24V Tube IRF610 MOSFET Hybrid HPA

[lineup]

This is a fun project.
The CCS for the MOSFET IRF610, at the same time will feed Heater of 6922 tube.

There are many more small HeadPhone Amplifiers (HPA) and interesting schematics
at the sijosae korean website

[DIY] MHHA v2.0 (Multi-Hybrid Headphone Amp)

headphoneamp.co.kr

[Dxvideo]

Hmm...
Interesting project. Looks like "my" old headamp project with one MOSFET and one op-amp..
However,
Do you think that tube works with just 24v supply?
And, he says, 1.05VAC at output while input was 0.1VAC, that means it has 10X voltage gain ! I dont understand from the tubes but I dont see any VAS stage in this amplifier.. How?

[lineup]

Tube will work at low voltage but will give some distortion.
Which can give the typical and good 'tube sound'.

Of course, raising the voltage to double, 50 Volt, will make one 6922, 6DJ8 or E88CC work better.
At 80-100 Volt these tubes are quite happy

150 mA x 24 V = 3.6 watt cooling
150 mA x 50 V = 7.5 watt cooling
Well, a nice little headphone amplifier at 50 Volt seems noot impossible.
Does it?

[Dxvideo]

As my calculations,

150mA bias is overkill for a headphone amplifier..
If you use 32R headphones (like me).. Then 70-80mW is fairly enough for you.. That means, you will source 50mA RMS and 70mA peak at full volume ... So if you take 20% more then just 85mA bias will be enough to have 80mW into 32R headphones..
So,
We can feed them with 50V and the total dissipation will be ~4.5W at output stage..
Do you think its ok?

[00940]

The amp uses the heater of the tube as load. Thus, you need 2x150ma to get the 300ma heater current required for the 6922, unless you add a new supply.

The same trick is used here: http://www.pmillett.com/starving.htm and also by the Melos headphones amp.

[Dxvideo]

As mentioned....
I dont understand from the tubes!
But this situation can be fixed with just a LM317 regulator added for tubes heater?

[lineup]

6922 needs 300 mA to get warm and glowing.
150 mA from each channel, will do this.

Dxvideo, you are correct about we may only need mybe 50 % of 150 mA.
For 32 Ohm headphones.

On the other hand IRF610 will not work worse at 150 mA SE, single end Class A.
In fact 150 mA Class A is a very good level for IRF610,
even if this MOSFET can work very well at 50 mA, too.

[Dxvideo]

I have tried various bias levels for IRF510 and hear no difference between 50 and 100mA bias... So I think the 610 will act as 510.
I think this circuit can be modified at least by the meaning of output stage;
- PSU will be 50v
- Bias will be 70-80mA (if you use higher impedance heaphones then just 50mA will be enough)
- CCS will be replaced with another IRF and zener...
- And the heater of the tube will feed with a seperate regulator tuned at 6v3..
I think if these mods are made, it works like a charm
But: I am still confused on voltage gain issue.. Is this tube have its own voltage gain? If not then is this circuit work as a buffer? Or what?

[lineup]

http://www.pmillett.com/starving.htm

This is at a very high output power level.
At the P. Millett website 'Starving Student Hybrid'

This shows the ear-friendly tube type of distortion we an expect.
Very Dominant is the 2nd harmonic.
At this extreme high output level THD is 5 %

In order to get any change in sound, we have to be able to hear a difference. Distortion is to change, add something to the sound.
And the distortion level we has to be well above 0.1-0.2 % to get a change in sound.
To try to get tube sound out of one tube amplifier with 0.01 % THD is impossible. We might as well use one Op-Amp with feedback.
If we want tube sound effect, we must give the amplifier chance to effect the input signal in some way.

MOSFET and Tubes are known to both have more of nice 2nd Harmonics distortion.

[lineup]

Originally posted by Dxvideo
I have tried various bias levels for IRF510 and hear no difference between 50 and 100mA bias... So I think the 610 will act as 510.
I think this circuit can be modified at least by the meaning of output stage;
- PSU will be 50v
- Bias will be 70-80mA (if you use higher impedance heaphones then just 50mA will be enough)
- CCS will be replaced with another IRF and zener...
- And the heater of the tube will feed with a seperate regulator tuned at 6v3..
I think if these mods are made, it works like a charm

But: I am still confused on voltage gain issue.. Is this tube have its own voltage gain? If not then is this circuit work as a buffer? Or what?

Yes, I think your idea will work very well.
And IRF510 and IRF610 is pretty much the same.

I also would want to know the actual voltage gain we could expect from tube.
I see the potentiometer at the input and the 2k2 pot from cathode(=emitter) to ground.
These are probably very necessary, to reduce gain/output voltage level.