Autotransformers - why does McIntosh use them?

[dnewma04]

Well, I am not bored with this topic. It's interesting to me that I still haven't heard a explaination of whether autoformers add or subtract from the sound and why others state that autoformers are not necessary in modern amplifier designs. I am not bashing McIntosh. I own McIntosh equipment. I am interested in why other people feel this way about autoformers. Just a friendly exchange in ideas. No reason to get worked up about this topic.
Jim

I've heard explanations of both in this thread.

Those who like them attribute potentially improved sound quality to allowing the amplifier to operate in a more consistent environment where the designers can design around that state and lower distortion The detractors think that having anything extra passive components in the signal path would cause some sort of detrimental effect. Both sides are most likely correct and the end result is that amps with or without autoformers can sound excellent.

One thing that is an absolute is that adding an autoformer isn't cheap (contrary to some implications above) and if there wasn't a sound engineering justification, the bean counters would have eliminated them long ago.

The one major item people miss is the autoformer enables an amplifier to have a large power output using lower voltage rails that use the output devices as current amplifiers well within their safe operating areas thus giving less distortion and higher reliability. The ability to give the devices an ideal match no matter what load is presented allows the designer to concentrate on lower distortion and higher reliability using lower voltage high current output transistors.

That does a nice job quantifying how employing an autoformer can improve performance. Unfortunately, distortion measurements like THD and IMD have no correlation to perceived sound quality.

 

[CarlV]

It's interesting to me that I still haven't heard a explaination of whether autoformers add or subtract from the sound and why others state that autoformers are not necessary in modern amplifier designs.

This is as good as you are going to get:

Jim, I believe your original question and continued interest is somewhat intuitive. You seem to be assuming that there might be an adverse affect just because of what you read on other forums. Unless you and others can be more specific, this notion is without merit. To my knowledge there is no imperical evidence that autoformer design necessarily improves or detracts from sound compared to other designs under optimum conditions. However, we do know that that autoformers provide a successful technical solution to some engineering problems outside of optimum conditions that other designs do not solve. I believe the previous explanation by Serge adequately addresses how autoformers improve overall operation and thus contribute to the avoidance of inferior sound characteristics.

I own both, only audible differences IMO are in the actually circuitry. Nobody hears the same, your nuances are not my nuances. I suggest you find out for yourself.

Carl

 

[Rob-F]

In addition to impedance matching, possibly the autoformer's tight electromagnetic coupling of the two output transistors might play a role in smoothing out the crossover distortion to a greater degree than quiescent bias alone can accomplish.

 

[damacman]

Good topic. I think the points made by Roger Russell and Terry DeWick are the most valid here. Both provide an insight into what McIntosh intended to achieve by using autoformers with their very first solid state designs - an idea that has obviously continued and has become part of the McIntosh DNA.

Keep in mind that McIntosh was late to the game with solid state amplifiers. While other companies rushed SS designs to market, McIntosh continued with tube products. Early SS designs were unstable and unreliable - two areas that McIntosh was committed to be the best. It is my understanding that the autoformer was the key to McIntosh releasing SS amplifiers that were both stable and reliable. Now, it's easy to agree that McIntosh couldn't have been more correct, as the MC2505 - their very first SS amplifier - is revered even today as being an incredible performer. [Here's a good read in regards: http://vinylhive.com/mcintosh-mc2505-the-solid-state-amplifier-that-change-the-world/ ]

Allow me to elaborate on Roger Russell's point in regards to the power capability of the amplifier with an autoformer (page 1 of this thread) at various impedances. Another way to think about this is to look at how traditional direct coupled designs are sold. Consider the following ratings:

100 wpc @ 8 Ohms
200 wpc @ 4 Ohms

So, really what you're buying is a 200 wpc power amplifier, and not a 100 wpc power amplifier. But, the only way you can get the 200 wpc (which you've paid in advance for) is to operate the amplifier into 4 Ohms per channel. OTOH, let's say that you purchase a McIntosh 200 wpc power amplifier. It may carry the following ratings:

200 wpc @ 8 Ohms
200 wpc @ 4 Ohms
200 wpc @ 2 Ohms

So, what you really have there is a 200 wpc power amplifier when it comes time to drive your loudspeakers. You get 200 wpc at 8 Ohms and 200 wpc at 2 Ohms as a bonus over the direct coupled design. Regardless of what you connect the amplifier to speaker wise, you absolutely have a 200 wpc power amplifier.

In regards to why other companies never adopted the idea, that is and isn't true. There have been other transformer coupled SS designs - the Sansui BA5000 is the one that comes most quickly to mind. But, the very idea of SS amplifiers with autoformer coupled output sections was very well . . . McIntosh. I believe most other brands had the philosophy that the autoformer would somehow degrade the sound, or act as a bottleneck between the amplifier and the loudspeaker. Some marketed against that based on those very ideas. 50 years later, some of McIntosh's competitors still proclaim this - albeit untrue. Here's a read you'll enjoy in regards to that, one of my first posts here at AK: http://audiokarma.org/forums/index....2300-mc2500-mc2600.371236/page-2#post-5918272 Keep in mind, that this is an absolutely true story.

Although the OP appears to have gone on to other things since his original post, this is a great discussion regardless.