Chip (gainclone?) type headphone amp

[tschrama]

Hi,

If the resistor is to be inserted, however, I don't think its value should exceed 1/10 of the nominal impedance of your headphones, based on my own experience.

This could depend very much on the type of headphone. I use Grado SR-60, and they sounded fantastic driven by the current output of a TDA1541a DAC. So I conclude that headphones can sound good driven by a very much higher impedance than the headphones' nominal impedance.



Hey Paulb,

Nice to hear from you once again.
Was the feedback taken before or after the current limmiting resistor? Do you agree that the LM1875 functions in class A in your application, and could this be a argument in favour of power-opamps for headphone use?


Regards,

Thijs

[paulb]

Originally posted by tschrama
Was the feedback taken before or after the current limmiting resistor? Do you agree that the LM1875 functions in class A in your application, and could this be a argument in favour of power-opamps for headphone use?
Regards,

Thijs

Hi Thijs.
Feedback was taken before the resistor - don't forget I used this for both speakers and headphones (just used a switch to turn off the speakers).
The intent of the resistor is not so much current limiting as voltage dropping. Taking feedback afterwards would work against this, and just drop your headroom, so I wouldn't recommend it.
The LM1875 does not operate in Class A, except at very low power levels. Maybe these low levels would apply to headphone use (probably without the resistor). I know others have reported good results from the LM1875 driving headphones; maybe this is why. I think the main argument against chip power amps for headphones were the noise levels.

[Tomo]

Hello,

I like the idea. Perhaps I should build a gainclone set at lower gain. I think it might just do very wonderfully with Gradoes.

T

[tschrama]

My thinking exactly... a dedicated headphone amp for low impedance Grado's with power opamps..

The LM3886 has a Iq of about 25mA (guessing a little from the datasheet) and a minimum gain of 10x for stable operation. So if feedback was taken after a 330 Ohm resistor, the Grado would be voltage driven (Low output imp) while still protected for overdrive and the LM3886 would function stable as a follower in class A. Remenber that a Grado only needs 150mV rms realistic SPL levels. This would (about) equal a Log potmeter volume control set at 12o'clock driving this baby from a CD player. Voltage offset would become a major issue.. hmmmmmmm ..

Another idea... take the 330 Ohm to the other side (first the headphone, then the 330 Ohm resistor to ground and take the feedback after the headphone: now it's current driven right? How about making it adjustable .. Rod Elliot has written something like that... hmmmm


I think I will give the first option a try...Peter .. are your reading this? How do you feel about this idea?

Regards,
Thijs

[Vladco]

Thanks for the replays. I’ve got more questions. Could it benefit from regulated power supply? If yes what kind is acceptable? Or it doesn’t mater? Should I use output capacitors? I ask these questions because I was able to find only one parts supply store in my area – rat shack.
Vlad.