Lme49600 headphone amp

[Dr. Popo]

Hello, I want to build headphone amp as per schematic. What kind of electrolytic caps should i use? audio grade, low esr, low impedance? any help would be greatly appreciated.

[tomchr]

I would move the 100R resistors from the inputs to the outputs of the LME49600 to ensure that any difference in offset, current drive, etc. doesn't make the amps fight. You can probably reduce the resistors to 10R if you move them to the output. The resistors will raise the output impedance slightly, but it'll still be very, very low due to the global negative feedback.

But, honestly, I don't see the point of running multiple LME49600's in parallel for a headphone amp. They can deliver upwards of 500 mA short circuit current. For my headphone amp (driving a pair of Sennheiser HD-600's) I implemented the circuit in Figure 4 of the LME49600 Data Sheet. I chose to use the DC servo to set the LF pole.

Anyway... To your question: I'd use any electrolytic or tantalum cap that has low ESR and decouple it with a good polypropylene film cap for the supply caps in your circuit. The opamp will need separate 100 nF polyprop decoupling.

~Tom

[Sebastiaan]

Originally Posted by Dr. Popo

Hello, I want to build headphone amp as per schematic. What kind of electrolytic caps should i use? audio grade, low esr, low impedance? any help would be greatly appreciated.

Dear Dr. Popo,

Nice idea!

Some comments:

1: Make the 100K input resistor 20K, so that the positive leg sees the same impedance as the negative leg. This will reduce DC offset and destortion. A 20K input impedance is a good value in my opinion, unlike u feed it with high output impedance tubes.

2:Like Tom commented, put resistors (something in between 47R and 100R in series with the output. Else the output opamps will become really hot and can even get destroyed. Then take the feedback AFTER those output resistors to keep the output impedance low.

Good luck!

With kind regards,
Bas

[Sebastiaan]

One rectification of my former post. The BUf634 and LME49600 don't need output resistors, since they have it itnernally already. So in that case u can keep the output connected as you have now, which I recommended then.

With kind regards,
Bas

[Dr. Popo]

Thanks to everybody! I will now look for low esr capacitors and start soldering. Does anybody have any suggestions as to how i should solder LME49600 to pc board? I am good at soldering but never have come across ic in power pad package!

[Sebastiaan]

Originally Posted by Dr. Popo

Thanks to everybody! I will now look for low esr capacitors and start soldering. Does anybody have any suggestions as to how i should solder LME49600 to pc board? I am good at soldering but never have come across ic in power pad package!

Don't forget to change the 100K input resistor to ground for 20K or a equal of the feedback resistor. Also, are you not afraid to fry your headphone voice-coil in case of DC offset? I would add a DC servo to the system. (no input cap! ) If u find it hard to design a DC servo I am happy to help you with your application and do that part for you

With kind regards,
Bas

[Dr. Popo]

Dear Bas, Many thanks for your reply. Any help would be greatly appreciated!

[Sebastiaan]

Originally Posted by Dr. Popo

Dear Bas, Many thanks for your reply. Any help would be greatly appreciated!

Dear Dr. Popo,

Hold on I will draw for you tonight. The reason why I like to do this is, is because I always liked te headphone amplifier concept with the diamond output buffers. Weather it is the Burrbrown Buff634 variant or the new National ones.

Is there a reason you didn't add a volume control?

With kind regards,
Bas

[theAnonymous1]

Originally Posted by Sebastiaan

If u find it hard to design a DC servo I am happy to help you with your application and do that part for you.

Why not just follow the Figure 4 headphone amp circuit in the datasheet to begin with? It already includes a servo.

More buffers can be paralleled as needed.

[Sebastiaan]

Originally Posted by theAnonymous1

Why not just follow the Figure 4 headphone amp circuit in the datasheet to begin with? It already includes a servo.

More buffers can be paralleled as needed.

Sorry, I didn't look at this sheet yet, I see it now. The difference is only that the OP choose a voltage gain of 20, and the circuit in this sheet has a voltage gain of 3. The servo design and the resistor values must be changed with a voltage gain of 20.

With kind regards,
Bas

EDIT: The circuit in the actually has a voltage gain of 3, since the 1K resistor to ground and the 1K resistor to the output of the opamp makes those seen as 500 ohm.