the reason I ask for a handdrawn schematic is when tracing the actual circuit you will find the problems in the process. When drawing using a computer you are putting down what you think it "should be", it is not the actual wiring any more.
about using resistors as rail spliters: 4.7K as spliters will only give you loose bass, nothing solid. If you can make sure both batteries are always the same %, center tap ground is always better. Because ALL of your headphone currents goes through the ground.
The problem might be the chip, per that schematic in the picture, if you use a BJT opamp you are guaranteed to have DC offsets which, when combined with the low ground current capability (due to using 4.7Ks as spliter), will give you off center voltage readings (one being higher than the other, instead of being the same). The 2132 being a FET input opamp should not do that, but if the opamp itself is damaged when handling-- the FET input is very sensitive to static discharges, that is why they are always shipped in a silver bag -- they can create DC offsets.
Try another chip and see if that solves the problem. IC's are pretty cheap.
about using resistors as rail spliters: 4.7K as spliters will only give you loose bass, nothing solid. If you can make sure both batteries are always the same %, center tap ground is always better. Because ALL of your headphone currents goes through the ground.
The problem might be the chip, per that schematic in the picture, if you use a BJT opamp you are guaranteed to have DC offsets which, when combined with the low ground current capability (due to using 4.7Ks as spliter), will give you off center voltage readings (one being higher than the other, instead of being the same). The 2132 being a FET input opamp should not do that, but if the opamp itself is damaged when handling-- the FET input is very sensitive to static discharges, that is why they are always shipped in a silver bag -- they can create DC offsets.
Try another chip and see if that solves the problem. IC's are pretty cheap.