory retention then I would probably look at replacing those.