Headphone amp

user 37518 said:

Whats the purpose of the resistor (R34) between the base and emmiters in the Studer schematic?

Click to expand...

I'm sure you probably know this already but you would need to add the 2200uF DC blocking cap like in your first schematic to use this Studer circuit. Otherwise you're fry your phones.

But if you're using a 9V battery you could try connecting the phones ground to the virtual ground and then put two large caps (100uF) between Vg and each supply (+4.5V and -4.5V). That way you could then remove the 2200uF altogether. But that would only work if  the supply is floating (like with a 9V battery). Then again it could be a little unstable which would cause distortion and thus defeat the purpose of removing the big cap.

squarewave said:

I'm sure you probably know this already but you would need to add the 2200uF DC blocking cap like in your first schematic to use this Studer circuit. Otherwise you're fry your phones.

But if you're using a 9V battery you could try connecting the phones ground to the virtual ground and then put two large caps (100uF) between Vg and each supply (+4.5V and -4.5V). That way you could then remove the 2200uF altogether. But that would only work if  the supply is floating (like with a 9V battery).

Click to expand...

You can't simply replace a 2200uF cap with 2x100uF. These caps are in the signal path. In a headphones amp, both channels share an identical output "ground". The resulting value of these two caps will determine LF crosstalk.
With 200uF, x-talk is -9dB at 100Hz, -26dB @ 1k.
With 2200uF it is about 20 dB better.
However, using separate caps will get rid of x-talk originating from this cause.

we had personal 4 ch headphone mixers in the studio the  last time ,  so each musician could dial in his own headphone mix of himself and the other band members,

looked like the studio owner DIY'ed these boxes does anybody make these nowadays?

CJ said:

we had personal 4 ch headphone mixers in the studio the  last time ,  so each musician could dial in his own headphone mix of himself and the other band members,

looked like the studio owner DIY'ed these boxes does anybody make these nowadays?

Click to expand...

(cough) Behringer.

Cheers

Ian

so just sit next to the drummer with no bones right?   

CJ said:

does anybody make these nowadays?

Click to expand...

Hear Technologies, Aviom, Allen & Heath, Digital Audio Labs, Roland, dbx, and yes, Behringer.
I use the Behringer system, and I must say I have no complaint at all. The main issue with any of these system is overcoming the defiance of musicians for something they think is "too complicated". For the most timid, I've made a couple of presets with which the talent just needs to adjust the volume pot to their liking.
In the 80's and 90's a large part of my income was derived from studio headphones systems, basic systems using power amps and distro boxes. Today, the concept of personal mixer is well refined and well accepted.
It's based on network technology, and as such it does not lend itself to DIY.

user 37518 said:

Thanks, it makes sense.

Click to expand...

I've seen the same strategy used in a high current power amp output stages to save a few parts (driver stage), but there is no free lunch and class AB delivers a true benefit in reducing the turn on time of power semiconductors (moot with modern class D switching stages).  Waiting for the devices to turn on is what causes crossover distortion at high audio frequency in output stages.

Headphone amps operate at much lower current and a decent modern (fast) op amp could probably deliver a respectable output. The schematic posted driving a transformer is probably not a very low Z, high current output.

JR

user 37518 said:

Im guessing that without that resistor, the opamp would have to provide a dc offset of around 0.7V to keep one transistor turned on and thus mantain dc feedback.

Now that you mention the transformer, how come the lower terminal of the primary is connected to a pair of resistors, one which is a pot that goes back to the non-inverting input of the opamp, why not just connect the lower terminal of the trafo directly to ground?

Click to expand...

As Abbey already mentioned that is some current feedback to linearize the transformer drive.

JR

PS: Speaking about personal monitor mix boxes for studios these only became popular after the bedroom recording boom made it possible for musicians to afford to record and mix themselves. Before that asking musicians to mix their own monitors would be met with a blank stare.  Typical studios would build their own headphone distribution systems and later personal mix systems. Now you buy almost anything you can think to make (cheaper and often better).

JohnRoberts said:

Headphone amps operate at much lower current and a decent modern (fast) op amp could probably deliver a respectable output.

Click to expand...

OPA551 will clean the wax outta your ears. Tested here with Sony 7506, Sennheiser HD280 and Grado SR125e, a couple of volts output was enough to be screaming loud.

abbey road d enfer said:

You can't simply replace a 2200uF cap with 2x100uF. These caps are in the signal path. In a headphones amp, both channels share an identical output "ground". The resulting value of these two caps will determine LF crosstalk.
With 200uF, x-talk is -9dB at 100Hz, -26dB @ 1k.
With 2200uF it is about 20 dB better.
However, using separate caps will get rid of x-talk originating from this cause.

Click to expand...

Ok. So output -> cap -> speaker -> ground is no different from output -> speaker -> cap -> ground.

squarewave said:

Ok. So output -> cap -> speaker -> ground is no different from output -> speaker -> cap -> ground.

Click to expand...

Yes if I understand the question, the order of components doesn't matter...

JR

squarewave said:

Ok. So output -> cap -> speaker -> ground is no different from output -> speaker -> cap -> ground.

Click to expand...

In a single-channel configuration, no difference. But in a stereo application, if the channels share a common path, that is a cause for crosstalk. You could have output->cap-> load->ground (times 2), or output->load->cap->ground (times two), but (output 1->load 1) + (output 2->load 2) ->common cap->ground has both channels interfering.