Yesterday we discovered through our new found listening and relating skills that despite what our meters were telling us, we made a much better sounding phono preamplifier by increasing the headroom beyond anything that made sense. We did that by raising the power supply voltages by double what they were.
In later years we've learned there's far more going on than just headroom, but that's for another set of posts. For now we're understanding how we came to learn cause and effect of circuit changes and what we hear on the stereo.
At the time this experiment was going on there weren't a lot of high voltage semiconductors available to us. Running at 60 volts across the devices was pushing the limits of what we might be able to do and perhaps we could increase to 80 or even 100 volts but getting another magnitude improvement as we did was probably out of the question. If you'll remember, we had duplicated the Audio Research RIAA curve in the feedback loop and discovered we could get closer to their sound by raising our headroom level - but now we reached an impasse and couldn't come close to the high voltages they used with tubes. What to do?
In my experience, whenever a door closes and I reach a roadblock I go from despair and resignation to "wait a minute, I must be focusing too hard on the wrong solution". And so it was. If headroom was the problem and we could no longer increase the headroom in the same way the tube circuit did, we needed to do one of two things: go to tubes ourselves or try a completely different approach to solving the problems.
Many people have asked me over the years "if you love the sound of tubes so much why don't you just design with them?" The simple answer is they are just too flawed that I could never bring myself to go there. Microphonic, fragile, way past end of life, supply problems, performance degradation over time and the list goes on. Just wasn't going to happen. No, we were determined to create products that last a lifetime and without compromise that sounded as good or better than tubes. Sort of a lifetime quest.
One of the reasons most phono stages have the RIAA in the feedback loop is because of how "hot" the output of a phono cartridge is on the high frequencies - nearly 100 times louder than the bass frequencies - and so rolling them off to flat reduces the headroom problems one might encounter if you don't flatten them out.
We had learned that another designer, Stuart Hegeman, had implemented with great success a passive EQ in one of his designs (probably a Citation preamp). In order to get away from the headroom issue he broke the necessary 60dB of gain into 2 amplifying blocks: an input 30dB block with no EQ, followed by a passive RIAA curve (2 resistors and 2 capacitors), and another 30dB gain stage. The beauty of this design was the headroom had been increased by twice that of a single stage WITHOUT the use of EQ feedback. Further, it allowed audio designers to use any 30dB linear amplifier they wished and manipulate the performance of the amplification stage at will. In fact, one could design two zero feedback stages, or the first stage could be one type of amplifier optimized for high speed and low noise and the second block optimized for large voltage excursions, and so forth. The design freedom this method afforded us was huge and we went for it.
So now we used the high voltage discovery we made and then added to that the passive RIAA and we had oodles of headroom. Sonically we had come of age and learned a real lesson in trusting our ears and intellects rather than blindly following our instruments.
We never did beat the Audio Research sound but we sure got close and learned a whole bunch.