Output Transformerless Audio Amplifiers (OTL)

[flyingtech55]

Hi all

Not sure if this is the correct section for this. If wrong, apologies and would the mods do the honours please. Here goes...

The subject of output transformerless valve audio power amplifiers has always interested me. I know that there are some commercial designs about. There was also the circlotron but that did/does in fact have an output transformer of sorts.

Anybody have any circuit diagrams, information etc on this subject?

TimR

[ppppenguin]

There was a thread about this a few years(?) ago but I can't locate it at the moment.

[flyingtech55]

Hi pppenguin

That may even have been me. It's been an ongoing interest! There is a small company who makes these things commercially but I can't remember the company name. ISTR that the output of the amplifiers is fairly low. Also it's very difficult to get diagrams etc. to see how they work.

TimR

[hotbulb]

Hmm- the circlotron is oneway of doing things, but ISTR power supply requirements are tedious. The other topology that comes to mind is the Futterman amplifier, a quick Google should bring up more details.
In the UK, Croft make/made OTL designs based on the Futterman using multiple PL519's as output valves

I've never been a lover of OTL designs, as they require paralled output valves to get the output impedance down to a sensible level, and they tend to double as room heaters.
Also, the 6C33 and 6AS7 devices that are comonly favored are high perveance/high transconductance devices. They might start out matched, but as they age they do tend to drift and I have known one or two self destruct-with consequent deleterious effects to the speakers, as they aren't protected by an output transformer.
PL519's are OK in this respect, but being designed for line output duties, aren't the most linear of things.
It is said that a good OTL amp has a high level of transparency and sounds extremely dynamic compared to a 'conventional' push pull design. Having listened to a number of them over the years I remain to be convinced.The best devices to use in OTL power amps are transistors IMHO
I ought to add one caveat here-I use an OTL headphone amplifier for a lot of my listening, which sounds superb. The topology is a 12SN7 mu-follower driving an 807 cathode follower. However, power output requirements are in the milliwatt range, and my Sennheiser headphones are relatively high impedance-it's not as if I'm trying to drive 50 or a hundred watts through 8 ohms

[flyingtech55]

Hi hotbulb

Yes...Croft are the people I was struggling to think of. Haven't ever been able to find any circuit diagrams of his amps though. I think most of the OTL amplifiers are fairly low power devices.

Presumably OTL amps are basically servo amplifiers and are in a bridge configuation. The old Boeing B29 Stratofortresses used valve servo amplifiers to power the remote gun turrets. ISTR they used loads of 6 (or 12)SN7s in the amplifiers.

TimR

[Framer Dave]

Lots of circuits here: http://pagesperso-orange.fr/franck.doucet/ click on DIY.

I have built many OTL amps over the last 23 years, most of which gave excellent results given the right speakers. I currently have, part-constructed, an OTL with 12 x PL519 in the output and an HT supply capable of 3,000 mA.
This may be a little bigger than Glenn Croft's.

Dave

[flyingtech55]

Hi Dave

Would you be prepared to let me see some of your own circuits that you have successfully used. No monster amps required, up to 15W or so is fine. These are purely for my own interest and eternal curiosity and NOT for any form of commercial use.

I'm never quite sure if an enquiring mind is a blessing or a curse.

TimR

[valveaudio]

The Sonab circuits are in JL Hoods book "Valve and Transistor Audio Amplifiers" one of them uses a small transformer just to generate NFB.

[Framer Dave]

Tim,

Yes, I'd be happy to let you have copies of circuits I've used. I don't have a scanner, so can't email them. If you send me a PM with your address, I'll post copies to you. Attached for interest is a photo of the current build.

Dave

Sorry for the poor photo, something amiss with the software.

[dominicbeesley]

If you can get hold of some high impedance speakers (800ohms or more) then look at some of the continental designs of the 60's.

I'm currently listening to a Saba Freudenstadt 8 that has four speakers connected to a OTL amplifier - sounds great! Not sure of the exact topology but I think its two EL84s in some sort of cascode arrangement?!?

Dom

[Sideband]

Hi Dom. That's probably a 'Bi-Ampli' type of circuit. I think paul has a description of it on the main site. Philips liked using this arrangement in some of their European designs. Most of them used an EL84 and an EL86, but there were other weird arrangments using PL81's of all things (because they needed a good heater/cathode insulation) and another used an EL84 and a UL84. They all sounded really good but used 800 ohm speakers which seem completely unobtainable now.


Rich.

[Duke_Nukem]

I have a design knocking about, and got as far as ordering a custom 400W toroid mains tranny to power it all, but change of company meant I lost the use of the metal bashing department so the project ground to a halt.

It'd use four EL509s to give around 12W of _real_audio. The 509's are good in these circuits as they not only work well with lower anode voltages (less waste power) but being derived for use in LOP stages they have excellent voltage ratings, helping the robustness of the design.

I have to admit one failure though ... I had to include one transistor, the purpose of which was to pull the upper 509's grid down if the -ve HT volts disapeared (fault condition) (the protection was in effect automatic if the +ve HT failed).

A friend of mine built one and it sounded good to me ... only he was driving electrostatics so in effect the OTL output stage was, in fact, driving a transformer !!!

Anyhow, I'll see if I can dig out the schematic.

TTFN,
Jon

[PWH]

Here's a prototype Circlotron OTL I made earlier . You would have to get the transformers specially wound,

It keeps the room warm too. 39 W RMS into 8R. 2Hz -50KHz +/- 0Db

Not for the faint hearted.
I think I have the Pspice diagrams somewhere if you want them

[Sideband]

Good heavens! What output bottles are you using?



Rich.

[flyingtech55]

These OTL amplifiers really are pretty scary in terms of their power supply requirements; and horrendously inefficient. However, they are potentially very good in terms of frequency response etc.

TimR

[PWH]

the output valves are russian triodes 6c33c-b

have a look at the link below which is how I derived the amp

http://members.aol.com/aria3/otlpaper/otlhist.htm

[oldeurope]

Look at this one:
http://members.aol.com/sbench102/inverted.html
I like this idea very much. Driving it with coupling triodes in the "vari µ mode"...
Nice project for 2008
Kind regards,
Darius

[kalee20]

Using a valve in 'inverted' mode, although OK in principle, could lead to excessive grid temperature. Generally of course, the grid won't dissipate any power, so there are no means to conduct heat away, and the manufacturers won't bother specifying maximum grid dissipation (Class C transmitting valves excepted).

By all means experiment - they're your valves - but don't use anything expensicve for development until you've got many hours clocked up with no adverse effects (gas released from overheated grid, or worst case, fused grid wires)!

Incidentally I seem to remember F Langford-Smith making a brief reference to inverted mode for obtaining phenomenally high input resistance (such as pH electrode amplifiers), rather than very low 'anode' resistance for OTL amplifiers.

Have fun - keep us posted!

[Brian R Pateman]

Kalee is right about the use of inverted mode in pH meter amplifiers.

I remember one (either Cambridge Instruments or Elliot I think) which used this sort of arrangement as the input stage in a pH recorder.

This was used in a power station which was commissioned in 1944 or '45 although I worked there in the 1970s so it was old equipment even then.

Regards,

[oldeurope]

Originally Posted by kalee20

Using a valve in 'inverted' mode, although OK in principle, could lead to excessive grid temperature. Generally of course, the grid won't dissipate any power, so there are no means to conduct heat away, and the manufacturers won't bother specifying maximum grid dissipation (Class C transmitting valves excepted).

By all means experiment - they're your valves - but don't use anything expensicve for development until you've got many hours clocked up with no adverse effects (gas released from overheated grid, or worst case, fused grid wires)!

Incidentally I seem to remember F Langford-Smith making a brief reference to inverted mode for obtaining phenomenally high input resistance (such as pH electrode amplifiers), rather than very low 'anode' resistance for OTL amplifiers.

Have fun - keep us posted!

Good morning,
I am a bit sureprised about these "arguments" against the inverted mode.
The same kind of arguments can be given against the conventional use of a triode.
Example given:
Using a valve in 'non-inverted' mode, although OK in principle, could lead to excessive anode temperature. ... Incidentally I seem to remember Heathkit making a VTVM in the non-inverted mode for obtaining phenomenally high input resistance (such as pH electrode amplifiers) ), rather than very low 'anode' resistance for OTL amplifiers.

This http://members.aol.com/sbench102/inverted.html is a very good idea.
Please take some time for it and think about it once again.

Kind regards,
Darius