Hello,
I will be modernizing old Yugoslavian radio called Iskra Savica SN120.
In the process I will remove all electronics and I need a suitable amplifier.
It has oval (10x15cm paper membrane) full range speaker -full range with high sensitivity.
Enclosure is not completly sealed, it has backplate with many holes, like radios did have back in the day.
Those radios (when working correctly) have really nice warm sound that somehow fills the whole house, it also has strong mids.
This is the sound that I would like to keep.
Here is the schematic of original amplifier and I'm thinking of building the same one, except for better power supply:
What do you think of it? Is this class A?
Do you see any room for improvements?
Edit: I had just realised that AC187K and AC188K output transistors are germanium... can I substitite them for something newer? Also should I substitute other transistors with something newer?
I know that this amp goes nice with the speaker, would you recommend different amp anyway?
This is not the exact speaker, but it's similar enough.
Thank you
I will be modernizing old Yugoslavian radio called Iskra Savica SN120.
In the process I will remove all electronics and I need a suitable amplifier.
It has oval (10x15cm paper membrane) full range speaker -full range with high sensitivity.
Enclosure is not completly sealed, it has backplate with many holes, like radios did have back in the day.
Those radios (when working correctly) have really nice warm sound that somehow fills the whole house, it also has strong mids.
This is the sound that I would like to keep.
Here is the schematic of original amplifier and I'm thinking of building the same one, except for better power supply:
What do you think of it? Is this class A?
Do you see any room for improvements?
Edit: I had just realised that AC187K and AC188K output transistors are germanium... can I substitite them for something newer? Also should I substitute other transistors with something newer?
I know that this amp goes nice with the speaker, would you recommend different amp anyway?
This is not the exact speaker, but it's similar enough.
Thank you
Last edited:
You may be able to get a similar sound with a simple more conventional amplifier IF the tone control network in the feedback is connected in the same way. And if the amp is stable at any setting of the controls.
it’s kind of an interesting circuit - boot strapping both T205 and T206’s loads. And using a fully class B Ge pair driven by a class A emitter follower. All this is done to get the absolute maximum out of such a low supply rail. Just subbing silicon for the output pair will give horrible crossover distortion, but one could use a properly biased push pull pair and make it work. It will lose a little output swing, but just raising the supply to 20V would fix that, and a 3 or 4 amp TO-126 pair (BD43x or 44x) is as big as needed. With a pair of bias diodes between the bases, using a paralleled resistor to trim for 5 to 10 mA of bias current in the output pair. Small emitter resistors are probably a good idea too. The whole front end driving it could be left as is.
it’s kind of an interesting circuit - boot strapping both T205 and T206’s loads. And using a fully class B Ge pair driven by a class A emitter follower. All this is done to get the absolute maximum out of such a low supply rail. Just subbing silicon for the output pair will give horrible crossover distortion, but one could use a properly biased push pull pair and make it work. It will lose a little output swing, but just raising the supply to 20V would fix that, and a 3 or 4 amp TO-126 pair (BD43x or 44x) is as big as needed. With a pair of bias diodes between the bases, using a paralleled resistor to trim for 5 to 10 mA of bias current in the output pair. Small emitter resistors are probably a good idea too. The whole front end driving it could be left as is.
Keep the original germanium amplifier: it is perfectly matched to the set.
Otherwise, if you really want to switch to Si, opt for something like TDA2611 or TCA760
Otherwise, if you really want to switch to Si, opt for something like TDA2611 or TCA760
Assuming the Ge pair is still any good. If not one may be SOL without a switch to SI. I know I can’t get NPNs. Most of the time chip amps just don’t sound right in these applications. They usually aren’t happy about tone controls in the feedback and just don‘t sound right flat when the tone controls have been voiced into the overall design. Would that “vintage amplifier with a new twist” of yours be stable using tone controls in the feedback? If so it would be a GOOD starting point, better than a precarious diode bias circuit.
It's all a matter of personal choice, but for me I wouldn't want a chipamp in there. if there's concern about stability over the feedback tone control, you could use an active one. There lots of choice of low power amplifier circuits out there.
A replacement amp would require the same gain, about 68, and be inverter as that s mandatory
to have the tone control working correctly, otherwise with a non inverting amp it would just set the oscillation frequency.
Best it to let it as it is, an AD161/162 pair would fit even better than the original AC187K/188K but they are in TO66 package.
to have the tone control working correctly, otherwise with a non inverting amp it would just set the oscillation frequency.
Best it to let it as it is, an AD161/162 pair would fit even better than the original AC187K/188K but they are in TO66 package.
T204 is also inside the feedback, making the whole thing noninverting. Its capacitively coupled, but part of the overall “amplifier”.
I don’t like chipamps either, unless being used with modern speakers. IMO, back in the day they were always a step backwards unless you like tinny, shrill and forward presentation.
I don’t like chipamps either, unless being used with modern speakers. IMO, back in the day they were always a step backwards unless you like tinny, shrill and forward presentation.
Hy, thank you all for anwsers.
Based on what you wrote, I would like to stick with this amplifier design, just change germanium transistors.
I think that transistors itself are bad, my attept to fully fix this radio was unsuccesful, I replaced all capacitors and also some other components and there is a problem with sound. Imagine listening to speaker with bad connection to the coil, it goes like "brr brr brr" to the bass, now imagine there is another speaker connected that plays ok... this is what it sounds like. Signal that comes to the output transistors is clean, so thats the reason for my believe.
I would modify what needs to be modified... is there anyway one could change the schematic... I don˙t have enough knowledge on transistor amplifiers to do it myself. I am usually building chip amps with schematics ready to go. If it`s not too much work, I would appreciate it.
Thank you
Based on what you wrote, I would like to stick with this amplifier design, just change germanium transistors.
I think that transistors itself are bad, my attept to fully fix this radio was unsuccesful, I replaced all capacitors and also some other components and there is a problem with sound. Imagine listening to speaker with bad connection to the coil, it goes like "brr brr brr" to the bass, now imagine there is another speaker connected that plays ok... this is what it sounds like. Signal that comes to the output transistors is clean, so thats the reason for my believe.
I would modify what needs to be modified... is there anyway one could change the schematic... I don˙t have enough knowledge on transistor amplifiers to do it myself. I am usually building chip amps with schematics ready to go. If it`s not too much work, I would appreciate it.
Thank you
Hello,
I didn't make any progress yet, because I was bussy with the scale mechanism.
I want to replace the amplifier with new transistor amplifier.
I'm planning to build active preamp for tone controls.
From what I understand a part of current amplifier is class A and other part is class B... is this correct?
What do you think would be the closest match, class A, class B or class AB?
I intended to use class A, but I would prefer something that can be driven with a (if possible single rail) 12v supply and would not require as big of heatsink.
Thank you
I prefer transistorised amplifier for this project, not IC.
Speaker is low power, but high sensitivity and I want strong mid tones as it originaly had. "Big sound with low power".
I think that Ic amp of such low power wont do this... correct me if I'm wrong.
Will go higher with supply voltage if needed, it's just that everything else in my project runs on 12v, 8v & 5v... Im kinda waiting with preamp until I choose the amp, to see what transformer I will use.
Speaker is low power, but high sensitivity and I want strong mid tones as it originaly had. "Big sound with low power".
I think that Ic amp of such low power wont do this... correct me if I'm wrong.
Will go higher with supply voltage if needed, it's just that everything else in my project runs on 12v, 8v & 5v... Im kinda waiting with preamp until I choose the amp, to see what transformer I will use.
I probably misunderstand you, but this looks like the loudspeaker isn't properly centred. There are tricks to solve that.
This circuit is good. You should try to keep and restore it, the sound quality of Germanium is better than that of Silicon transistor. You should try to keep that circuit and you already have new AC187, AC 188 bulbs, it's so convenient. Now just need to check the working voltage points of the previous bulbs, replace the broken ones. Seeing the coupling capacitors, output capacitors, and power filter capacitors, it will probably sound good....
If I would keep the amplifier, I would need to make new pcb for it anyways.
I'm not too keen to do it, because this will be kitchen radio, working all the time (ge transistors....) and probably for years.
To replicate the sound I was thinking of recording original amp frequency response, then plugging a PC between tuner and amplifier & adjust the EQ until I get close and build a preamp with resposne curve as close as I can get.
I know that it wont sound exactly the same, but I believe that main charm with sound of many Yugoslavian radios were loud mid tones.
Our audio legacy was build to be afordable at the time, so there isn't and special "quality" about it.
I'm not too keen to do it, because this will be kitchen radio, working all the time (ge transistors....) and probably for years.
To replicate the sound I was thinking of recording original amp frequency response, then plugging a PC between tuner and amplifier & adjust the EQ until I get close and build a preamp with resposne curve as close as I can get.
I know that it wont sound exactly the same, but I believe that main charm with sound of many Yugoslavian radios were loud mid tones.
Our audio legacy was build to be afordable at the time, so there isn't and special "quality" about it.
Last edited:
Sorry, how could I forgot, its 4ohm 3w, power supply is not determined yet. So far I have/need 12v, 8v regulated & 5v regulated.
12v would be ideal, but I'm affraid that it is too little for transistorised amp. I was thinking maybee +-12v, if it's not too much to borrow 0.2a - 0.5a from positive rail for microcontroller & stepper motor.
I have some space to work with, but not too much. Also it is in wooden enclosure... I would like to avoid using 2 transformers, but if not else I could get away with using switcher for logic side and open more to audio power transformer specs.
12v would be ideal, but I'm affraid that it is too little for transistorised amp. I was thinking maybee +-12v, if it's not too much to borrow 0.2a - 0.5a from positive rail for microcontroller & stepper motor.
I have some space to work with, but not too much. Also it is in wooden enclosure... I would like to avoid using 2 transformers, but if not else I could get away with using switcher for logic side and open more to audio power transformer specs.
If you very efficiently used all 12 volts, you could get to 3W. Requires 10v p-p, so it’s all that could be expected. I’d start with 18 to 20 giving it room to sag 10-15%. And you’d still want an amplifier circuit that can get very close to the rails. +/-12V would give you room to get sloppy or press the “easy” button.
One does not especially need a +/-supply - you could start out with an 8-0-8 transformer and put the ground on the far negative side, and have both +10 and +20-ish volts (after rectifier and losses).
What kind of amplifier circuit to use depends mostly on one thing - how badly you want to keep the tone control network in the feedback loop. Not everything is going to be happy about doing that. I wouldn’t bet on chip amplifiers (even TDA2003, which is about as close a rail-to-rail chip amp as you can get) to do it without at least some grief. One could wrap it around a unity gain stable op amp (ie, LF351) and do it ahead of a “flat” amplifier. Even a TDA2003. If I wanted to keep the tone controls in the feedback loop of the power stage I’d rework the original topology. It’s not “bad”, but silicon is harder to do right than germanium here. Another method would be to put emitter followers in the feedback loop of a unity gain stable op amp, and then wrap the tone controls around it. The minus input of the op amp is the emitter of T204 - keep all ratios the same. But then that necessitates the higher voltage supply because the circuit is inefficient (does not swing very close to the rails). One can make a rail to rail amp with an LF351 in the front end and still be unity gain stable, but that design exercise is NOT for beginners. There is much potential for magic smoke.
One does not especially need a +/-supply - you could start out with an 8-0-8 transformer and put the ground on the far negative side, and have both +10 and +20-ish volts (after rectifier and losses).
What kind of amplifier circuit to use depends mostly on one thing - how badly you want to keep the tone control network in the feedback loop. Not everything is going to be happy about doing that. I wouldn’t bet on chip amplifiers (even TDA2003, which is about as close a rail-to-rail chip amp as you can get) to do it without at least some grief. One could wrap it around a unity gain stable op amp (ie, LF351) and do it ahead of a “flat” amplifier. Even a TDA2003. If I wanted to keep the tone controls in the feedback loop of the power stage I’d rework the original topology. It’s not “bad”, but silicon is harder to do right than germanium here. Another method would be to put emitter followers in the feedback loop of a unity gain stable op amp, and then wrap the tone controls around it. The minus input of the op amp is the emitter of T204 - keep all ratios the same. But then that necessitates the higher voltage supply because the circuit is inefficient (does not swing very close to the rails). One can make a rail to rail amp with an LF351 in the front end and still be unity gain stable, but that design exercise is NOT for beginners. There is much potential for magic smoke.