Andrew Barta started out as a guitar player, albeit one who studied electronics and took a job fixing amps in a New Jersey music store to keep himself fed between gigs. This led to work customising amps; there was very little technical literature available at the time, so he did his own experiments to find out what effect various circuit modifications would produce. Eventually, he moved to Manhattan, where more work was available, and began to wonder whether there was any alternative to using tube amp circuits. I started by asking him what was the first thing he built from scratch.
"The first thing I built was a tube amp of around 15 watts with two EL84s in the output stage, three preamp tubes and a single 12-inch speaker -- a friend of mine still has it somewhere. It took a lot of work, and was hard to reproduce with a consistent sound. That was one reason I started to look into other ways of getting the tube sound. The other reason was that tubes are heavy, unreliable and temperamental. Also, my band was going into the studio and spending the first couple of hours getting a guitar sound by moving mics or trying different microphone types, and I wanted to cut down on this time. I wanted to be able to just plug in and play.
"I'd already done some designs for distortion in the '70s, because none of the commercial boxes around gave the same smooth tone as a tube amp. They had a very processed sound with very little dynamics, and I felt that touch-sensitivity was very important to the guitar player. As you play louder, it isn't just the level that changes, but also the harmonics."
The first product we saw from you was the little SansAmp Classic box with the DIP switches on top -- it was a great product. What kind of research did you have to do to get to that point?
"That was an interesting story, because when I built it originally, I had the DIP switches in it only to figure out what settings to use. But I wanted to get other people's opinions first -- I was doing some amp modifications for Foreigner at the time with Mick Jones, and he asked if he could try it out for a couple of days. When I told him I was planning to get rid of the DIP switches, he said that was one of the best features, because you could use them to create different amp setups and different sounds.
"Initially, I built the thing for myself, but people kept asking to buy one -- so I made 10, then 20. I tried to sell the technology to some of the larger companies, but no one was interested, so I decided I should put it in the music stores myself."
The production SansAmps use some form of hybrid circuit design, but I guess your prototypes relied on traditional, hand-built circuitry?
"Yes, and there was a lot of circuitry which didn't fit comfortably into the box -- so I was looking for a way to miniaturise the circuit. I was fortunate to hook up with a company which was prepared to work on a much smaller scale at the beginning. Back then, there was no surface-mount PCB technology; only the hybrid technology which uses a ceramic substrate. Metal is evaporated onto this, and then the resistors are formed by laser cutting. It is a very involved and expensive technology, but it was all that was available at the time, and we still use it. It took four or five attempts before the hybrid performed properly, mainly due to layout problems. We could change it now to surface-mount, but you know how people are -- they would say it doesn't sound quite the same. We've decided to stick with what we have, because that's what people want, and the same circuit is at the heart of all our products."
One area where tube amps do seem to sound different to their solid-state counterparts is at the bottom end, where the speakers can really thump. Is this a result of a damping factor, or is there more to it?
"There's more to it, and one thing I discovered recently -- which I built into the XXL overdrive pedal -- has to do with the even harmonics. They change with the level of the amp, creating an almost envelope filter-like effect. When a good tube amp decays, you can hear the harmonic content changing -- almost like an 'OW' sound. There is no single secret to making tube amp simulation work properly; it's a matter of combining the results of a lot of different observations, to give you the effect you're looking for."
One thing that has always surprised me is how much difference it makes whether you put the EQ before or after the overdrive stages.
"Oh yes, that was one of my very early discoveries, 15 or 20 years ago. I had an old Laney Klip amp, and though I liked the sound, there was something I didn't like about the attack -- so I made what went into the clipping circuit a bit brighter, and cut out some of the low end, which made it less of a fuzz type of effect. The combination of pre- and post-EQ gives you a tremendous amount of variation -- which is what we do in the SansAmp PSA1 preamp. In fact, the PSA1 doesn't entirely rely on preamp EQ, because some of the EQ, like the Crunch control, is in the feedback loop. If you look at a tube amp, the presence control is fed back from the power amp output. It's another small detail, but they all add up.
"To develop the speaker simulator, I miked up cabinets and drove them at different volumes, and recorded the results. Again, I don't want to reveal what I came up with, but the bottom line is not to believe anything until you're tried it. People tell you that it's important how a loudspeaker breaks up when it's played loud: maybe it is, maybe it isn't, but you have to do experiments to get at the truth. I used to arrange double-blind tests for myself, so that I could compare recordings I'd made without knowing which was which until afterwards. The psychological effects are very strong.
"There's also a myth about the soft clipping in tube amps, which is true in some ways, but false in others. If you look at an overdriven Marshall on an oscilloscope, you see a very strong clipping waveform. That's because you're overdriving the output tubes, which are biased in a way, in that they become non-linear very quickly at a certain level. If you do the same experiment with preamp tubes, you find that they'll clip softly -- but only on one side, because the circuitry is class A."
After building a career based on removing the need for an amplifier -- as the very name SansAmp implies -- you introduced a guitar combo into your range. What is special about this design?
"The preamp in there is based on existing SansAmp technology, and the combo has a ground-lifted XLR DI output with speaker simulation for recording. Many big rock names like Eddie Van Halen run their stage amps into a speaker simulator, combined with a dummy load. Then he runs everything through HH amplifiers, which are very clean. Once you've got a good sound, you just need to amplify it; you don't need to change it any more, and that's what we've done in the combo, by putting in a very clean power amplifier. It's the same with the speaker, because you've done all the simulation before it.
"Our power amplifier has lots of headroom and is limited by the power supply current rather than by the power supply voltage, which means loud transient signals won't be clipped. One of the wonderful things about power MOSFETs is that they can handle a lot more power than you actually use them for.
"There is also a real difference between tube and solid-state rectifiers, not in sound but in dynamics. A tube rectifier lets the voltage sag when you play a loud chord or note which pulls down the level, then lets it rise again rather like a compressor. You can simulate this very closely with a solid-state rectifier, by putting a resistor in series with it -- because this simulates the impedance of the tube rectifier.
"I haven't done anything like that in the power amp stage, because that would be cheating -- the DI sound would be different to the sound you get from the loudspeaker. If you listen to the XXL pedal, you hear this effect recreated in the preamp stage. Some people think there is a compressor in there, but that's not it. What happens is that there is a natural sagging effect, but again, I don't want to say how I've done it. On the combo amp, we call it the Weep button, because it gives you that kind of crying quality. You can hear it on the old Queen recordings from the AC30s, but the modern reissue amps don't seem to have it. The combo is also fitted with a genuine Accutronics spring reverb, because digital reverbs don't sound as warm on guitar."
In the course of the interview, it became clear that Andrew's passion for the guitar and his insight into electronic design have enabled him to develop a unique product, where the end result is judged purely by its sound and not by the technology used to create it. I look forward to trying out one of his new combos in the studio.
You once said that the secret of your sound was your circuit -- which is really a tube amp built in miniature, using FETs, right down to a push-pull output stage.
"Exactly. When you use a simple clipping circuit there is no real dynamics to it, but when you listen to an amp distorting, it changes its harmonic structure depending on how you play the guitar, what pickup you use... there are so many variables. I wanted to find a way to make a solid-state amp do what a tube amp does, and it's important to emulate both the preamp and the power amp. There are tube products out there that use a single 12AX7, and they just don't work properly. A lot of things happen in the output stage, including how the output transformer reacts, and although we can't put a transformer on a chip, we can simulate it. The transformer is really a huge inductor.
"The phase inverter is also very important, because it does some funny things which I became very familiar with in my tube amp days. When you put a master volume before the phase splitter, they never sounded good, but if you put them after, using a stereo pot, they were better. I took all these effects into account when I was building the SansAmp."
I should imagine that the transfer characteristics of the loudspeaker also play a large part in the sound.
"That's absolutely true, and there is the damping factor too. There are interactions between the loudspeaker and the transformer, but what you have to do is look at which effects actually influence the final sound and which are irrelevant. Some of the things that are different between a solid-state amplifier and a tube amplifier don't make as much difference as you might think -- I'm not going to tell you what is important and what isn't, but you have to do a lot of A/B testing to find out. I built a lot of prototypes, including a 12AXT tube overdrive, but I discarded that idea, because it sounded really thin and useless.
"One problem is that people say solid-state devices can't sound like tube devices, without even trying it. I started with a clean sheet of paper, not believing anything that other people said, and I tried everything I could. I used silicon transistors, germanium transistors, JFETs, MOSFETs, LEDs, different op amps -- everything. Another important point is the impedance of the circuitry, because most solid-state circuits have a relatively low input impedance, whereas tubes have high impedances. That's why I used FETs, because they can be used in very high-impedance circuits. It isn't the same using a transistor circuit and putting a high resistor in series with the input."