On 18 Ago, 00:00, Miguel Gimenez wrote:
Are you sure? For a volume pot the scale should be logarithmic. I
mounted a linear pot in my first amplifier, and the effect was the same
you observed: the volume rised quickly, and only the first quarter of
the pot was practical.

Well, I did not check if it was indeed logarithmic or if the minimum
volume was actually 1/16Vpp but the audio wave was almost invisible on
the scope, so it was at least under 0,1Vpp. In the LC7533 datasheet
says: Attenuation curve: Pseudo-curve A.

The output amplifier gain of 100 seems excessive to me, a 1Vpp signal
at full volume becomes 100Vpp, which is obviously badly clipped at
32V.
1/16Vpp (min volume) is 6Vpp output to the speakers.

For now I reduced LA4280 gain to 30 by adding 680 ohm resistors in
series with the Feedback pins. But I still wonder if it rings a bell
to someone.

This solution looks far better than the first; the 56 Ohm resistor
reduced the volume but didn't solve the clipping, and you were wasting a
lot of energy.

Yes, I calculated since it runs on 32V a 1Vpp signal would be
amplified to 30Vpp at highest volume thus getting the maximum
undistorted swing at the output. I wonder why the designers didn't
make this simple calculation.

Thanks