[Roger A. Modjeski]

Hi,

My apologies for being away for so long. I have been working up some new tube applications for the low powered OTL with matching Autoformer. As I previously stated, an autotranformer has many advantages over a push-pull plate transformer. Small OTL's naturally like high impedances around 100-400 ohms. This is still low compared to the RM-10 transformer which is 12,000 ohms. I will keep you posted on the progress.

This posting was partly inspired by the following discussion on diytube.com: http://www.diytube.com/phpBB2/viewtopic.php?t=3235. Thanks to Lafish for letting me know and responding to the nay-sayers. I will be referring them to our gentle circle so be ready.

Tube applications are many and varied. Although the data books list 2-5 these are just examples and are meant to help engineers get their circuits going without having to find the right load impedance, voltage and bias currents. There are an unlimited number of other applications that can be created and that is what I find interesting. None of the amps I have produced use published applications. Instead I study the tube, have a look at the actual curves on my curve tracer, see what I can do without exceeding the tube specs. In fact I usually get far longer life than most other applications, in the range of 5000-10,000 hours. The RM-9 was the first amp I know of to state 10,000 hour tube life which has been well proven by the users of 930 units some as old at 24 years now. I have a local customer who bought his in 1987, uses it daily and still has the original Siemens tubes. We did have to replace one that got gassy after 12 years. The other 7 are still going strong and still well matched!

Exploring tube applications starts with choosing the right tube(s) for the job. Tubes are designed and classified by the expected usage. For instance Horizontal output tubes for TV are especially suited for that but they also make fine output tubes yet not for all circuits. These tubes are not well applied in triode or ultralinear circuits but are excellent in certain pentode and OTL circuits. My new OTL series will be using a selection of them based on power desired.

The ultralinear/triode/pentode tubes with which we are so familiar were designed with that in mind and will tolerate the high screen voltages that are found in these traditional circuits, though not if the B+ exceeds 500 volts.

Audio triodes like the 45, 300B and Emission labs series we sell were designed for high voltage low power amps. These tubes allow for very simple circuits and their popularity is in those applications. Western Electric had good reason to create the 300B and thousands of movie theaters around the globe had amplifiers using that tube.

Triodes with low mu (typical 2) were designed as pass tubes in regulated power supplies.  These are not good choices for audio output tubes, though many may disagree. Those who have experienced amps built with these tubes find exactly what I predict; drifting bias, extreme power consumption for very little output and generally unreliable operation. These tubes were not meant to hold a steady bias current, instead they are made to vary current widely just as needed in a voltage regulator as the load varies. They have high heater demand and in today’s energy crisis they are the dinosaurs of audio tubes.

Any questions?

[DustyC]

Working in an engineering environment, I see the need to reuse the previous design work as a basis for problem solving (for problem "A", use formula "C") because it saves time and money. However, when a client has a unique problem, it's time to get out the slide rules! 
Past history has shown that there are disbelievers when new research and development has shown a better way. I think there just a handful of true audio designers, the rest are practicing "textbook engineering".

[6BQ5]

Exploring tube applications starts with choosing the right tube(s) for the job.

The ultralinear/triode/pentode tubes with which we are so familiar were designed for ...

Audio triodes like the 45, 300B and Emission labs series we sell were designed for ...

Triodes with low mu (typical 2) were designed as ...

Isn't the objective, ultimately, to amplify a signal with as true fidelity as possible, in driving a reasonably efficient 4 - 16 Ohm speaker? It seems the choice of tubes and topology are simply no more than various means to the same end objective.

[Roger A. Modjeski]

Like any other product, amplifiers come in all sizes, flavors and complexities. I am moving toward simpler and simpler circuits because I find that results in better sound.

I can give several examples where the designer went overboard for one parameter over another. The Citation 2 power amp had instability problems due to the high feedback and wide bandwidth that the designer felt was more important. Read the HK lit, they bragged on the ultrawide bandwith. The results were horrible. Dyanco was designed at a price, down to the penny. Not a bad design but the driver has problems, the power transformer runs very hot and the filter caps are on the edge. Some modern amps have computers inside to monitor the bias which should not have to be monitored in a good design. Some modern amps have more transistors (often mosfets) than tubes in the audio path and regulators that don't regulate well. Putting mosfets in tube circuits is rather silly, not to mention the reduction in reliability. Most amusing is amplifiers where the wrong tube is chosen because of how it looks or its availability at a very low price because they are stacked to the ceiling in warehouses all over Russia.

The previously mentioned amplifiers and most I have tested have poor clipping behavior. Many amplifiers oscillate into certain loads. In many cases the problem is "birdies" that occur on certain portions of the wave causing some odd distortion. They can only be identified on a scope and usually under music conditions.

I tested a 120 watt mono amp recently and found that the designer had achieved very low distortion at the expensive of a lot of heat and very short tube life. The low distortion was obviously his focus at the expense of everything else I hold dear.

To sum it up I think "design focus" is a good way to look at why amps are so different. I strive to have a balanced focus between decent specs, long tube life, reliability, beauty, simplicity and value. I would love to hear other designers speak on their focus and choices that follow.

[Ericus Rex]

Speaking of OTL amps, I've heard some OTL critics say something along the lines of 'speakers need transformers to operate properly' or something like that (sorry, my technical know-how is still in its infancy).  What could they be talking about and how would you address their criticism?

[Roger A. Modjeski]

The challenge for OTL amplifiers is supplying current. Commonly used OTL tubes have current capabilities from 250 mA/tube for the 6AS7 to 1.2 amps/tube for the ones I use. To get 100 watts into 8 ohms it takes 5 amps peak and 7 amps peak to get 100 watts into 4 ohms.  OTL amps have one bank of tubes to go positive and one bank to go negative. Before you know it, you have a lot of tubes.

There are several ways to make a 100 watt OTL. One pair of tubes can yield nearly 100 watts into 64-100 ohms, but only 10 or so into 8 ohms. An autotransformer can allow the single pair to put 100 watts into 8 ohms and with other taps can put 100 watts into any load. To get 100 watts into 8n ohms directly takes 4 pair of the same output tubes, 4 times the power consumption and 4 times the replacement cost.

Whether an OTL needs a transformer or not is totally dependent on the output current of the amplifier and the speaker's minimum impedance.

I will be building OTL amps with and without autotransformers. I will also be building them at various power levels because not everyone needs 100 watts. I will also be making available separate autotransformers that exceed the performance of the toroids currently available.

[zybar]

The challenge for OTL amplifiers is supplying current. Commonly used OTL tubes have current capabilities from 250 mA/tube for the 6AS7 to 1.2 amps/tube for the ones I use. To get 100 watts into 8 ohms it takes 5 amps peak and 7 amps peak to get 100 watts into 4 ohms.  OTL amps have one bank of tubes to go positive and one bank to go negative. Before you know it, you have a lot of tubes.

There are several ways to make a 100 watt OTL. One pair of tubes can yield nearly 100 watts into 64-100 ohms, but only 10 or so into 8 ohms. An autotransformer can allow the single pair to put 100 watts into 8 ohms and with other taps can put 100 watts into any load. To get 100 watts into 8n ohms directly takes 4 pair of the same output tubes, 4 times the power consumption and 4 times the replacement cost.

Whether an OTL needs a transformer or not is totally dependent on the output current of the amplifier and the speaker's minimum impedance.

I will be building OTL amps with and without autotransformers. I will also be building them at various power levels because not everyone needs 100 watts. I will also be making available separate autotransformers that exceed the performance of the toroids currently available.

Roger,

Any idea on possible costs for say 100 watts or even 140-150 watts?

What will be different about your autoformers than say Paul Speltz's?

Thanks,

George

[EDS_]

The challenge for OTL amplifiers is supplying current. Commonly used OTL tubes have current capabilities from 250 mA/tube for the 6AS7 to 1.2 amps/tube for the ones I use. To get 100 watts into 8 ohms it takes 5 amps peak and 7 amps peak to get 100 watts into 4 ohms.  OTL amps have one bank of tubes to go positive and one bank to go negative. Before you know it, you have a lot of tubes.

There are several ways to make a 100 watt OTL. One pair of tubes can yield nearly 100 watts into 64-100 ohms, but only 10 or so into 8 ohms. An autotransformer can allow the single pair to put 100 watts into 8 ohms and with other taps can put 100 watts into any load. To get 100 watts into 8n ohms directly takes 4 pair of the same output tubes, 4 times the power consumption and 4 times the replacement cost.

Whether an OTL needs a transformer or not is totally dependent on the output current of the amplifier and the speaker's minimum impedance.

I will be building OTL amps with and without autotransformers. I will also be building them at various power levels because not everyone needs 100 watts. I will also be making available separate autotransformers that exceed the performance of the toroids currently available.

Mark me down as being extremely interested in OTLs sporting between 15 & 30wpc.

[Roger A. Modjeski]

Thanks, the low power one with the autotransformer will be much better than getting the power from lots of tubes.

[Phil]

Roger,

I'm curious what you think about the 6H30 tube used in a preamp? 
Thanks.

Phil

[Roger A. Modjeski]

It's ok, a bit of overkill, but ok.