Balanced Gainclone

[ErikdeBest]

I've been looking at this forum for information to build a Gainclone. I wanted to keep costs low, and with patience I got most components for free. Last week I gathered the last components and courage, took the iron and build a mono Gainclone a la Peter Daniel. And it worked.

Now I want to build a stereo balanced gainclone for my 8 ohm speakers (they don't go below 8 ohms). I do this for two reasons. I believe it will sound better, and I hope I can build a balanced Gainclone with a single supply, which won't cost me almost nothing. (I saw that more people wanted to build a Gainclone with Single Supply, and someone said you could buid a balanced with single supply.)

I know there are opamp's that transformates an unbalanced signal in balanced. But then I would need a separate PS for those opamp's... To eliminate the need for those Opamp's, I thought I could combine Peter Daniel's inverted Gainclone with Steve Eddy's non-inverting Gainclone. The result is shown below. I want to supply it with a single supply. (OPA's can work with single supply.)

What are your comments? Would this work? Problems could be impedance matching at the input, and speakers must be at least 8 ohm.

[theChris]

though i am new to this all, isn't the non inverting gain formula differnet then inverting gain. it looks like the gains would be different.

[BrianGT]

Originally posted by theChris
though i am new to this all, isn't the non inverting gain formula differnet then inverting gain. it looks like the gains would be different.

This is true. The gain of the non-inverted amplifier is Av = 1+Rf/R1, and the gain of the inverted gainclone is Av = -Rf/R1. (R1 is the 10k in the picture, and Rfeedback is the 220k) This would mean that with the potentiometer at full signal, then the gain of the non-inverted would be 23, and the inverted would be -22. This issue could be fixed with a 230k resistor for Rf in the inverted amp.

Another issue that I am not sure about is that the potentiometer in series with the parallel combination of the amplifiers should effect the gain of the inverted amplifier, decreasing the gain as the volume decreases for only the inverted amplifier, since it is essentially in series with the 10k. Putting a voltage follower/buffer in there should fix this. Anyone else have any comments on this?

I see this taking away from the simple nature of the gainclone, and if you want to bridge them, National has an application note on bridging the lm3886, which might be helpful, as it seems better for bridging. Also, keep in mind that each side of the amplifier is driving half the load impedence, meaning that if you are driving a 4 ohm load, then each amp sees a 2 ohm load, which can cause problems.

added link to national note on higher powered solutions:
http://www.national.com/an/AN/AN-1192.pdf

--
Brian