If that output transformer has a secondary higher than 8 ohms, you would get more power with a higher impedance speaker. Start with a much higher impedance, and drop it till the maximum voltage swing starts to take a dive. That will tell you what the speaker impedance needs to be (aside from measuring turns ratio directly). Power will fall as the speaker impedance is raised too, but you will get full voltage capability with a high impedance load.
Here we get drop in circuits for CRT TVs, called 'kits', as sometimes the old ICs are not available, the tube is good.
In those the audio output is more powerful, the sound is much better with the same speakers.
Maybe you could change the amp to a more powerful one?
Provided the speakers can handle it of course.
In those the audio output is more powerful, the sound is much better with the same speakers.
Maybe you could change the amp to a more powerful one?
Provided the speakers can handle it of course.
Easily! Add a follower to the existing scheme: complementary pair or Diamond in class AB or A. A resistor instead of a transformer. In class A, you can use the output stage circuitry of JLH, Delapiano or Sugden, but more current from the power supply is required.
Any replacement IC amplifier, for example PAM8406
3 - 5 Watt Class-A Audio Amplifier - RED - Page80 (Transistor Q2 and surrounding R and C.)
An output transformer from a (very) old car radio and Darlington can be used.
Any replacement IC amplifier, for example PAM8406
3 - 5 Watt Class-A Audio Amplifier - RED - Page80 (Transistor Q2 and surrounding R and C.)
An output transformer from a (very) old car radio and Darlington can be used.
Last edited:
I don’t think I’d go modifying that particular turd by adding followers to an output stage eating 3 watts. If this were part of some stereo set you might actually want to improve, I’d just put in an efficient-running class AB amp using the existing power supply. Being able to supply six watts continuously to the existing amp, it should run as a higher power amp (around 6 watts a channel, sine wave on that voltage) with a normal music signal just fine without overload. Chip amp is the “easy” way. I’d go discrete myself, to eke out a couple more peak volts using bootstrap, and be able to optimize the output stage bias, rather than being stuck with the setting in the chip. A well stocked junk box could build it for “free”. The existing heat sinks are big enough.
The tube amp guys like really efficient speakers, because many of those amps don’t put out much power. Use a 6” bookshelf speaker on a 2A3 SET that cost a thousand bucks, and it won’t play much louder than a clock radio, either.
The tube amp guys like really efficient speakers, because many of those amps don’t put out much power. Use a 6” bookshelf speaker on a 2A3 SET that cost a thousand bucks, and it won’t play much louder than a clock radio, either.
Thanks.
I started studying Engineering in 1969, until 1974 and adding specialties along that time (Core being Industrial Engineering, adding modules from Electronics because my product was Guitar amps and then Business Administration because it grew exponentially so it was needed) and SPICE was available, in its early form.
Almost unusable for everyday tasks because our computer was an IBM1620 (proud of having "the same as Columbia University" thanks to a millionaire donor)
you could do only very limited, specific tasks because otherwise compuer couldn´t handle it and to boot output was a literal bedsheet of zig zag printer paper.
In real use, and when/if available, it was better to write needed snippets in FORTRAN and run them.
NO GRAPHS , only tons of numbers, not exactly user friendly, best was to have output as tables and pick results from there.
So the practical everyday way was to hand write needed equations and solve them using the MIGHTY Staedtler slide rule (I still have mine):
VERY fast , as in seconds to give you accurate results .
Accurate here meaning 1% or 2% error tops, well within Engineering or Technical needed precision; after all purchased parts come from a "preferred" series of values, besides manufacturing tolerance, so .....
Back to the OP problem: a humble TDA2030 fed from his nominal 20 V supply, needing a capacitor output of course, will fit in available space and give him incredibly good 4W RMS into 8 ohm, per channel.
Day and night compared to current results, I suggest he breadboards one , connects it to current supply and speakers and give it a try.
His Friend/Customer will LOVE results.
If unavailable or suspect, use guaranteed LM1875 from Mouser, same pinout and circuit.
Just try it: simple, cheap, available, do not polish the current turd.
As mentioned above, supply will feed it well if used clean, with Music program.
I started studying Engineering in 1969, until 1974 and adding specialties along that time (Core being Industrial Engineering, adding modules from Electronics because my product was Guitar amps and then Business Administration because it grew exponentially so it was needed) and SPICE was available, in its early form.
Almost unusable for everyday tasks because our computer was an IBM1620 (proud of having "the same as Columbia University" thanks to a millionaire donor)
you could do only very limited, specific tasks because otherwise compuer couldn´t handle it and to boot output was a literal bedsheet of zig zag printer paper.
In real use, and when/if available, it was better to write needed snippets in FORTRAN and run them.
NO GRAPHS , only tons of numbers, not exactly user friendly, best was to have output as tables and pick results from there.
So the practical everyday way was to hand write needed equations and solve them using the MIGHTY Staedtler slide rule (I still have mine):
VERY fast , as in seconds to give you accurate results .
Accurate here meaning 1% or 2% error tops, well within Engineering or Technical needed precision; after all purchased parts come from a "preferred" series of values, besides manufacturing tolerance, so .....
Back to the OP problem: a humble TDA2030 fed from his nominal 20 V supply, needing a capacitor output of course, will fit in available space and give him incredibly good 4W RMS into 8 ohm, per channel.
Day and night compared to current results, I suggest he breadboards one , connects it to current supply and speakers and give it a try.
His Friend/Customer will LOVE results.
If unavailable or suspect, use guaranteed LM1875 from Mouser, same pinout and circuit.
Just try it: simple, cheap, available, do not polish the current turd.
As mentioned above, supply will feed it well if used clean, with Music program.
Last edited:
JMFahey brings up a good point, and yes, in some cases I've modified some previous customer units with an appropriate IC conversion with excellent results.
I once built up an STK amp 20/20W board to replace the completely destroyed SA16 tube amp in an early 1960's Fisher console stereo.
The customer wanted to keep costs down, and accepted my proposal, since all three transformers, tubes, etc were ruined (likely due to neglect and powering up with oversized fuse)
As it turned out, the customer was thrilled at the "new punchy dynamic sound".
However, this amp's owner insisted on staying original as possible, end of story.
I once built up an STK amp 20/20W board to replace the completely destroyed SA16 tube amp in an early 1960's Fisher console stereo.
The customer wanted to keep costs down, and accepted my proposal, since all three transformers, tubes, etc were ruined (likely due to neglect and powering up with oversized fuse)
As it turned out, the customer was thrilled at the "new punchy dynamic sound".
However, this amp's owner insisted on staying original as possible, end of story.
The NTE175 is a 35W TO66 power transistor which would be required if the standing current of this circuit was higher yet the measurements suggest it is only of the order of 150 m.a.
While the transistor appears more than robust enough for the application it is possible the the 4R7 resistor connecting to the emitter has been damaged through exposure to heating over the lifetime of over 50 years and the actual value has "gone high". Strangely values this low can also serve as fuses under fault conditions and not show any visible signs of any issues. I have observed this possibility with an external pre-amp power supply when one or other of the rails had been short circuited by accident.
While the transistor appears more than robust enough for the application it is possible the the 4R7 resistor connecting to the emitter has been damaged through exposure to heating over the lifetime of over 50 years and the actual value has "gone high". Strangely values this low can also serve as fuses under fault conditions and not show any visible signs of any issues. I have observed this possibility with an external pre-amp power supply when one or other of the rails had been short circuited by accident.
Last edited: