3 transistor classA 50w/25w with variable output impedance as SE triodes.

[Hayk]

Function description is here.

Can this 3 transistor classA amp 50w@4ohm with 1.3A bias still be called classA? - Page 1

Can this 3 transistor classA amp 50w@4ohm with 1.3A bias still be called classA? - Page 1

www.eevblog.com

This amp initially I designed to function with constant current, where the power CCS modulates itself to keep a constant current from the power supply.

NEW class A :Constant Current Push Pull amplifier with single SMPS and hybrid preamp

Floating Single Supply This type of power supply provides very high transient energy , the SMPS works in constant current mode , on condition that the two outputs are perfectly complimentary . I tried this technique with great success on sublimed JLH . Constant Current Push Pull To obtain a...

www.diyaudio.com

This is upside down version, negative supply, to be very expressive sounding amplifier, but, adding an phase inverting transformer at the input, it becomes the contrary to be soothing, background sounding amplifier.
Probably this the last amplifier I have to present, as the TPA3110 ABD mod and TPA3255 in BD mod replaced all my discreet designs and I don't have any other interesting exotic amp left in my collection.

 

[Hayk]

The PCB is 56x56mm the bottom layer is only ground.
If you can't buy the mosfets anywhere else then Aliexpress, buy only used ones, I bought 4 pairs for very cheap, and all were excellent. The new ones are not genuine with unknown characteristics. They don't need to be matched pairs.

 

[Hayk]

I built this amp. Very easy to adjust. No plop or thumb on both start up and shut down. The sound is the most dynamic version compared to the 2 precedents.
I will start on DIY forum a new thread with gerber files available.

 

[Hayk]

This amp is designed to be nonlinear, the NFB plays important role in tuning the sound quality. As it is intended to compete with single ended triode amps, the feedback resistor hence the gain need to be adjusted. Instead of 20db, I will try out diffrent values to find the best spot. The 220uf is wrong, it must be 1000u for better pronounced bass.

 

[Hayk]

I added 10k adjust with 3k resistor for the feedback. With 20db gain the amp at low power about 2w is has neutral sound as a standard amp. It wakes up above 26db with vigorous bass until 32db where it decreases because of low damping factor. This sweet spot depends upon the speaker, this why I will leave the user to adjust it at his convenience.
To repeat again that nonlinear output impedance distortion doesn't mean dirty sound. The positive going voltage decreases it and negative ones increase it. This makes more articulated voice particularly the low frequencies that are higher in amplitude. I will reverse the input phase by a transformer and see how it becomes a snoring amp good for harmonic type music as were the 2 precedent versions.
To adjust it, I used the following music.

 

[Hayk]

I installed a transformer and an inverting switch. At speech leve,l 89db speakers, the effect is negligible. It takes effect at higher powers.
The heatsink used is minimal. With 30°C ambient, it measures 71°C, it is 200x100x40mm. It can be replaced by uP fan cooled, cheaper and smaller radiator.
For the sound quality, being only 2 stages, it has higher resolution than JLH 1969 that I have one modified.
The TPA3110 in ABD modulation remains the highest resolution sounding amp for the same power.

 

[Hayk]

Gerber file ready to be uploaded to JLCPCB. Will cost less than $5 5 boards with econo shipment.
C2, C3,C5 can be 50v low ESR. C6 is 63v.
Bias=1.25A. Be cautious, the ground is positive.
SMPS of 48v 4A is sufficient for 2 channels.
Volume potentiometer needed of 10k.

 

[Hayk]

I was hearing something wrong in high frequencies. Silly me, I have used 10uf 16v input capacitor biased 25v. Once replaced, I hear now very smooth, outstanding high frequencies. Associated with punchy bass, now I am sure I have designed a jewel.