by John E. Johnson Jr.
Class A, single ended tube preamplifiers and power amplifiers have enjoyed a renaissance over the past few years, but Class A, single ended solid state power amplifiers are not very common. Only a handful of manufacturers produce them. Alternate Audio is one of those manufacturers. They have been in business since 1986, which isn't very long, but then, there are lots and lots of young companies in the hi-fi business now.
Dan Patten designs the components marketed by Alternate Audio, which he owns. When you own the company, design the products, and market them yourself, you can save a bunch of money and put it into the product. I think that is what happened with the two products I report on here.
The BCAP Preamplifier and CA-35 Power Amplifier are similar in that each has only two gain stages (most amplifiers have at least three). The first is the voltage gain stage (the input stage), which is differential (one MOSFET for the + leg and one for the - leg of the signal). The second stage is for current gain (the output stage), and is single ended.
The BCAP has one MOSFET for this purpose, while the CA-35 has three, in parallel. In order for the input stage to be coupled to the output stage without a driver stage in between, the input stage has to be high biased so that enough current is available to drive the output stage by itself. This makes the components run quite warm, at least for the CA-35.
The power supplies for the two stages in each component are separate. In the case of the BCAP, there are ± 22 Volts on 8,000 µF of capacitors for the input stage, producing 2 Joules of energy storage. For the output stage, there are ± 15 Volts on 8,000 µF of capacitors, producing 1 Joule of energy. The CA-35 has a larger power supply, with the same 2 Joules of energy for the input stage, but ± 15 Volts on 660,000 µF of capacitors, for 148.5 Joules, handling the output stage. This is a LOT of energy storage for 25 watts per channel, but that is what it takes for Class A, and it allows for more headroom when short term transients require it. With a high performance Class A/B power amplifier, it might have half or one third as many Joules of energy storage in the power supply as the rated output (8 Ohms) in watts (all channels added together). For Class A power amplifiers, you will have something around three times as many Joules of energy storage as rated output in watts. This is a major reason why Class A amplifiers are so expensive.
The looks of the two components are simple and elegant. On the front are on/off and standby toggles. LEDs indicate when the internal solenoids have switched to "ready". Overload or too much DC offset from inputs will switch the component back into standby mode. The rear panel has gold plated RCA input jacks for four inputs, labeled 1,2,3, and 4 on the BCAP, and one set of inputs on the CA-35.
The BCAP has one set of RCA output jacks, and the CA-35 has one set of very heavy five-way speaker binding posts. There are also on/off rocker switches on the back of each component which allow trickle current to keep the circuits warm. One thing I found to be a little odd was the fact that the left/right input and output labels were on opposite sides for the two components. This would mean that if you had the BCAP on the shelf above the CA-35, you would have to cross the interconnects right to left, and left to right, assuming you used the labels as is.
The BCAP tests showed it to have + 3.5 mV of DC offset, and was 3 dB down at 95 kHz from the measurement at 10 kHz. The CA-35 had a DC offset of + 0.75 mV and was 3 dB down at 267 kHz. Square wave response (± 10 Volts, 10 kHz) was much different between the two, reflecting not only the bandwidth, but some ringing that was present in the CA-35 (see closeup photo of CA-35 response at right). Notice that on the upswing portion of the waveform, it is almost perfectly straight for most of the upswing (where the electron beam path on the oscilloscope is faint). This means the amplifier is very fast. Ringing can be present when the amplifier has a large bandwidth, such as the CA-35. It can be reduced by lowering the bandwidth, but a smaller bandwidth can alter the phase response in the audio band. Ringing can also be reduced by adding negative feedback. There is no negative feedback in either the BCAP or the CA-35, except for DC back to the input. With single ended designs, it really isn't necessary if the circuit is linear to begin with. Negative feedback reduces 2nd order distortion, but increases 5th, something you would definitely not want with an amp like this. That would be like taking part of the sugar out of a recipe and replacing it with a tablespoon of vinegar. C'est la vie.
I listened to the BCAP and CA-35 with an Audio Alchemy CD package, Monitor Audio Studio 20 SE Speakers, and Nordost SPM Reference Interconnects and Speaker Cables. For a 25 watt/channel amplifier, the CA-35 had plenty of dynamics. That is the speed and overhead in the power supply of a Class A amplifier. The preamp and power amp together also sounded very sweet and mellow, with no harshness even when cranked. That's single ended performance, which eliminates odd ordered harmonics (odd ordered harmonics, e.g., 3rd and 5th, sound irritating, while even ordered harmonics, e.g., 2nd and 4th, can actually be pleasant).
Although this package performed nicely with symphonic music (Rachmaninov Piano Concerto Number 2) and rock, I really like Class A single ended designs with popular vocal music, such as Tony Bennett, Michael Bolton, and Natalie Cole. There is just something about the design that makes voices mellifluous. When listening to music like this, you don't need a lot of volume, so the dynamics can be spectacular even with just 25 watts.
In summary, Dan Patten has a pair of winners here, and I look foward to seeing more from this bright young designer in the near future.
- John Johnson