• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Designing an OTL / Circlotron

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi guys, i'm new to these forums but definitely not new to audio, been building my own amps since, forever?
In the past i've built myself a few amps, ss power amps: 120W/8r BJT, 320W/4r Mosfet (both with Toroids, diode bridge, and big caps), god knows how many guitar amps and effects, and a SS-moded version of Fred Nachbauer's phono preamp (http://www.platenspeler.com/diy/phono/hounddog/hounddog_phono-l.gif - his site is down but i found the schema elsewhere).
Basically, I replaced either r105 or 106/7 (so long ago i cant remember, probably r105) with a double-bjt constant-current source, the 3 neons with Zener diodes, unswitched magnetic riaa eq with whatever polystyrene capacitors than fell into my bag at Uni, and fudged a power supply with 2 back-to-back toroids (i think 240V->12V, then fed that into a 9V->240V to get 320ac, 450DC, and a 7812-driven heater supply).
It worked great, wish i had some photos of it all, or could at least listen to it, it's all on the wrong side of the world from me.

So now, I thought i'd try my hand at building my first tube power amp. I've been working with SMPS-repair for the last 6 months, and i've learnt a lot about Power Factor Correction and Switching circuits, so i can now build a very nice, clean, PFC-driven 400V rail, and switch it down to 12V for the heaters.

So i was thinking of trying an OTL circlotron, the main thing that's put me off from building a tube power amp in the past is the availability, expense, and designing of an output transformer.

So first, some theory questions. I've designed (very) high power RF (80kW-peak 10%-duty 50MHz) in the past, so i definitely know what happens when the source-transmissionline-load are not completely impedance-matched. But I was also taught that for any voltage source, output impedance should be as low as possible to minimise internal losses in the source and send all the power to the load. This is certainly true of the solid-state amps i've built, lowest possible output-impedance is best.
So which is right? Or is it that at low (audio) frequencies, the wavelength is so long that impedance mismatch doesn't matter so much?

So, to valves. I can certainly understand that a 400V swing at low current has to be transformed to a lower-voltage higher-current for a speaker-sized load. But for the schematics i've been browsing so far, many people just say 'add more parallel output tubes for lower output impedance', which i certainly understand. But (presuming i can get a spice program for linux and good models) how many is enough, and how low should the output impedance be for (eg) an 8ohm load? exactly 8ohms? lower? and what happens to the output power, distortion, frequency response, all of that, if i have a higher output impedance? (of course, i could just add x more tubes, but i'm on a limited budget here).


So, to the circuits. I've been googling OTL and circlotrons, and a fair few seem to favour EL509s, which i think i can source a few of. Are there any other suitable tubes with low output-impedance that can be used in OTL designs?

OK, it doesn't have to be a circlotron, but i certainly have the capability to make multiple isolated supplies with high Power Factor (and i'm obviously not one of those people who are simply allsolidstate=bad, anynegativefeedback=bad, addingharmonicdistortion=good, i believe in using the best of both worlds), and this amp is more an experiment than anything else.
Class A, AB, whatever sounds better and is cheaper and easier, i'm in Holland so i don't mind having a spae-heater in winter, i'll just turn down the central-heating whenever i listen to something if it's class A. Single/balanced inputs also don't matter, i can hack together a balanced-driver in an hour if i have to...

If anyone can help out with schematic suggestions from projects they've built, design tips, whatever, that would be much appreciated.

cheers
Doc C


ps, i also have the (dis?) advantage that I also don't have any speakers yet, on this side of the world. (my homemade dual 10" subs weigh a nice 44kg each, real 1"MDF in a 120L box, and 8" monitors are 15kg, so they're staying down under for now).
So i was looking for some very high efficiency fullrange, probably narrow-tower speakers, i'm in a small apartment so size and power handling can be fairly low, if i can get them 16 or 32 ohm that would certainly help. So if anyone can recommend any for <€200 a pair thanks for that too.
 
Even the lowest impedance tubes will have to be paralleled in the ~50 quantity to get anywhere near ideal load impedance. Even for 16 ohm speakers.

That said, if you don't mind that you are always operating in the current limited region of the amplifier's performance, and that the only way you will EVER get a usefully low output impedance is with lots of NFB, then go ahead, you can do a fine job (given that restriction). Use high current tubes (6AS7, sweeps, 6S33S, et al.), use grid current if necessary.

Tim
 
6as7g or 6c33c are your best choice as far as output tubes go.

In regards to output impedance, the 6c33c is the bigger of the two and has an internal impedance of about 80R, so figure it would take 10 in parallel to match an 8R load. Adding NFB will lower the distortion and output impedance (stabilize the output voltage in relation to load impedance changes) but it doesn't increase your power output or current delivering capability. That's because the internal I2R losses of the tube remain the same regardless of the amount of feedback. Most OTL designs which are intended to drive real world speakers use NFB, but if paired with the right speakers, you can get away without it.


I suggest class AB unless your planning to build high impedance or high sensitivity speakers. 1pair of 6c33c's per ch will output 20-25Wrms into 8R class AB. 4 6c33c's per channel can give 80-100Wrms into 8R class AB. If you notice, doubling the output tubes quadruples the potential output power. That's because current is primarily the limiting factor, not power.
 
Last edited:
Hi.

Tubes like 13GB5, 21KQ6/29KQ6 are similar to EL509/PL509 and are $1.00 each from thetubecenter.com . You can series the heaters and run it directly from the power line/mains. The problem with a circlotron (besides the drive circuitry) is that you need two isolated power supplys (iron transformers).
If yer gonna build an amp, built your own speakers too. Towers with 8 8-ohm drivers wired in series should work.

Jim
 
Last edited:
Hi,

Low-power speakers and OTL is the worst mismatch one could imagine. Think about it, you'll probably only swing 5% of the supply voltage into the load and still need a huge amount of tubes to get a reasonable damping factor. There is no valid reason for doing this (apart from novelty I suppose).

Since you're thinking of high efficiency speakers may I suggest something like EL 84 or 2A3 tube single-ended *with* output transformer. OPTs for these tubes can be found all over the place in all price categories, so your concern over availability and expense is unfounded. As a bonus such an amp would be far more reliable and consume far less heater power than any OTL.

If you build, say, a BIB with small fostex drivers for speakers and build a modest EL84 SE amp, you'll be all set for perhaps half the price of the OTL amp alone.

Kenneth
 
Hi,

Low-power speakers and OTL is the worst mismatch one could imagine. Think about it, you'll probably only swing 5% of the supply voltage into the load and still need a huge amount of tubes to get a reasonable damping factor. There is no valid reason for doing this (apart from novelty I suppose).

Since you're thinking of high efficiency speakers may I suggest something like EL 84 or 2A3 tube single-ended *with* output transformer. OPTs for these tubes can be found all over the place in all price categories, so your concern over availability and expense is unfounded. As a bonus such an amp would be far more reliable and consume far less heater power than any OTL.

If you build, say, a BIB with small fostex drivers for speakers and build a modest EL84 SE amp, you'll be all set for perhaps half the price of the OTL amp alone.

Kenneth

This is pretty well untrue. OTLs **shine** when driving high efficiency speakers! You don't need many tubes to do it either. Its easy to build a small OTL that will outshine any 2A3 amplifier made, for about the same cost.

originally posted by Sch3matic: Even the lowest impedance tubes will have to be paralleled in the ~50 quantity to get anywhere near ideal load impedance. Even for 16 ohm speakers.

That said, if you don't mind that you are always operating in the current limited region of the amplifier's performance, and that the only way you will EVER get a usefully low output impedance is with lots of NFB....
None of this is true either! 50 power tubes would get over 500 watts, with maximum power into less than 3 ohms. If you are driving 8 ohms you only need 8 6AS7Gs to make reasonable power -60 watts- and that is class A with zero feedback.

Take a look at the link to 'what tubes for a tube amp?' posted above.
 
So, to valves. I can certainly understand that a 400V swing at low current has to be transformed to a lower-voltage higher-current for a speaker-sized load. But for the schematics i've been browsing so far, many people just say 'add more parallel output tubes for lower output impedance', which i certainly understand. But (presuming i can get a spice program for linux and good models) how many is enough, and how low should the output impedance be for (eg) an 8ohm load? exactly 8ohms? lower? and what happens to the output power, distortion, frequency response, all of that, if i have a higher output impedance? (of course, i could just add x more tubes, but i'm on a limited budget here).

BTW:
I think the output impedance (Zout) of your amp mainly affects the Qes ( -> Qts) of your speakers, so if your amp end up with a highish Zout it's just a mater of designing the speakers accordingly. More or less like adding a serise resistor.

Your new Qes and Qts would be:

Qes = 2*Pi*fs*Mms*(Zout + Re) / (BL)^2
Qts = Qms*Qes / (Qms + Qes)

/Leif
 
It's obvious that some of you who answer this thread have no experience with OTLs. Please do not bable about OTLs then.

If you want to suggest another design, that is fine, but don't slander something you don't have any clue about.

There is no reason to believe OPT are unavailable and underperforming, there are enough choices for just about any tube out there. OTLs are unefficient, but can drive speakers just fine. If high dampening factor is all you aim for, go transistor, if sound quality is your thing, try just about any good design, being tube OTL, OPT-SE or PP, SS, hybrid, etc. All can be very good indeed.

My five cents goes towards that Circletron by Atmasphere if OTL is what you'd want to try.
 
If you are driving 8 ohms you only need 8 6AS7Gs to make reasonable power -60 watts- and that is class A with zero feedback.

I'm curious how you reach that output? Even idling the tubes at their max current rating (125mA per section x 16) should only yield 32Wpeak or 16W rms in to 8R before leaving class A.
 
I'm curious how you reach that output? Even idling the tubes at their max current rating (125mA per section x 16) should only yield 32Wpeak or 16W rms in to 8R before leaving class A.

Class A2 actually. When you put the amp on the bench and run an 8 ohm non-inductive load, then measure the waveform across the current limitiing resistors, you will find that the tube goes into cutoff only once the amp is clipping. Now a lot depends on power supply voltage and bias voltage, but that is how in practice it can work. Obviously you have to have a driver that is linear with some grid current.

What is not obvious about the 6AS7G is that is is pretty linear while being driven into grid current, so this is practical as long as you are careful with the tube. This is not something you can do with a 6C33 BTW.
 
The operative term in text you quoted was "low power speakers", not "high efficiency speakers", the notion being that efficiency (Pout for Pin) of such a setup would be extremely low. I don't see how this would be untrue.

Yes, that was a question mark for me but when I looked at the context in which it appeared, it seemed to be referring to high efficiency speakers.

'low power speakers' really does not make any sense otherwise- certainly this is not referring to a low efficiency speaker.

Regardless OTLs can drive either quite well- as in many situations, you need an amp with enough power for the job. OTLs come in all sizes, and while it is true that smaller ones are less efficient than bigger ones, they can do a much better job of driver various speakers than has been portrayed earlier on this thread. IOW I'm sticking to my story.

If it helps I've been building OTLs in one form or another for the last 38 years. I've seen them drive all sorts of speakers. At home I have a set of amps that have 2 tubes per channel- one 6SN7 and one 6AS7G. They play my 16 ohm speakers easily.

A lot depends on the intention of the speaker designer- whether they want it to be easy for tubes or not, for example.

see
Competing paradigms in amplifier and speaker design
for more about that.
 
Class A2 actually. When you put the amp on the bench and run an 8 ohm non-inductive load, then measure the waveform across the current limitiing resistors, you will find that the tube goes into cutoff only once the amp is clipping. Now a lot depends on power supply voltage and bias voltage, but that is how in practice it can work. Obviously you have to have a driver that is linear with some grid current.

What is not obvious about the 6AS7G is that is is pretty linear while being driven into grid current, so this is practical as long as you are careful with the tube. This is not something you can do with a 6C33 BTW.

I'm a great fan of OTL's and have built three myself, including a circlotron. I haven't heard the Atma-Spheres, though I would love to. Since the class A vs. class AB question has come up again, I would be really interested to know what is the counter-argument to the standard one, along the lines that Jeb-D was mentioning, that seems to imply that any realistic OTL amp is going to go into class AB mode once the power exceed a fairly low limit. (Essentially, that in class A mode the plate current can only go up to twice the idle current at the "top" of the sinusoid, if it is not to go to zero and then cut off down at the bottom of the curve.)

I hear Ralph's statement that experimentally, at least in the Atma-Sphere, the current though the tubes only cuts off at the bottom of the sinusoid if clipping has been reached (i.e. full power). But I still don't understand WHY this can happen, since it appears to contradict the standard argument that implies class AB beyond a rather lower power. Is the class A2 issue playing a role here? I can see how that means the maximum plate current can be bigger at the top of the sinusoid than might otherwise be achieved, but I don't see how it could help with preventing a cut-off down at the bottom of the sinusoid.

Can anyone shed light on this, with some mathematics of what it happening?

Thanks!
Chris
 
Well, my statements were pretty much based on class A1 vs. AB1. Under the assumption that the grid current demand would be too large for A2/AB2 to be practical without sand. I know that has been my experience with 6c33c. The typical driver tubes tend to spew their guts trying to drive the 6c33c's grid positive. This seems much easier to do with the 6as7g based on Ralph's comments.

Class AB consumes much less power on average, especially when music is the signal. The 6c33c class AB OTL's I've built run much cooler than my 6c33c SET, even though power output is roughly the same between the two. The 6C33C SET does a fine job as a space heater or marshmallow roaster. I could not imagine the amount of heat would be generated by having multiples in parallel running class-A.

I noticed the M-60 amp can switch between class-A or AB, which is a really interesting feature. Best of both worlds it seems. Winter mode and summer mode :D

I can see how that means the maximum plate current can be bigger at the top of the sinusoid than might otherwise be achieved, but I don't see how it could help with preventing a cut-off down at the bottom of the sinusoid.
I can see why you are puzzled. My only guess on how this works is that gm step between the positive and negative cycle is dis-proportionate to the extent where the saturation bound half is allowed to swing much more plate current than the cutoff bound half per given Vgk. By looking at the steep load line on the tubes curves this seems probable. This effect would create large amounts of even-order harmonics if it were an SE class-A, but since push pull cancels even-order it's no problem.
 
Last edited:
Well, my statements were pretty much based on class A1 vs. AB1. Under the assumption that the grid current demand would be too large for A2/AB2 to be practical without sand. I know that has been my experience with 6c33c. The typical driver tubes tend to spew their guts trying to drive the 6c33c's grid positive. This seems much easier to do with the 6as7g based on Ralph's comments.

Class AB consumes much less power on average, especially when music is the signal. The 6c33c class AB OTL's I've built run much cooler than my 6c33c SET, even though power output is roughly the same between the two. The 6C33C SET does a fine job as a space heater or marshmallow roaster. I could not imagine the amount of heat would be generated by having multiples in parallel running class-A.

I've seen comments in the past about how a class A OTL is not possible. That is a myth- like a transformer coupled amplifier, there is a set of parameters that you have to meet to get to class A. The issue is really one of the load, the lower the load impedance, the more likely the amp is not running class A, depending on power supply voltage, number of tubes, etc. For example the the M-60 runs class A2 into 8 ohms but is a class AB2 amp driving 4 ohms. BTW, half of the power generated when running into that load is dissipated in the output section. The amp will run cooler and draw less power if you have it drive 16 ohms.

Now the saturation of the 6C33 is such that it just does not appear practical at all to operate it in A2. The grid current is prodigious. Some people have thought that maybe if they have a beefier driver than the one used in the M-60 that they will get more power; if you go totally over the top with that idea it gets you maybe another watt. Even the 6AS7G will have significant grid current if you push it hard enough :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.