Tube Power Amp Reviews

For anyone who's been around the audiophile block a few times, Conrad-Johnson Design is a brand that needs no introduction. My first acquaintance with Conrad-Johnson was before I began writing for Stereophile (more than two decades ago—time sure flies fast when you're having fun!). I was in the market for a new preamp, having become convinced that my Dayton Wright SPS Mk.II was the weak link in my system, and had narrowed my choices to two similarly priced products: a solid-state model made by PS Audio (I'm not sure of the model number), and the tubed Conrad-Johnson PV-2ar. They were carried by different dealers, who allowed me to take their preamps home over the same weekend for a direct comparison. I was impressed by both preamps, and was sure that either would represent an improvement over the Dayton Wright, but in the end decided to go for the PV-2ar. I later traded it in on a dealer's demo unit of another Conrad-Johnson preamp, the PV-5. And, as it turned out, one of my first reviews for Stereophile was of Conrad-Johnson's PV-11 preamp.

Through the years, I've often thought of reviewing another Conrad-Johnson product, but the timing never seemed right. Then, when discussing my review plans with John Atkinson, he mentioned that C-J had two new power amps, the LP66S and LP125M, and that I might consider reviewing one of them. Well, why not? One of the factors I consider in selecting a product for review is the likelihood that I'll have a good time listening to it, and based on my previous experience with C-J products, that likelihood seemed high. The LP125M is in a price range that I'm comfortable with—just—and I had on hand a similarly powered and priced solid-state amplifier, the Simaudio Moon Evolution W-7, which I knew to be an excellent performer, and which would provide for a good comparison. The review was on!

Description and design
With few exceptions—the ART preamplifier comes to mind—Conrad-Johnson products have always had a functional look, and the LP125M is squarely in that tradition. There's a touch of elegance in its 3/8"-thick, champagne-colored, anodized faceplate, but no one could accuse the LP125M of being audio jewelry. There's a substantial power switch, an easily removed cage for the tubes, and, in the back, a pair of RCA inputs (no balanced inputs) and a single pair of speaker connectors. Rubber rings are provided to place on the small 6922 and M8080 tubes (two rings per tube), to help damp resonances. Next to each output tube is an LED bias indicator and a bias-adjustment screw; a long plastic screwdriver is provided to make adjustments.

The bias adjustment is a bit tricky, especially if, like most audiophiles, you have in you a touch of the obsessive-compulsive. The idea is to turn up the bias control until the LED comes on, then back off until it just goes off. With new tubes, you're supposed to do this about five minutes after the amp is turned on, and repeat after 30 minutes. Then, if the amplifier has been off for some time, when you turn it on again the bias LEDs are likely to stay on for up to 30 minutes, some even longer, which might tempt you to readjust the bias. My advice: don't! Just leave them alone, and eventually—it may take more than 30 minutes—they'll go off. If any of the LEDs is still on after an hour, then I would re-bias that tube. After I'd gone through this routine a couple of times, the biases required no further adjustment during the review period—but it took a deliberate effort of will not to reach for that long screwdriver and start fiddling again.

You may wonder, as I did, why Conrad-Johnson doesn't use the autobiasing now embraced by many (most?) makers of tube amps. I asked Lew Johnson about it. His response: "Manual bias avoids the unnecessary feedback path if the servo control operates dynamically, and in any case introduces audibly inferior parts into the circuit to execute the adjustment simply for the sake of convenience. Simpler is usually better."

"Simpler if usually better" seems to be the Conrad-Johnson mantra. In the LP125M, a triode input stage is direct-coupled to a cathode-coupled phase inverter. The output stage uses a pair of 6550 tubes in an ultralinear configuration. A small amount of negative feedback is used to reduce distortion and provide a sufficiently high damping factor. Discrete DC power-supply regulators are provided for the input and inverter stages, in order to isolate each stage from the output signal. Both the audio circuit and the related power supplies use precision metal-film resistors and polypropylene capacitors, with no electrolytic capacitors in either the audio circuit or the power supplies. (I remember C-J's former marketing director, Tor Sivertsen, now retired, saying that the audio circuit and the power supply are like a pair of scissors, each making an equal contribution to the effectiveness of the "cutting action.") The use of proprietary wide-bandwidth output transformers produces what is claimed to be excellent high-frequency performance, and the input/output connectors and internal wiring have been selected for optimal sound quality.

Best-laid plans
My reference loudspeaker is the Avantgarde Uno Nano, which combines a horn-loaded midrange and tweeter with a powered subwoofer. Although capable of producing sounds that are at times astonishingly lifelike, and quite revealing of differences in the sounds of amplifiers, the Uno Nano's high sensitivity (over 100dB) and use of a powered sub make it less than ideal for evaluating amplifiers. Mind you, the sonic signature of the main amplifier is surprisingly audible through the sub amp, which can accept a speaker-level signal—but some characteristics of an amplifier that are present when the speakers are the Uno Nanos are not necessarily evident to the same degree when more conventional speakers are used, and vice versa.

To help me deal with this limitation, I borrowed a pair of a more conventional dynamic speaker model to serve as an additional reference: the Monitor Audio Platinum PL200, which I reviewed in April 2010, and with whose sound I was quite familiar. Kevro International, North American importers of Monitor speakers, kindly provided me with a loaner pair.

Although my Convergent Audio Technology SL-1 Renaissance preamplifier—also tube-based—might have seemed an ideal companion to the LP125M (a brief listen showed that the combination was indeed synergistic), I decided to use Simaudio's Moon Evolution P-7 preamp because it allowed a better-controlled comparison between the C-J LP125M and Simaudio's own W-7. The advantages of this arrangement were: 1) I could use the Ayre Acoustics CX-7e^MP CD player's preferred balanced outputs with both amplifiers, 2) the connections between the P-7 and the W-7 were the more optimal balanced rather than the unbalanced cable I would have had to use with the CAT SL-1, and 3) the P-7 allowed for more precisely matched levels when making comparisons than I would have been able to obtain with the CAT, which controls volume in much larger increments.

I began my listening with the Monitor PL200s, but there was a hitch: noise (hum/buzz) coming through the speakers. This was higher in level than I was expecting, and much greater than I had experienced with the dead-quiet Moon W-7. The noise was independent of the preamp's volume control; in fact, it was present with the amplifiers' input cables disconnected, but not when I inserted shorting plugs in the input jacks. I tried floating the ground by using cheater plugs on both amplifiers—a trick used by many audiophiles that electricians frown on because of the potential danger of shock in the event of AC being present on the chassis. In any case, this made no difference. The LP125M's manual suggests that attention be paid to the routing of interconnects (keep them away from AC cords). I followed this instruction, and it helped, but only to very minor extent.

Both amps were plugged into the PS Audio Power Plant Premier power conditioner; plugging them into the wall outlet (the same outlet for both amps) didn't solve the problem, and neither did unplugging the power cords of the Uno Nanos' subwoofer amps. I called Lew Johnson to see if he had any suggestions, and he pointed out that since the noise was gone when the inputs were shorted, this meant that the problem did not originate in the power amp. Fair enough, but the practical problem remained: Shorting inputs may be an acceptable procedure for measuring amplifier noise, but you'll have a hard time listening to music that way! I resigned myself to listening past the noise to determine the LP125M's sonic characteristics—and the noise really wasn't that bad.

This noise issue was again brought to the fore when the time came to return the PL200s to Kevro and switch to the Uno Nanos, whose sensitivity is a good 10dB higher than that of the Monitors: The level of noise was intolerable. I called Lew Johnson again, and he told me to try floating the ground of just one LP125M. Voilà! With the ground from the left-channel amp disconnected through use of a cheater plug, the hum/buzz virtually disappeared from both channels. Apparently, the problem was a difference in the amps' ground potentials; floating one of the LP125Ms (not both, as I had previously done) broke up the ground loop.

Things were hunky-dory for several weeks—but then a new noise appeared, in the right channel only: an intermittent sound, like paper being torn, that is familiar to tube fans as indicating a problem with a tube. I followed the standard troubleshooting steps of switching tubes between amplifiers one at a time, and isolated the problem as originating from one of the 6550 output tubes. A new tube was dispatched promptly from C-J; inserting it solved the problem.