Adcom GFP-565 preamplifier Page 3
The specific comments I've made about the GFP-565's phono preamp can also be applied to the line-level stage as well. It's clear that the designers had the same goals in mind for both sections of the preamplifier, but a few specific comments are in order here as well. Dynamics of the line-level stage are also superb. The brass crescendo on the CD version of Siegfried's Funeral Music (Solti Götterdämmerung, disc 4, track 8) is reproduced with tremendous impact. At the same time, the soft tympani roll at the very beginning of the Funeral Music is delicate and well articulated. In fact, if a high-quality LP playback system and CD player are used for the comparison, the two versions of this recording sound remarkably similar.
The Forging Scene "depth test" I referred to above, when heard on London's CD version of Solti's Siegfried, passes with flying colors as well. Again, there can be no question that the LP and CD have the same master tape in their ancestry. In my experience, it is rare that an LP and CD of the same source sound similar enough to consider the differences of little consequence. Both the LP and CD must be extremely faithful replicas of the original, and their respective playback devices must exhibit high degrees of sonic accuracy. Beyond that, the design of an RIAA phono preamp and line-level stage with such similar sonic characteristics is an impressive feat. The greatest difficulty does not lie in the design of the line-level stage, but rather the phono preamp, where so many design parameters must be considered and so many things can go wrong.
I recently acquired the new Chesky CD of Jascha Horenstein conducting Dvorak's "New World" Symphony 9 and Wagner's Flying Dutchman Overture and Siegfried Idyll. This important addition to the Horenstein CD discography is sonically Chesky's finest orchestral CD to date. The accuracy of instrumental timbres and clarity of inner lines are reproduced by the 565 with breathtaking realism. The line-level stage's ability to reproduce the low-bass region with virtually unrivaled impact and clarity can be heard on any Telarc CD, but I find Maazel's Rite of Spring to be particularly impressive. I have not heard any commercial preamplifier with better low-frequency extension, impact, and clarity. (Only my custom-built reference preamp is superior to the 565 in this respect, but I went to the extreme of using separate supply regulation for the line-level stage. My phono preamp and tape-output buffers are powered by a different set of regulators.)
The GFP-565 is an extremely accurate preamplifier. Whether or not it is musical depends strictly on the source material. The 565 does not add those euphonic colorations which make every recording sound sweet, pleasant, and musical. On the other hand, it adds none of the stridency or harshness that many still believe characterizes IC-based audio circuits. Instead, it tells the truth, and accuracy is, in my opinion, what high-performance audio is all about.
I've had the opportunity to compare the GFP-565 with several well-known preamps, including the older GFP-555, the Forté 2, and the Audio Research SP-11 II. The GFP-565 is, by far, the quietest of any of these preamps, particularly the phono section. I've never encountered a quieter phono preamp. The GFP-555 sounds gritty in the high frequencies when compared with the 565. The 555 has a more analytical, colder, and less musical sound. Dynamics are not as impressive, particularly in the bass region. The 555's phono preamp was worse in this regard than the line-level stage. The phono preamp also loses inner detail, becoming congested during orchestral climaxes.
Several design weaknesses in the GFP-555 have not been repeated in more recent Adcom products, most notably the GTP-400 and the GFP-565. First, the phono preamp is powered by an essentially de-regulated power supply, since the designers put 33-ohm decoupling resistors in series with the supply lines. These resistors, used in many preamp and CD-player circuits, are often called safety resistors, since they protect the supply regulators in the event of a catastrophic circuit failure. This "safety" feature is unnecessary with properly designed supply regulators, which should shut down if the output rails are shorted to ground. Nearly all IC regulators are protected in this fashion. Decoupling resistors substantially raise the output impedance of the supply, defeating the purpose of the regulation, in turn degrading the dynamics and soundstaging.
The GFP-555's power supply also has fuses in the secondary circuit of the power transformer. This is another bad practice. To maintain as low a supply impedance under dynamic conditions, any fuse should be confined to the primary side of the transformer. Finally, the 555 has a rather dirty signal path, with the output IC op-amps driving the tone-control circuits and filters. No less than eight switch contacts appear in the signal path between the IC's outputs and the jacks on the rear panel! The contacts are bad enough, but when they appear just prior to the preamp's output, the loading effect of cables, etc. can make them perform in a non-linear fashion. From a perfectionist point of view, there should be no switch contacts between the output IC and the output jacks. The sound of the GFP-555 is noticeably inferior to the newer and cheaper GTP-400, and not the least of the reasons for this are the GTP-400's superior power supply and cleaner signal path.
The $900 Forté 2 was a disappointment. This op-ampbased preamp sounds very musical and pleasant and reveals more inner detail, particularly at climaxes, than the GTP-400 or the GFP-555. The high end is a bit smoother than the other two, and the dynamics are better than the 555. When compared with the GFP-565, however, the Forté 2 falls short in every aspect of performance. The Forté's soundstage is narrower than the 565's, in both the phono preamp and line-level stages, and the Forté's depth perspective is not nearly as impressive. The 565 maintains a deeper soundstage, which remains rectangular right to the back of the stage. The Forté's soundstage collapses to near mono, like a triangle, at the rear of the stage, a characteristic I've consistently observed in equipment with inadequate power supplies. Both lateral imaging and depth are less precise than the GFP-565.
The 565 reveals more inner detail and is far more dynamic than the Forté. The bass region, in particular, is reproduced with greater impact by the GFP-565. The 565's phono preamp is also considerably quieter than the Forté's. The Forté 2's Tiffany-like gold-plated connectors and machined aluminum knobs give a true high-end appearance and feel. Then why should this $900 preamp be so easily beaten by an $800 contender? I was shocked when I first removed the Forté 2's covers. This $900 preamp has an unregulated power supply!
The Forté 2 has an outboard supply containing a skimpy 6VA transformer, rectifier diodes, and a pair of 2200µF capacitors. This minimal raw supply is connected to the main chassis with 6' of cable which is no heavier than 20 gauge. Inside the main chassis are a pair of simple R/C filters for the positive and negative rails, each containing a pair of 1000µF caps. The R/C networks are on the line-level end of the main PC board. Six inches away, a pair of 4.7µF caps are all the bypassing the phono preamp is given, one hundred times lower in value than the bypass capacitors used by Adcom in the 565. In terms of dynamics, the Model 2 might better be described as mezzo-Forté, but contrary to what is often believed, the inadequate dynamics have nothing to do with the power-supply operating voltages, instead being directly related to the lack of supply regulation and the resulting high output impedance.
Both the phono preamp and line-level stages of the Forté 2 are buffered using a discrete transistor arrangement biased for class-A operation. Like the GFP-565, the buffers operate inside the feedback loop and, in the case of the phono preamp, drive the RIAA feedback network. It's strange, then, that Nelson Pass chose a high-impedance RIAA feedback network, with its resulting higher noise level. The Forté's RIAA network looked awfully familiar, and is, in fact, the same network used by Lampton and Zukauckas in their late-'70s Mark I preamp (footnote 5). The Forté's buffers may not be capable of enough current to drive a lower-impedance RIAA network, which is unfortunate since the Linear Technology LT-1028 op-amp used by Pass is capable of a much lower noise level than he has realized in this RIAA circuit. Finally, the IC op-amps are not soldered to the PC board; they are socketed.
Victor Campos challenged me to compare the GFP-565 to the Audio Research SP-11 II, recently discontinued but considered by many to be a definitive statement on preamplifier design. Since Stereophile owns one, arranging to borrow it was no problem. I really didn't expect the 565 to hold its own with a $5000 preamp, but the results were not exactly as I had anticipated. I had expected the SP-11 II to easily outperform the 565. Only half of it did so, and only after a long warmup (after the unit had been on for ten minutes it sounded pretty bad). The SP-11 II's line-level stage is absolutely gorgeous; it is the best line-level stage with gain I have heard. It is incredibly detailed, and that detail and transparency are maintained in the most complex, heavily scored passages. The soundstage is huge, both in width and in depth.
Subtle changes in dynamics which are masked by other preamps are revealed with stunning realism by the SP-11 II. Its ability to recover ambience is also exceptional. In all areas of performance, the SP-11 II's line-level stage was an order of magnitude better than that of the GFP-565. But the Audio Research preamp is seven times the price. Is it seven times as good? Absolutely not. It's difficult to quantify such differences, but I would say the SP-11 II's line-level stage is roughly twice as good as the 565's. In terms of cost vs performance, the law of diminishing returns is definitely at work here.
The SP-11 II's phono preamp was an entirely different matter. The GFP-565 easily outperformed the Audio Research phono section. The 565 has lower noise, greater inner detail, and is free of a small amount of extraneous grit which the SP-11 II adds to the music. This grit does not appear in the form of high-frequency stridency or edge. It is simply a subtle layer of grundge which separates the listener from the innermost details of the music. Since phono preamps must also be fed through a line-level stage, the 565 would appear to be placed at a disadvantage. But, even though the 565's line-level stage is inferior to that of the SP-11 II, the capabilities of its phono preamp were still able to be revealed. As a system, the GFP-565's phono and line stages are superior to the SP-11 II for the playback of long-playing records.
One obvious listening test would be to connect the 565's phono stage, via the tape outputs, to a line-level input of the SP-11 II. This is really not as straightforward as it seems, since the signal would have to pass through the tape output buffers of the 565. The 565 would have to be internally modified to bypass the buffers, so I did not attempt such a test.
Those who love the sound of the SP-11 IIthere is good reason for feeling that way about the line-level sectionmay be shocked to find out that the SP-11 II is not giving them that great tube sound. The SP-11 II is a hybrid design using both 6DJ8 triode vacuum tubes and field effect transistors (FETs), but tubes do not play a major role in the gain characteristics of this preamp. The line-level section consists of two stages, each using two FETs and a single 6DJ8. The two FETs are cascaded, with the second FET feeding the triode tube in a cascode configuration. The audio signal actually passes through twice as many transistors as tubes, with the transistors providing most of the voltage gain.
As someone who finds it difficult to use a piece of commercial equipment without eventually modifying it in some way, I'm sure that many tweakers will succumb to the temptation to modify the GFP-565. I must honestly warn you that most attempts to alter the 565 will result in a degradation of its performance. First, the 565's passive components are the finest available today from any source. Better-sounding resistors and capacitors simply do not exist, and I don't care whose Super-Duper-Wonder-Deluxe resistors and capacitors you care to consider. Second, the active devices have been so carefully selected for their respective circuit topologies that inserting your favorite op-amp will invariably degrade the sound. In particular, there is not a single op-amp I know of which, if substituted in the Adcom phono preamp, would result in anything but poorer performance. The GFP-565 is a thoroughly engineered product that delivers the maximum performance that can be extracted from these circuit topologies in this chassis with this control layout. Before you mess with it, you'd better know exactly what you're doing.
The GFP-565 is a superb preamplifier that can hold its own with all but the most expensive products. If an $800 preamp can be compared to a $5000 product, with the results being a toss-up depending upon which source material you prefer, it speaks very well of the lower-priced contender. Based on the listening I've done over the past six months, I'd rank the preamps I've compared here, and in my GTP-400 review, as follows:
Phono section Line-level section
1) Adcom GFP-565 1) Audio Research SP-11 II 2) Audio Research SP-11 II 2) Adcom GFP-565
3) Forté 2 3) Forté 2
4) Adcom GTP-400 4) Adcom GTP-400
5) Adcom GFP-555 5) Adcom GFP-555
6) Hafler DH-110 6) Hafler DH-110
Considering that the Forté 2 serves as an example of solid Class C sound, and the SP-11 II, prior to being discontinued, stood near the top of Class A, the GFP-565 easily deserves a Class B recommendation. The GFP-565 now sets the standard for preamps under $1000, and should remain a worthy contender for years to come.
Footnote 5: Lampton, Michael and Don Zukauckas. "A Low Distortion IC Preamplifier/Control Unit," Audio Amateur, February 1979.