Home OTL Asylum

OTL, Output Transformerless Amplifier User Group.

[ Follow Ups ] Thread: [ Display  All  Email ] [ OTL Asylum ]
[ Alert Moderator ]

Re: What is wrong with Futterman?

I was chief engineer for New York Audio Labs during OTL production years and I have been upgrading and redesigning Futterman OTL's for 23 years. Go to http://moscode.home.att.net/ for my website and background. Let me give you the low down on the Futterman OTL circuit.
FIRST OF ALL- In the 23+ years of being associated with this amplifier design I have NEVER received a report of speaker damage!!! NOT ONE!! 2 things contribute to this. First, each tube has a 2 amp fuse so tube shorts and arc - overs are quickly removed from the circuit. Second, with the exception of a few custom OCL direct-coupled designs there is an output capacitor that effectively blocks all DC from the output. There is no more danger to your speakers from a Futterman than there is from any transformer coupled amplifier. I know for a fact that there are many more toasted voice coils from solid state amplifiers.
Let's talk about instability. There is a fundamental flaw in Julius's design that led him to use a kluge to fix. The flaw is that the low frequency pole point of the speaker and output capacitor is very close to the pole point of the coupling capacitor and grid resistors of the output stage. Varying speaker impedance can cause the amplifier to oscillate or motor boat. Julius alleviated this problem by splitting the feedback loop inot two paths and reduced the interaction of the speaker in the feedback loop. This sort-of worked but didn't eliminate the problem. This is easily corrected by moving the pole point of the output stage RC coupling circuit far away from the output cap-speaker pole point by using followers to drive the outputstage which allows a higher resistance in the grid of the follower circuit. Problem solved.
Other instability problems are solved by regulating the G2 of the input tube 6EJ7. This gets rid of another pole point in the amplifier.
RE: Positive feedback. We are really not thinking positive feed back but rather a bootstrap circuit to add the output voltage to the drive voltage for the upper tubes so that the upper tube control grids get the correct equal but opposite drive RELATIVE TO THEIR CATHODES, which are sitting on the output rail!
Without the bootstrap the amplifier has very high distortion. I serviced a number of Fourier designs that didn't use the bootstrap but their output was severely limited. Adding the bootstrap solved the problem. It is also necessary to use a pentode for the phase splitter in this design for wide voltage swings. Triodes just don't cut it. So Ralph's contention that the bootstrap introduces distortion is incorrect. It actually reduces distortion. Also the gain of this bootstrap is less than 1 so it can't create oscillations.
Re: Negative feedback. This amplifier runs at about 60 db of negative feedback. With an overall gain of 27-30 db we are looking at 90 db of open loop gain. With pentode's or triodes the open loop waveform is very distorted around the 0 point. Also bear in mind that the input circuit is a "current starved" circuit and runs at extremely low plate current which allows for enormous gain at low frequencies but it rolls off very quickly within the audio band starting around 100 hz. So the negative feedback at 20khz is much less than the NFB at 30 hz because the open loop gain is much less. The issues of HF phase shift are therefore alleviated due to the lower open loop gain at 20khz and beyond.
I recently built 3 Futterman style OTL's using the above techniques. The output power with 10 EL509 tubes in my Triode boost circuit is 200 watts into 8 ohms - CONTINUOUS! Frequency response is excellent and the amplifier produces controlled solid state like bass and the magical midrange and top octave that only an OTL can give.
Now I hate to say anything to dispariage Julius. But the truth is that 6LF6's cost under $3 each in the early '60's and Julius wanted to produce a cheap amplifier with no output transformer. I believe that that circuit's sonic performance was a bonus. However at that time the advent of the power transistor started the decline of the power tube and ended commercial hopes for the amplifier.
RE: Output tubes. The output tubes that were used in the JF OTL's were horizontal output tubes that were designed to switch peak currents of 1.4 amps for short periods of time. These are the only class of pentode tubes that work well in the OTL and are more efficient than triodes by a long shot. You need over 45 watts to light up a 6C33 and you can't get 1.4 amps of peak current. An EL509 (same as a 6LF6 except it has 9 pins, very European) uses 12 watts to light it up and can produce 1.4 amps very easily at lower plate voltage if wired right so the losses are less. My above mentioned OTL's produce 200 watts and draw about 1000 watts at full output which is 20% efficient. Early NYAL OTL 1 & 3's only produced 60 watts at the same current draw off the line. You do the math.
The biggest problem right now is getting reliable tubes. There is now an Octal based EL509 Tesla that looks promising but I haven't yet tested it. There was the EL509 type II that was reported to be the same thing but only passed .5 amps. The Sovtek EL509's that have the heavy glass bottle are also fake and are more like the 6pi45c ( I think this is the right #). At any rate, I tested a number of them at New Sensor a few years ago and they only passed1/2 amp. They aren't the EL509 that OTL's need. The EI EL509's NOS were quite good at that time but recent ones have over a 40% failure rate due to weak glass bases that crack and insufficient evacuation of gasses. I hope the Tesla's are better.
I hope this will help clear up a lot of misinformation about the Julius Futterman OTL. It really was a brilliant design and serendipitously proved to be one of the best sounding designs ever. They are my reference amplifiers for developing the Moscode 401HR. The HR is for Dr. Gizmo, Harvey Rosenberg or Harvey Resolution. See www.Moscode.com for the Moscode 401HR.
Happy Tinkering and Listening,
George Kaye
Moscode@att.net


This post is made possible by the generous support of people like you and our sponsors:
  Schiit Audio  

Follow Ups Full Thread
Follow Ups
You can not post to an archived thread.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]