Higher levels

Join Our Community Subscribe to Paul's Posts

Back to op-amps. We were just getting set to find out why we needed a high voltage op-amp.

The vast majority of integrated circuit op-amps are power supply limited to 15 volts. What this means is that, with few exceptions, it’s tough to get more than a clean 10v rms out of the device. While this isn’t usually a problem for preamplifiers because most power amplifiers clip when you input more than 2v rms, there are examples where we need more. One perfect example is our original phono stage.

I threw together this schematic for you (sorry it’s not better. I am on vacation and quickly threw it together while Terri was waiting to go to dinner). This is what we originally came to market with. The two triangular blocks represent op-amps, each with a gain of 30 (R4 is 30KΩ and R3 is 1kΩ same with R’s 6 and 7). R5 is 47kΩ to set the phono cartridge loading. Look at the middle grouping. This is the RIAA curve (R1 is 31.6kΩ, R2 is 4.64kΩ, C1 is 0.0233mf and C2 is 0.68mf) if my memory of the last 45 years is still working. This gave an RIAA with accuracy to better than 0.1dB along the entire 40dB curve.

The problem here is that first op-amp. The pre-equalized signal from the phono cartridge came in hot because there’s no filter in this first stage of a passive phono preamplifier. This caused a sonic “strain” on higher frequency loud passages because we were bumping up against the outer ranges of the op-amp’s linear area. You could hear it strain.

There were two ways to eliminate that strain: divide the RIAA up so the front op-amp had the first leg of the curve in its feedback loop, or increase the op-amp’s power supply voltage beyond 15.

We chose the latter because we wanted the negative feedback to be the same at all frequencies (if the first have of the curve were in the loop then feedback would be different at every frequency, changing the sound in ways unacceptable to us at the time). That required us to keep the passive curve and figure out a way to increase the power supply voltage beyond what the chip IC could handle.

Tomorrow we’ll see what happened.