My suggestion is to find a
Germanium transistor of lets say about 30W
it will be probably PNP type,
Set the single ended topology with transfered load to collector
of 600-1000 ohms, start with 5H primary inductance,
riasing the Ub from 0 and monitoring the temperature...
The TR must be mounted on the heat sink...
When it become just below the warm stop riasing curent throught...
I think that we can have 5-7W in class A single ended Germanium
And most likely sounds much better then any transistor amp...
Germanium transistor of lets say about 30W
it will be probably PNP type,
Set the single ended topology with transfered load to collector
of 600-1000 ohms, start with 5H primary inductance,
riasing the Ub from 0 and monitoring the temperature...
The TR must be mounted on the heat sink...
When it become just below the warm stop riasing curent throught...
I think that we can have 5-7W in class A single ended Germanium
And most likely sounds much better then any transistor amp...
Since no one answered my question, I have ordered germanium transistor replacements by NTE from Mouser. NTE part numbers are: NTE 155 and NTE 131.
Mouser part numbers are: 526-NTE155 and 526-NTE131.
NTE155 ( replacement for AD161 ) $ 3.60
NTE131 ( -----------//------ AD162 ) $ 3.11
Those are the best prices which I could find in US.
Mouser part numbers are: 526-NTE155 and 526-NTE131.
NTE155 ( replacement for AD161 ) $ 3.60
NTE131 ( -----------//------ AD162 ) $ 3.11
Those are the best prices which I could find in US.
The problem is with transistors in US. The only germanium transistors, which I was able to find ( not to mention e-bay ) are ones sell by NTE.
Why not from here?
http://www.donberg.ie/descript/a/ad162.htm
Why not from here?
http://www.donberg.ie/descript/a/ad162.htm
yes,I think so...
if using other germanium transistor or silicon transistor,maybe the sound is difference...
Mr.gerard told me that his transformer power is 120VA each,
But I want to using one 120VA AC15V+AC15V toric transformer for dul-channel,I am trying to make this amp's 3D as small as possible...
TB said:
if using other germanium transistor or silicon transistor,maybe the sound is difference...
Mr.gerard told me that his transformer power is 120VA each,
But I want to using one 120VA AC15V+AC15V toric transformer for dul-channel,I am trying to make this amp's 3D as small as possible...
I was happy to observe this unit with calm... watching details.
I was observing, fascinated, the amplifier... and observing the colors used, the love that Gerard paint into the letter he sent me.
I feel very honored Gerard, as I can understand those subtle details of man behavior, I have some studies of Psychology that made me a little bit more
“wake up” related to those “hidden” messages.
It seems to me that you not only love this amplifier,but also that you trust entirely on it.
Reason why I was watching it deeply.
Using the higher value resistor, the current will be around 23 miliamperes crossing this bias resistor, to produce the needed 400 milivolts, that will bias the upper and lower transistor…each one of them will have around 200 milivolts from base to emitter, and the voltage at the output line will be half the supply voltage…if 25 volts, the voltage there may be something around 12.5 volts…the upper transistor base will have that output line plus 200 milivolts in its base….so…. 12.7 volts…..and the lower one 12.3 volts.
I cannot remember exactly, but those germanium transistors normally had low gain….something around 30….. but I have not sure about that, as more than 35 years passed since I used them last time.
If we accept that the gain could be 30…..with 23 miliamperes in the base to emitter circuit, the collector to emitter current will, probably, result in that base to emitter current multiplied by the gain…this could result around 700
miliamps crossing the output transistors…this is class A for sure!
The other option, reducing the main bias resistor, the current could be 350 miliamps….. well guys, observe that those number are not exact, as I do not know the exact bias voltage (from 150 to 240 milivolts), also I do not know the main supply voltage (can be from 20 to 25) and I do not know the exact transistor gain…so this is only some “parrot reflection”…talking while watching the circuit…. A real bla! Bla! Bla!.
That output current will produce a voltage over the 0.47 ohms resistors, and
The voltage can be 330 milivolts in the highest current I imagined…this will make the base increase it voltage related the output line…the center amplifier horizontal line….so…. the more real voltage for base may be 13.2 in the upper transistor and 12.6 in the lower transistor base…those may be more realistic numbers….dissipation in each one of the transistors could be around 8 watts to each one of them in the worst case…but I believe that in the real world it may be less than that, even for the worst case.,
Pirinnees are cold, they are having winter…temperature is very low now…but even this way, using the reduced bias resistor…the one with 8 ohms….power dissipation may reach 4 watts to each one, and it is slightly more than the case can dissipate by itself, without the heatsink use.
I suggest you to use heatsinks…..at least 5 by 5 centimeters each output units, as those guys turns noisy when overheat, and they are happy with “avalanche” temperature effect…more heat more current…and having more current the internal resistance is reduced, and this will produce more current, that will produce more heat…and this way going and going till burn the junction!
Yes, the numbers I am thinking can be twice the real numbers…the heat for instance can be a half of that, this depends from the transistor’s gain, voltages and currents.
Those circuits are not too much efficient, because class A, by itself is not efficient…. Normally you have 10 percent of the consumption of power transformed in sound…and 90 percent appear as heat!….but this one, being simetrical, may produce twice my expectations…. So, let’s imagine 20 percent efficiency…. 16 watts of consumption can produce happy and delicious 3.2 watts of pure sound…and do not think this is not a good power, as we use that power almost all the time, even when our amplifiers produce 100 watts, we adjust to 1 watt normally, so…the amplifier will have condition to show dynamics, undistorted audio power to audio peaks…nice!
The use of 4400uf electrolitic condenser in the output will block some deep bass, will reduce the deep bass volume slightly…so…if you can increase it…make a test with 10000uf and use a 100n capacitor in parallel with it, as those condensers…the big ones, have a lot of inductance, because the internal construction is alike a coil…and that inductance is a big problem creator…. The 100n will let the high frequency cross the damned electrolitic without return to circuit creating mess….treble will result clearer too.
The wonderfull bass is guaranteed by the 1000 uF electrolitic condenser that is in series with 33 ohms, near the BC107 emitter (BC547 is the modern unit for that application)…this one is Silicon, high gain and low noise unit…BC559 is the perfect one there, if you did not find the BC107.
The AC153 will be the voltage amplifier….this unit gain is very important, also the frequency response is interesting, it will produce the voltage and current needed to output…dissipation of heat there can reach 300 miliwatts, and this is too much heat to avoid heatsink…destruction is guaranteed in hot weather alike a have (30 degrees environment now, at night!)…It’s collector voltage will be 13,2 volts….the emitter will present your positive supply voltage and the base will present the supply voltage less 300 milivolts, approximately 24.7 in stand by mode, with the supply with 25 volts.
The supply must have full wave rectification, diode bridge and 10000uf, because of the high stand by current of class A…maybe more than that…listen close the speaker…if you listen some mains noise…increase the filter till the mains noise disappear.
The BC107 emitter will have 12.5 volts, the base will have 12.8 and the collector will have 24,7….current there will be very small…something around 2 miliamperes, and it will work very cold…very nice.
The second one, have perfect substitutes….those AC188 or AC128 and even 2N408 can hold this job….2 centimeters side, squared aluminium…rounded into a pencil…just a little to fit the transistor…small rouded corner, will produce perfect heatsink for it….the leads use to broke very easy…carefull when holding those transistors…avoid vibration on them…because the leads..do not let them flying…glue the case into something solid, not to shake if over some speaker, or your conection will broke, when years passed, result of long leads and vibrations…very fragile there…I broke hundreds in my life, in special those AC188 transistors.
I am afraid that the Stiroflex capacitor, in the feedback line, will be too big, killing softly some high end treble….i would try something smaller for experience only.
The input electrolitic is nice in size…but something non polarized…some black gates there may be very interesting.
Voltage produced in the output, because of the non simetrical supply, will not be much bigger than 6 volts in any case….i am talking undistorted, unclipped sound!…. this will represent, the maximum capability of 8 watts
RMS over 4 ohms and 4 watts RMS over 8 ohms….but I believe will be a little more than lovely, pure, clear and happy, 3.2 watts.
The output, clean voltage will be 3,7 volts…that means more than 10 volts peak to peak…the normal multimeters will read 3.7 or a little bit more…if you inject a constant 60 hertz tone in the amplifier input, as those meters only can measure that frequency with some precision (bad precision, error alike 10 percent).
Intersting is that the biggest majority of speaker I could have in my hands in those 45 years constructing amplifiers, are happy to receive that voltage, that will make it reaching the end of the diafragm movement…full compliance, with small speaker distortion, as this voltage is perfect for many speakers…and it will follow the wave without produce problems…this is something special in this amplifier…perfectly dimensioned to reality.
Search for soft speakers…. Not needed big magnets…. Acoustic suspension that have deep movement and large gap to avoid coil to touch the magnet…observe if goes deeply backwards and the same movement upwards….simetrical movement…to stop in the middle..to rest in the exact half way position, related completely out and completely in…use 3 volts batteries to make them move!….they may reach the end…so!…they will be perfect to a nice enclosure, that will send you to paradise….with some noise…some hissing, because Germanium..but very low , reproducing delicated sonics that will surprise you.
Be happy guys, Gerard knows what he is offering…. Wonderfull that he is sharing with us.
If something goes wrong…re-check, as the circuit is perfect…has no errors, if did not worked properly…hundred percent that was human error!
Regards,
Carlos
I was observing, fascinated, the amplifier... and observing the colors used, the love that Gerard paint into the letter he sent me.
I feel very honored Gerard, as I can understand those subtle details of man behavior, I have some studies of Psychology that made me a little bit more
“wake up” related to those “hidden” messages.
It seems to me that you not only love this amplifier,but also that you trust entirely on it.
Reason why I was watching it deeply.
Using the higher value resistor, the current will be around 23 miliamperes crossing this bias resistor, to produce the needed 400 milivolts, that will bias the upper and lower transistor…each one of them will have around 200 milivolts from base to emitter, and the voltage at the output line will be half the supply voltage…if 25 volts, the voltage there may be something around 12.5 volts…the upper transistor base will have that output line plus 200 milivolts in its base….so…. 12.7 volts…..and the lower one 12.3 volts.
I cannot remember exactly, but those germanium transistors normally had low gain….something around 30….. but I have not sure about that, as more than 35 years passed since I used them last time.
If we accept that the gain could be 30…..with 23 miliamperes in the base to emitter circuit, the collector to emitter current will, probably, result in that base to emitter current multiplied by the gain…this could result around 700
miliamps crossing the output transistors…this is class A for sure!
The other option, reducing the main bias resistor, the current could be 350 miliamps….. well guys, observe that those number are not exact, as I do not know the exact bias voltage (from 150 to 240 milivolts), also I do not know the main supply voltage (can be from 20 to 25) and I do not know the exact transistor gain…so this is only some “parrot reflection”…talking while watching the circuit…. A real bla! Bla! Bla!.
That output current will produce a voltage over the 0.47 ohms resistors, and
The voltage can be 330 milivolts in the highest current I imagined…this will make the base increase it voltage related the output line…the center amplifier horizontal line….so…. the more real voltage for base may be 13.2 in the upper transistor and 12.6 in the lower transistor base…those may be more realistic numbers….dissipation in each one of the transistors could be around 8 watts to each one of them in the worst case…but I believe that in the real world it may be less than that, even for the worst case.,
Pirinnees are cold, they are having winter…temperature is very low now…but even this way, using the reduced bias resistor…the one with 8 ohms….power dissipation may reach 4 watts to each one, and it is slightly more than the case can dissipate by itself, without the heatsink use.
I suggest you to use heatsinks…..at least 5 by 5 centimeters each output units, as those guys turns noisy when overheat, and they are happy with “avalanche” temperature effect…more heat more current…and having more current the internal resistance is reduced, and this will produce more current, that will produce more heat…and this way going and going till burn the junction!
Yes, the numbers I am thinking can be twice the real numbers…the heat for instance can be a half of that, this depends from the transistor’s gain, voltages and currents.
Those circuits are not too much efficient, because class A, by itself is not efficient…. Normally you have 10 percent of the consumption of power transformed in sound…and 90 percent appear as heat!….but this one, being simetrical, may produce twice my expectations…. So, let’s imagine 20 percent efficiency…. 16 watts of consumption can produce happy and delicious 3.2 watts of pure sound…and do not think this is not a good power, as we use that power almost all the time, even when our amplifiers produce 100 watts, we adjust to 1 watt normally, so…the amplifier will have condition to show dynamics, undistorted audio power to audio peaks…nice!
The use of 4400uf electrolitic condenser in the output will block some deep bass, will reduce the deep bass volume slightly…so…if you can increase it…make a test with 10000uf and use a 100n capacitor in parallel with it, as those condensers…the big ones, have a lot of inductance, because the internal construction is alike a coil…and that inductance is a big problem creator…. The 100n will let the high frequency cross the damned electrolitic without return to circuit creating mess….treble will result clearer too.
The wonderfull bass is guaranteed by the 1000 uF electrolitic condenser that is in series with 33 ohms, near the BC107 emitter (BC547 is the modern unit for that application)…this one is Silicon, high gain and low noise unit…BC559 is the perfect one there, if you did not find the BC107.
The AC153 will be the voltage amplifier….this unit gain is very important, also the frequency response is interesting, it will produce the voltage and current needed to output…dissipation of heat there can reach 300 miliwatts, and this is too much heat to avoid heatsink…destruction is guaranteed in hot weather alike a have (30 degrees environment now, at night!)…It’s collector voltage will be 13,2 volts….the emitter will present your positive supply voltage and the base will present the supply voltage less 300 milivolts, approximately 24.7 in stand by mode, with the supply with 25 volts.
The supply must have full wave rectification, diode bridge and 10000uf, because of the high stand by current of class A…maybe more than that…listen close the speaker…if you listen some mains noise…increase the filter till the mains noise disappear.
The BC107 emitter will have 12.5 volts, the base will have 12.8 and the collector will have 24,7….current there will be very small…something around 2 miliamperes, and it will work very cold…very nice.
The second one, have perfect substitutes….those AC188 or AC128 and even 2N408 can hold this job….2 centimeters side, squared aluminium…rounded into a pencil…just a little to fit the transistor…small rouded corner, will produce perfect heatsink for it….the leads use to broke very easy…carefull when holding those transistors…avoid vibration on them…because the leads..do not let them flying…glue the case into something solid, not to shake if over some speaker, or your conection will broke, when years passed, result of long leads and vibrations…very fragile there…I broke hundreds in my life, in special those AC188 transistors.
I am afraid that the Stiroflex capacitor, in the feedback line, will be too big, killing softly some high end treble….i would try something smaller for experience only.
The input electrolitic is nice in size…but something non polarized…some black gates there may be very interesting.
Voltage produced in the output, because of the non simetrical supply, will not be much bigger than 6 volts in any case….i am talking undistorted, unclipped sound!…. this will represent, the maximum capability of 8 watts
RMS over 4 ohms and 4 watts RMS over 8 ohms….but I believe will be a little more than lovely, pure, clear and happy, 3.2 watts.
The output, clean voltage will be 3,7 volts…that means more than 10 volts peak to peak…the normal multimeters will read 3.7 or a little bit more…if you inject a constant 60 hertz tone in the amplifier input, as those meters only can measure that frequency with some precision (bad precision, error alike 10 percent).
Intersting is that the biggest majority of speaker I could have in my hands in those 45 years constructing amplifiers, are happy to receive that voltage, that will make it reaching the end of the diafragm movement…full compliance, with small speaker distortion, as this voltage is perfect for many speakers…and it will follow the wave without produce problems…this is something special in this amplifier…perfectly dimensioned to reality.
Search for soft speakers…. Not needed big magnets…. Acoustic suspension that have deep movement and large gap to avoid coil to touch the magnet…observe if goes deeply backwards and the same movement upwards….simetrical movement…to stop in the middle..to rest in the exact half way position, related completely out and completely in…use 3 volts batteries to make them move!….they may reach the end…so!…they will be perfect to a nice enclosure, that will send you to paradise….with some noise…some hissing, because Germanium..but very low , reproducing delicated sonics that will surprise you.
Be happy guys, Gerard knows what he is offering…. Wonderfull that he is sharing with us.
If something goes wrong…re-check, as the circuit is perfect…has no errors, if did not worked properly…hundred percent that was human error!
Regards,
Carlos
No, i think i will not do it...because this one works fine when the transistors are
Germanium and matched units.
This is something almost impossible to find here...maybe some junk material, old radios maybe...but to find a pair, with the same gain...oh!...better to search for legs in snakes.
When PNP normally have more gain than NPN, specially in those days...oh!...very hard work, those parts may be difficult to see as some with fly, using baby doll and smoking a pipe.
regards,
Carlos
Germanium and matched units.
This is something almost impossible to find here...maybe some junk material, old radios maybe...but to find a pair, with the same gain...oh!...better to search for legs in snakes.
When PNP normally have more gain than NPN, specially in those days...oh!...very hard work, those parts may be difficult to see as some with fly, using baby doll and smoking a pipe.
regards,
Carlos
Hello,
NTE has also replacement for AC153K germanium transistor. Their part number is NTE158. Mouser has it under number: 526-NTE158 for $ 2.24.
I also have a question about part selections. Schematic notes mention selection of 2 Watts KIWAME resitors. This would be a rather expensive selection so, I ordered Xicon carbon comp. ones rated at 1/2 W. I hope it is enough ( I could not find 2 W carbon resistors). What about power rating of 0.47 Ohms resistors placed between output transistors. Suggested power rating is 3W-5W metal film resistors. Mouser catalog does not list any metal film resistors with that power rating for 0.47 Ohms. And even if they would carry one, price would be very steep. The only listed ones for a good price are wire wound ones for 0.5 Ohms ( note on the schematic says not to use wire wound ). Could anyone comment on that? The only reason which I can think about why not to use wire wound resistors would be their inductance. If metal film resistors would be recommended as the best can their rated power be lower than 3 W?
Thanks
NTE has also replacement for AC153K germanium transistor. Their part number is NTE158. Mouser has it under number: 526-NTE158 for $ 2.24.
I also have a question about part selections. Schematic notes mention selection of 2 Watts KIWAME resitors. This would be a rather expensive selection so, I ordered Xicon carbon comp. ones rated at 1/2 W. I hope it is enough ( I could not find 2 W carbon resistors). What about power rating of 0.47 Ohms resistors placed between output transistors. Suggested power rating is 3W-5W metal film resistors. Mouser catalog does not list any metal film resistors with that power rating for 0.47 Ohms. And even if they would carry one, price would be very steep. The only listed ones for a good price are wire wound ones for 0.5 Ohms ( note on the schematic says not to use wire wound ). Could anyone comment on that? The only reason which I can think about why not to use wire wound resistors would be their inductance. If metal film resistors would be recommended as the best can their rated power be lower than 3 W?
Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.