Electronic – Operational amplifier design

You are not losing gain, and your scope traces even show that since they are both about the same height.

The difference is the DC level. At the left side of C2, the DC level can only ever be between power and ground. The circuit is therefore biased so that the signal is centered far enough from either extreme to allow swing in either direction. That's why the signal at the left side of C2 is not centered around 0.

However, a net DC bias is undesirable in a delivered audio signal. Even HiFi audio only goes down to 20 Hz, so doesn't include DC. Any DC on the audio signal would cause dissipation in a resistive load, but without doing anything useful. In fact, DC bias will make ordinary speakers work less well and heat them unnecessarily. Since DC bias is not needed and can cause trouble, audio amplifiers usually produce a output signal without DC bias, which is another way of saying it is centered around 0 V. This is the purpose of C2 in your circuit. It blocks DC but passes the AC audio signal.

So you're not loosing gain across C2. It's merely doing its job of blocking the DC level.

Well, this is exactly what should happen ! The reason is as follows:

Your input signal is a sinewave between 0 - 2.5 V

I could also say that your input signal is a 1.25 V (2.5 V /2) DC voltage + a sinewave with an amplitude of 1.25 V.

The thing that confuses you is the 1.25 V DC ! It is converted to differential as well, resulting in the DC voltages you see at the outputs.

If you want to prevent this you need to make that DC voltage at the input 0 V. This might be impractical.

Fortunately what you can also do is shift the DC reference voltage of the single to differential converter. I see that is has a Vocm input, if this is a common-mode level input then you could apply 1.25 V DC to it (the DC voltage I mentioned above) then the single to differential converter will consider 1.25 V as the commonmode level and the outputs will get the same DC voltage.