Best sounding headphone amp from Wavebourn to Kubeek :)

[kubeek]

Yes, that seems to be quite practical.

I also thought about using power supply like -4V and +8V, and connecting outputs without capacitor, and using DC servo to maintain stable DC voltage.

I cannot decide between this and delayed relay in series with output, or usual power supply and relay with resistor in parallel with output.

No cap, but DC servo + complex power supply, or with cap, but simple power supply?
I just don´t know....

 

[Wavebourn]

No need at all, just increase capacitance of the filter capacitor in the 1'st base bias and try. To discharge it quckly on turn-off use a diode from it to +rail (cathode on + rail)

 

[kubeek]

hmm, with R7 being 3k9, I have only 40mA idle current, you wrote it should be like 200mA, that´s strange.
I get 200mA with 700ohms.

 

[Wavebourn]

hmm, with R7 being 3k9, I have only 40mA idle current, you wrote it should be like 200mA, that´s strange.
I get 200mA with 700ohms.

It depends on beta of BD135 on bottom (current source amp)
Also, a diode error corrector took some current. Place them together, for better temperature contact.

 

[kubeek]

it is set to 100, so it should be the stated 200mA: http://www.diyaudio.com/forums/showthread.php?postid=1056954#post1056954

 

[kubeek]

the diode takes 600uA, base current of the bottom transistor is 2.5mA, the diode can´t be responsible for such difference.

 

[Wavebourn]

Peak current should be twice of an iddle current. If base current is 2.5 mA, collector current is 100 mA, it means beta is 40.

 

[kubeek]

no, I was talkin about idle currents. I always try to make it work in DC mode, before I get to measuring with signal.

 

[kubeek]

Sorry, here are the right numbers: Ib=3.14mA, Ie=207mA, Ic=204mA, Hfe=100.
R7 is 600ohm to get the 200mA idle current.
With 3k9, Ib=243uA, Ie=19.4mA, Ic=19.2mA.

I just wanted to say that 3k9 can´t create 200mA idle current in the output.

 

[Wavebourn]

Sorry, here are the right numbers: Ib=3.14mA, Ie=207mA, Ic=204mA, Hfe=100.
R7 is 600ohm to get the 200mA idle current.
With 3k9, Ib=243uA, Ie=19.4mA, Ic=19.2mA.

I just wanted to say that 3k9 can´t create 200mA idle current in the output.

Sorry, my mistake in calculations on a napkin.

I used such output stage in both 300W/4 Ohm and line 600 Ohm driver, so 3.9K could come on mind from the last one.

 

[kubeek]

Ok then, I wasn´t sure whose nubers are wrong

I just found out that the distortion gets higher when the load resistance is higher, is that normal? (32ohm->0.5%@0.8V output, 320ohm->0.8%@0.8V output, both with 170mA idle current)

 

[Wavebourn]

Ok then, I wasn´t sure whose nubers are wrong

I just found out that the distortion gets higher when the load resistance is higher, is that normal? (32ohm->0.5%@0.8V output, 320ohm->0.8%@0.8V output, both with 170mA idle current)

Yes, it is normal. Both BJTs and FETs are nonlinear beasts, so it is hard to balance them perfectly when they are working on a real load.

What is most significant,

1. Distortions raise up with power
2. Order of distortions is low.

It makes the amp to fool human perception, as if the amp does not exist, just an amplifying straight wire between signal source and speaker.

 

[kubeek]

Yes, the distortion should raise with power, but here it gets lower with increasing power.
With the same voltage applied to 32ohm and to 320 ohm, the distortion with 32 should be bigger than with 320, but it isn't.
Please read the numbers again, the distortion on 320 is nearly two times higher than on 32.

 

[Wavebourn]

Yes, the distortion should raise with power, but here it gets lower with increasing power.
With the same voltage applied to 32ohm and to 320 ohm, the distortion with 32 should be bigger than with 320, but it isn't.
Please read the numbers again, the distortion on 320 is nearly two times higher than on 32.

When I say "With power", I mean the same load, louder sound.

Yes, the amp is intended to give the best results when driving the real load, so shouting to the air it indeed overcompensate own errors.

 

[Wavebourn]

Oops... The post went into the wrong topic...

 

[kubeek]

Allright.
I am going to sleep, the time here is 0.30.

Now I understand the design, so tomorrow I will start creating a board for it. I am really looking forward to hear the amp

 

[Wavebourn]

Allright.
I am going to sleep, the time here is 0.30.

Now I understand the design, so tomorrow I will start creating a board for it. I am really looking forward to hear the amp

Good nignt!

You will be really pleased, as people and singers on my concerts are.

They say they never heard such sound. But I know there is no magic, just knowlege how to fool imaginations using nonlinear devices.

 

[kubeek]

What can local PCB gurus say about this board?

Is eagle a common design program, or should I next time use some other formats (like jpg) for presenting on web?

 

[lineup]

What can local PCB gurus say about this board?

Is eagle a common design program, or should I next time use some other formats (like jpg) for presenting on web?

hello, kubeek

can you post the actual circuit you are using now
that goes along with PCB layout
don't forget to add Power Supply voltage used
.. maybe some details are different from Wavebourn's original quick drawing

I have to say this Wavebourn guy
has got some good and interesting audio ideas

Thanks
lineup

 

[kubeek]

the acutal circuit is actually included in the zip with the board, but I will post them both as pics.