[Tommy Thong]

Recently, i bought LME49720HA n i want to bias this chip become class a (on my desktop cmoy), do i need put resistor or diode? if resistor, the value?(my PS is running DC+-18 volt)

 

[jcx]

Class A operation requires knowing the load current - so what headphones?
 
then you can assume a max SPL and calculate the current required by that load
 
"CMOY" doesn't seem to have any real meaning here any more - can you point to the schematic, identify the rail splitter, its current capability?

 

[wakibaki]

If your opamp is lightly loaded, for example if it is driving a second, high current opamp driver, then you might do something like this:-
 

 
This will pull about 2mA out of the opamp.
 
But as jcx says, you need to know the current required by the load before you set off down this route.
 
 
The max. guaranteed current the device will supply is 23mA, so at most you would want to pull ~11.5 mA out of it. You can see that the max. available symmetrical current swing would be halved.
 
The LME49720 is a good opamp, but most phones require a quite substantial current (in opamp terms) to achieve a reasonable volume and it is unlikely that if you are driving the phones directly from its output, that sucking current out of it with a current sink will do anything other than reduce its capacity to drive the phones adequately.
 
w 

 

[Tommy Thong]

Quote:

Class A operation requires knowing the load current - so what headphones?
 
then you can assume a max SPL and calculate the current required by that load
 
"CMOY" doesn't seem to have any real meaning here any more - can you point to the schematic, identify the rail splitter, its current capability?

to JCX n Wakibaki,
my opamp chip is 1x lme49720 ha for left-right channel n 1x opa211 ID for ground channel, the gain setup are 2.2 for all, my testing headphone is panasonic with 32ohm (impedance). i'm new around diy so would you explain it more detail? because i read the other thread about biasing the opamp chip to operate class a. Also some of them, they  installed 2.2Kohm resistor ( but i don't know which pin on the chip?) thank's tommy 
 

 

[Pars]

You might find Tangent's article on this useful. Pay attention to the last two paragraphs of the Theory section, where it explains why in your case, doing this may not be a real good idea. JCX and Wakibaki have given you good advice.
 
http://tangentsoft.net/audio/opamp-bias.html

 

[Tommy Thong]

Thank's to Pars, so i need to put the buffer chip( opa634 or diamond buffer) after the opamp chip, is it right?

 

[wakibaki]

OPA634 is no good in this application, you have +/- 18V, OPA634 is good for +/- 5V only.
 
You could try TLE2062 or NJM4556, preferably 2 in parallel per channel.
 
w 

 

[Tommy Thong]

oh..my mistake...i mean BUF634 or diamond buffer or AD797 for the buffer stage, is it possible?

 

[wakibaki]

Yes, in general, you can, although the AD797 is only +/-15V and is a bit weak in this role. Make the feedback loop enclose both gain stage and driver. You may wish to bandwidth-limit the gain stage or apply a compensation cap.
 
w

 

[Tommy Thong]

ok..i will try to put 7815-7915 on my Power supply tonight n i ordered some ad797 (they will arrive soon) btw thank's for the advise bro

 

[peranders]

Tommy, the LME49720 has 0.00003% (typ) distortion if you have a proper pcb and power supply. Adding a class A circuit will hardly create even lower distortion. I think you will get a decreased sound quality in fact. It's better to put your efforts in the power supply.

 

[Tommy Thong]

yeah...i put the 7815-7915 on my PS and some film caps between +ground-, it sounds better than before. I ain't do anything about the opamp yet