Totem-Pole OTL with 6c33c tubes and 0.079 % THD at 20 Watt

[gkn202]

Hi all,

I've been looking into a series of OTL designs, but none really made me happy, namely because they lack symmetry and hence require "adjustments" which I do not trust to be stable over time.

My goal was to design an OTL which is symmetrical in nearly all regards.



Efficiency, costs, heat dissipation were no issue, only the (yet theoretical) quality of the amplified signal was my goal.

The mid-point referenced totem-pole allows such a design.

For each output tube, I've added its own driver stage with local feedback. The overall feedback loop allows more power but will (moderately) increase THD values, now very low at 0.079% at 20 W.

Now this is a LTspice simulation only, so please go ahead and give me feedback on this.


Please send me an email for getting the design and its library : email address removed Please PM



Thanks, Gary

[Alastair E]

I posted an OTL design a little while back based on the Inv. Futterman, Be interesting to see yours for comparison....

[gkn202]

Originally Posted by Alastair E

I posted an OTL design a little while back based on the Inv. Futterman, Be interesting to see yours for comparison....

Hi dear, I'm pretty new using this Blog, so I've some issues to upload my project files. I hope that a JPG comes thru...

Gary

[gkn202]

Here now the LTspice model and its library.

Feedback is highly welcome, go for improvements, there must be...

Gary

[Alastair E]

Ho-Hum!

Are you Serious!

ALL those 6C33C, and ONLY 20W! 10 6C33 Per Channel--Guess you have a HUGE Air-Con in your house!

--With THAT many, you should be running a couple of Hundred Watts O/P power!

In Circlotron, Ive got 56W RMS from just a Pair at 200V per rail!

[gkn202]

Hi, 10 pairs of 6C33 are OK for getting a low Outpiut impedance, I didn't want using a Futterman design as it is - to my appraisal - not symmetric, and this was a design goal.

Power was NO design goal, Low THD was.

Gary

[Alastair E]

Well Totem-Pole aint really symetrical either IMO!
--Thats why Futterman put the phase-splitter riding on the O/P to compensate for the drive requirements of the two sets of O/P tubes....

A Circlotron could be done that suits your requirements, 'symetrical' Much easier, and with much less tubes, and a great performance too!

A little neg feedback will lower the O/P impedance so you dont NEED Loads of mega hot tubes!

Have you a three-phase supply in your house? Guess the power cons. of your design will be horrendous!

As a matter of interest, Have you made amps using the 6C33C tube, They Get Amazingly HOT...?

[gkn202]

Hi, symmetrical as I use it semantically means that I use two floating Power Supplies and that the upper side of the amp is strictly identical to the lower part, actually it's a cut and paste operation. The Futterman solution is not symmetrical in that sense.

Now the design's goal was not efficiency, there I do fully agree that a Circlotron will likely be more reasonable. In this sense, I could have taken out the Cathode Follower stage for each 6C33 tube and leave everything to the overall feedback loop.

My goal is very low distortion levels at nominal power, 20 W are largely enough. I can likely take out 2 pairs of 6C33, lower a bit the feedback controls and still get 20 W out at a slightly higher THD level, but I have not yet estimated the impact on the damping factor.

Rgds/Gary

[tubetvr]

"Futterman design as it is - to my appraisal - not symmetric"

But the Futterman is symmetric, i.e it is fully balanced and gives very low 2nd order distortion, I get about -55dB, (measured) 2nd order with no global feedback. The uncompensated totempole is as Alastair write not at all balanced and gives hugh amount of 2nd order. What model did you use for 6C33C? most available models are just awful and are not even close to the real thing, the model presented here is the best I have found but it doesn't give correct distortion either.

BR Hans

[Alastair E]

Hmm--Lets look at power/heat etc for a minute.....

280mA through each 6C33C (1.1 odd volts across each 4 ohm cathode-resistor)

225V per rail.

Each tube will be dissipating 225 x 0.280, = 63W or therabouts

(BTW, Running a 6C33C at Just Over MAX PLATE diss which is 60W is asking for trouble!--Should keep it well below that. Recommended is below 40W P diss.)

This times number of tubes, (10) is 630W just in plate diss. for a Single-Channel.....

Add then the 30 ish watts Heaters for each one, so 300W ish....

We are up to nearly a Kilowatt of Power Cons, plus the drivers and pre-amp tubes, to make just 20W of sound output power....

And, You are gonna build a PAIR for stereo! Eee gads, Thank all thats Holy I aint paying your elect bill!

TBH, Its a non-starter in my book, Others may think otherwise, but its not for me personally!