In the web he said this is an OTL amplifier.
SINGLE ENDED TUBE AMPLIFIER "TRSAT-20"
(FUNERAL BELL FOR NO OTL)
http://www.super-otl-audio.com/Description.htm
But, this is really an OTL?
It looks like a parafeed.
Best regards Jaime
SINGLE ENDED TUBE AMPLIFIER "TRSAT-20"
(FUNERAL BELL FOR NO OTL)
http://www.super-otl-audio.com/Description.htm
But, this is really an OTL?
It looks like a parafeed.
Best regards Jaime
Factual errors aside, the text is pretty hard to follow.
I will concede that his bias adjustment scheme is pretty clever in its simplicity.
I had a pretty bad experience with 6C33 in otl, described in detail here
www.kta-hifi.net under amplifier projects.
Considering the number of output tubes used both the damping factor and output power seem low. I guess if I were to attempt something similar I would use an autotransformer in place of that choke..
Kevin
I will concede that his bias adjustment scheme is pretty clever in its simplicity.
I had a pretty bad experience with 6C33 in otl, described in detail here
www.kta-hifi.net under amplifier projects.
Considering the number of output tubes used both the damping factor and output power seem low. I guess if I were to attempt something similar I would use an autotransformer in place of that choke..
Kevin
As I have described many times before I have very good experience with using 6C33 in OTL circuits, however I wouldn't ever consider building a cathode follower as an output stage.
For me it seems that the the inverted Futterman type of coupling is the optimum solution for an OTL amp.
It has much lower output impedance than a cathode follower and of course not higher as is written in the article. It is also very easy to apply quite high amount of feedback while keeping high phase margin thereby achieving very low output impedance.
BTW, can anyone make sense of the text concerning amplifiers with and without feedback in the article? I can not, if any the described amplifier with adamping factor of 2 have very much higher output impedance than what is achievable using feedback.
Regards Hans
For me it seems that the the inverted Futterman type of coupling is the optimum solution for an OTL amp.
It has much lower output impedance than a cathode follower and of course not higher as is written in the article. It is also very easy to apply quite high amount of feedback while keeping high phase margin thereby achieving very low output impedance.
BTW, can anyone make sense of the text concerning amplifiers with and without feedback in the article? I can not, if any the described amplifier with adamping factor of 2 have very much higher output impedance than what is achievable using feedback.
Regards Hans
He's from Croatia, and it would seem that there's an "Engrish" problem here.
If you take 3.0 X V^2 and 30.0 X V^2, where V is a constant, the difference will be 10db, like he says. However, a "constant voltage generator" is a battery. While this could drive a loudspeaker, the result probably wouldn't be real interesting to listen to.
He probably meant a constant current generator. Either way, NFB or no NFB, won't do anything for the change in impedance with signal magnitude and frequency that occurs with the speakers, especially since a big component of that is mechanical, not electronic.I wonder what he does with that air coil. "Output tubes are fed with DC via voltage very massive coil. It is an air coil, with maximum linearity. Total load of the output". Very massive coil means lots of DC resistance ans stray capacitance. So how does he get the bandwidth he claims?
600W go in and 20W come out: that's an efficiency of 3.3%
Not very good for all those 6C33's.
Miles Prower said:
600W go in and 20W come out: that's an efficiency of 3.3%
Not very good for all those 6C33's.
In fact it's easy to obtain more than 50W on 8ohms with just 4 6C33 per channel... that's Ciclotron!see here http://www.audiocostruzioni.com/a_d...ati/otl-ciclotron-bebo/otl-ciclotron-bebo.htm
A circlotron is unnecessary complex and doesn't offer any advantages compared to an inverted Futterman circuit which also have lower output impedance.
It is quite easy to reach 80W or more with 4 6C33C tubes with very low output impedance and excellent linearity while still maintaining high reliability.
Regards Hans
tubetvr said:
Mmmm I see. Sorry, I wasn't saying that Ciclotron is best, because I'm not technical enought to know what are the difference between the two, I was just saying that everyone is obtaining more than 20W from 5 6C33
Could you point me at some simple links where OTL in general and Ciclotron/Futterman is explained?
The best summary I have found describing different OTL circuits is here http://members.aol.com/aria3/otlpaper/otlhist.htm however I don't agree with their choice of a circlotron as the best circuit. Output impedance is very important in an OTL as it decides how many output tubes that will be necessary, an inverted Futterman will have about half of the output impedance compared to a Circlotron.
An advantage of the Circlotron is usually described as that it is balanced and therefore have low even order distortion, but that is true for all push-pull type amplifiers including the Futterman. That there are no real advantages together with the increased complexity is the reasons I prefer the inverted Futterman instead.
See here for a summary description of my OTL http://www.tubetvr.com/otl.html I get 25W from just 2 6C33C. I am working with a version that uses 4 6C33C and in that one I get 80W in 8 ohm with 2nd order distortion of only 0.2% without any global feedback, that is a proof of excellent output stage balance.
Regards Hans
What's wrong with a cathode follower output? I'm still in the thinkin' and cypherin' stage, but I was planning on putting 6 to 8 6C33C's in parallel as CF's with a choke load OLT to run a still undesigned line array (8 drivers in series with 64 ohm total imp.)
Is the 600 ohm per tube why you say that a CF is no bueno or is there more.
Is the 600 ohm per tube why you say that a CF is no bueno or is there more.
Brian Donaldson said:
The problem is the low efficiency. Take another LQQK at the design in question. It takes 600W of DC power to deliver 20W of output. That's an absurd waste of both 6C33s and DC. You can get enough power to drive that array, with fewer tubes and less wasted power (and heat, what doesn't go to making sound becomes waste heat) with a better design. That would include something like a Circlotron (used because it's teh sexxors
) or the Inverted Futterman topology that's been mentioned here before. In either case, you get lower impedances than what a straight cathode follower will give, and so you won't need all those 6C33s.
By my rough guesses, I can get 100 to 150 Watts RMS at 64 ohms of single end triode magic. I don't have my notes here now, but the poor efficiency is due to the impeaadance mismatch of driving 8 ohms with 1 or 2 tubes. By paralleling many tubes and wireing the speakers as a high impeadance, I'm getting my 100+ watts without needing 5A plus current. Instead I only need 1.7A and I can utilize more of the voltage swing. I may even ad some GNFB to help reduce output impeadance.
I guess you didn't catch my way around the efficiency problem.
The thing I don't like about Futterman is the asymetric gain of the top and bottom halfs.
The circlotron is interesting and I like not having big caps in the output, maybe food for thought for some Pass susy thing in the future, but not now.
The only drawback is the amps and speakers will be an inseperable pair. Considering that, they may be built in.
I guess you didn't catch my way around the efficiency problem.
The thing I don't like about Futterman is the asymetric gain of the top and bottom halfs.
The circlotron is interesting and I like not having big caps in the output, maybe food for thought for some Pass susy thing in the future, but not now.
The only drawback is the amps and speakers will be an inseperable pair. Considering that, they may be built in.
A class A OTL amplifier will always have low efficiency and of course a CF will need to run in class A. In order to not exceed the max anode dissipation of each tube you need to connect many tubes in parallell and that is why the efficiency will be so low. You can always get much higher efficiency by using class AB and that is why all comercially high power OTL amplifiers are made to run in AB. I don't like to connect tubes in parallell because of relaibility problems so I am trying to use as few tubes as possible > conclusion use the inverted Futterman circuit, using only 2 6C33C tubes you be able to easily get more than 100W in 64 ohm in class AB1.
Brian Donaldson wrote
You have not understood the Futterman circuit where feedback around the output stage and the phase splitter equalises the gain of both output tubes, a Futterman is not an ordinary Single Ended Push-Pull where the tubes have unequal gain.
As I wrote before my amp with 4 6C33C give 80W at 0.2% 2nd order distortion before feedback is applied, that is a proof that the gain of lower and upper tubes are equal, it is not possible to achieve better symmetry in a circlotron.
Regards Hans
Brian Donaldson wrote
You have not understood the Futterman circuit where feedback around the output stage and the phase splitter equalises the gain of both output tubes, a Futterman is not an ordinary Single Ended Push-Pull where the tubes have unequal gain.
As I wrote before my amp with 4 6C33C give 80W at 0.2% 2nd order distortion before feedback is applied, that is a proof that the gain of lower and upper tubes are equal, it is not possible to achieve better symmetry in a circlotron.
Regards Hans
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.