Opamp based ground for CMOY.No TLE2426.
Sep 6, 2006 at 6:34 AM Thread Starter Post #1 of 5

]|[ GorE

500+ Head-Fier
Joined
Mar 16, 2004
Posts
796
Likes
11
Somethings seems to be wrong with the resistor values mentioned by Tangent in his VGrounds article for an opamp based virtual ground(I cant get a TLE2426) in : http://tangentsoft.net/elec/vgrounds.html

That's diagram No : 8 from the top in the above mentioned page.
In the diagram 2 Nos of 220K and 1 Nos of 1K resistors are mentioned.

Now,Tangent himself has recommended the use of 22K and 10K resitors(instead of 220K and 1K),and it is this config that i had used resulting in a distorted cmoy with bad splitted loads.That is the 22K and 10K config gave me perfect +/- 4.5V till i plugged the actual op-amp for the amp stage in its socket.

So,what are the correct resistor values ?? And NO,there was NOTHING WRONG with my soldering etc(i have already made 3 normal WORKING cmoy's.)

Note: Tangent mentions at the page's bottom that he has not tried the opamp based virtual ground config.Dont want to sound insolent but i almost killed myself trying to get that cmoy working !

PLEASE HELP !
 
Sep 6, 2006 at 10:48 PM Post #3 of 5
I doubt that the PINT schematic should give you good answers, simply because almost all of the headphone amp designs have buffered grounds which are for the most part immune to uneven splits. The only amp that uses a simple resistor divider is the CMOY.

If I'm guessing correctly, you're talking about the 220k resistors in the buffer-based virtual ground driver. Those are buffered, and they won't have uneven splits no matter what resistor values you use. In my CMOY, I used 4.7k resistors; those are probably what you should use, too. In general, the higher the resistor values, the greater the offset but the smaller the current draw (and thus the longer battery life you get). For the details on how this works, re-read the "resistor divider" section. It should makes sense that the split becomes uneven after plugging in the opamp, since the offset comes from the opamp itself. According to tangent's calculations, it's not unreasonable that the virtual ground should split to +3.7V and -5.3V with an opamp in place.

To calculate how much current draw the resistor divider will take, use ohm's law: I=V/R; so with two 4.7V resistors, the divider will draw less than 1mA. The lower and more accurate your resistor divider is, the less battery life the cmoy will have. If you really want an even split, you could try the buffer-based virtual grounds with a cheap opamp
 
Sep 7, 2006 at 12:13 AM Post #4 of 5
Quote:

Originally Posted by ]|[ GorE
Tangent himself has recommended the use of 22K and 10K resitors


Please remind me where I said that. Lowering the divider resistors can make sense sometimes, but I can't imagine why you'd want to raise the value of the output resistor any higher. If anything, you'd want to lower it if you can.

Quote:

what are the correct resistor values ??


Try the design as originally spec'd. Try using a lower output resistor, or even no output resistor. The output resistor value depends on the op-amp. Which begs the question: what op-amp did you use? Not all chips are equally suitable for this job.

Quote:

PLEASE HELP !


Please stop shouting.
 

Users who are viewing this thread

Back
Top