Class A, no feedback, 5W, minimum parts

I've been thinking... class A amplification is great, transistors are cheap, and usually a peak of only 3 watts is reached during normal listening.

If just a single transistor is biased to the middle of its linear range, could it be driven by a headphone amplifier for low power, high fidelity sound? A 2 stage amp with the first being a headphone amp?

I can't design the circuit for an amplifier (it's beyond my abilities), so I'm asking: is a headphone amplifier that can put out ~70mW (typical) be used to drive a single transistor which is biased into the middle of its linear range so that it can be used with no feedback? I know this won't result in the absolute best possible sound quality, but there'd be no crossover distortion, less harmonic distortion, and as long as the speaker connected doesn't have a very reactive low end, I think the frequency response should be close enough to flat

If it's possible, what would the simplest implementation of this look like? It'd be great if we could have inexpensive, easy, high fidelity, low powered, Class A amplification.
 
You can improve the quality of your questions by learning some very basic electronics. Enough to understand that quoting those 70mW is quite meaningless. You need to start backwards by defining what is the expected voltage at output.

There must be close to a hundred amps on this forum which do pretty much what you want. Do some research. Start from the basics, then read the Pass diy articles.
 
You can improve the quality of your questions by learning some very basic electronics. Enough to understand that quoting those 70mW is quite meaningless. You need to start backwards by defining what is the expected voltage at output.

There must be close to a hundred amps on this forum which do pretty much what you want. Do some research. Start from the basics, then read the First Watt articles.

70mW isn't meaningless, I gave it in the context of the output voltage of headphone amplifiers. And I acknowledged the help I was given already and will be reading
 
Zen or Son of Zen (balanced version)
Son of Zen | Pass DIY.

"The performance depends entirely on noncritical matching of gain Mosfets and resistors"

I have a Pioneer SX-950, which at this point is probably only good for parts unless I recapped the whole thing and repaired the power supply. The reason I stopped using it is one day when I turned it on, it blew the breaker. That was over 10 years ago. Anyway, it has 4 pairs of mosfets. One pair blew in 82 and they were replaced.

I know it says newer mosfets are preferred, but I have these ones and I assume they're matched pairs - would they be OK? By OK I mean would they sound good but not as good as it could?

I have two linear power supplies, 300 watts each (variablr 0-30v 10A). They're very fast.

The heatsink in the SX-950 is probably good for about 100 watts without a fan, so 50 per channel. They say 500 watts gives you 20 on the output, so I guess I'd get about two watts. Maybe I'll add a fan for 3 watts per channel.
 
You made a headphone amp version? How does it sound? Obviously good, but details!
No,
I made a PCB for the 4W-amp, see post #67 in that thread.
Then I said, if it´s as good as lineup´s headphone amplifier (which I already built) it´s going to be very good:
3 Transistor HP Amplifier with low dist

If you´re interested, I can give you the gerber files
but then you´ll be the beta tester ;-)
 
Last edited:
Yes, it is meaningless as it is not voltage. That is why i suggested learning some basics first. I know, not gonna happen.

I don't know if my headphone amp has the current for 70mW at 0.2V or 0.5V or 1V. Besides they all vary.

I picked 70mW because it's a typical rating for a headphone amp - they're all different. Just pretend I didn't give a power rating, ok? I'm asking if the first stage can be a typical headphone amp.

Change your attitude or go away
 
The Noir headphone amp in the diyAudio Store has two transistors per channel rather than one. Its output current vs. output voltage, just before clipping, is plotted below. If you happen to be interested in power rather than voltage or current: power equals voltage times current, you can read them off the plot.

The red squares on the plot show the required drive to get 110dB sound pressure level, for each of the individual headphone products in the comprehensive table on this thread over at head-fi dot org.

(The one red dot which lies above the blue line, is the Hifiman HE-500. It requires 2.5V RMS and 73mA RMS to play 110dB.)


_
 

Attachments

  • NOIR_output_voltage_and_current.png
    NOIR_output_voltage_and_current.png
    23.2 KB · Views: 173
The Noir headphone amp in the diyAudio Store has two transistors per channel rather than one. Its output current vs. output voltage, just before clipping, is plotted below. If you happen to be interested in power rather than voltage or current: power equals voltage times current.

The red squares on the plot show the required drive to get 110dB sound pressure level, for each of the individual headphone products in the comprehensive table on this thread over at head-fi dot org.

(The one red dot which lies above the blue line, is the Hifiman HE-500. It requires 2.5V RMS and 73mA RMS to play 110dB.)


_

I'm familiar with the basics, despite analog_sa's claims otherwise