Thursday, August 18, 2011

More Thoughts on Single Ended/High Efficiency Speakers

I've been very critical of SET amplifiers (single ended triode) because of their inability to drive speakers accurately. To briefly recap SET amps have high output impedance, low damping factors and depending on the plate resistance of the tube, amount of feedback,etc can have varying abilities of producing accurate response from a speaker. Variations of more than 10dB "mistakes" either side of flat response can be quite common, and indeed these misrepresentations are often what some people find so appealing about SE operation.
I was building SE amps back in the early 90's when the SE resurgence had just started, back in the good old Sound Practices days with the joelist , and had a lot of fun trying just about every DHT tube known to us then.
Of all those amps i've kept one pair of 300B amps, one pair of 211/845 amps and my favorite SE amps - the technically challenging and innovative 6EM7-300B-6c33C-B .
( I see they are now cheaply built copies of my Cabernet preamp available, at twice the price - watch this design get copied too)
I've also had a long standing 10 year project to complete a 310-10-212E SE pair of monoblocks  that has been in partially  built mode for too long. (Will do a blog on this interesting design soon)
Recently I purchased a pair of Tannoy Gold 15" monitors and tried my 100 watt KT88 push pull amps with them with very pleasing results, so much so that I modified another pair of amps to use triode connected EL34 tubes for 50 watts of  musical power.  (using a big Peerless output trans built for me by Mike Lefevre)
Very nice sound, reasonably accurate. And beyond that to extremely satisfying, almost jaw-dropping sound when placed into a large, very good sounding room.
And with room to eventually end up with half a dozen amps, 3 preamps, various dacs and digital playback equipment and a Micro-Seiki RX-5000 turntable (with another sore back from lifting the almost impossibly heavy plinth)
Much fun, and eventually I couldn't resist to lug the big 6EM7-300B-6C33C-B SE amps into the room to try with the Tannoys, even though I expected the same old midrange prominent, SE sound.
I was very surprised to hear a  sound very similar to the big push pull tube amp sound, and with that touch of euphonics you get with SE sound, but certainly not the soggy bass, lucious but too promiment midrange- instead it was just incredibly natural sounding and very enjoyable.
So what's going on- are all my beliefs about the weakness's of SE amps completely wrong? Am I a goose?
Then I remembered - it had been so long since I used these amps I had forgotten the output stage was configured as a cathode follower . Instead of the usual practice of connecting the output transformer to the anode the output tran was connected between the cathode and earth.
Those of you familiar with cathode followers will know they are commonly used in preamps to provide a low output impedance, and often used as buffers on the end of dacs etc. They aren't my choice to use in preamps- the disadvantage of them is they are a feedback device and the disadvantage of the veiled sound they produce is too high to outweigh any advantages.
The cathode follower principle is rarely used in power amps, but again has the advantage of drastically lowering output impedance. The 6C33C-B tube is a low plate resistance tube, it's Rp is 100 ohms, compare that to a 211 with a plate resistance of around 4000 ohms! The 6C33C-B in SE operation has the lowest output impedance of any tube  and when used in cathode follower configuration is further drastically lowered .
Another advantage is that distortion is lowered- the clipping characteristics with cathode follower mode is symmetrical and clean. Whereas typical SE operation with anode connection is very much higher distortion and with the 6C33C-B quite an asymmetrical clipping character.
Here is the circuit- you can see the output trans under the cathode of 6C33C-b tube.

The disadvantages of cathode follower operation; a closed in,veiled sound doesn't seem to have the same degree of effect when used with power amps- it is audible, but the advantages of that clean power balance out the disadvantages to some degree.
Initially I had the amp set up with a switch so i could toggle between cathode follower and anode follower mode. Both methods sounded very good, but with the conventional speakers I was using I prefered the more balanced, accurate and clean sound of the cathode follower.
As much as I enjoyed the sound of these amps, I valued them mainly because of their technical ability and with the speakers I use preferd the big push pull or solid-state amps for my listening.
So they sat on a shelf for a couple of years till they were paired up with the Tannoys.
It would be very interesting to investigate just exactly what sort of load the Tannoy Golds and their crossovers present to a SE amp- I suspect it would be a very easy load, as others report that even 211 and 845 amps work well with them.
I'm not convinced a 211 or 845 would get anywhere close to the accuracy of a 6C33C-B SE amp in either anode or cathode follower configuration however and I'm astounded how good the combination with the Tannoy Golds is, especially in a big room where the sound is glorious and magnificent.
Reference quality in terms of accuracy and reproduction? No. 
Emotionally and musically satisfying? Yes , and in a big way.

For those of you interested in the effects of output impedance on speaker response this is an interesting page-

http://www.stereophile.com/content/primaluna-dialogue-seven-power-amplifier-measurements

http://www.stereophile.com/reference/60