Not quite my experience. My 4 x 6c33 pr. channel are producing > 65W robust watts pr. channel and has been working perfectly without any hickups, not even a tube change for several years now and is the best sounding amp. of the very many I have made (and commercial amps too). The nearest competitors were a Goldmund Telos clone and My Fremen Ref. But the mighty OTL has some authority and natural timbre and a transparancy that I never have heard from a transformer coupled tube amp.
Attachments
This is complete nonsense. If it were true I'm pretty sure I'd not have been able to make a business of the last 44 years
Our smallest amp uses five 6AS7Gs per channel and makes 30 watts into 8 ohms. Just as a FWIW, most of the actual statements about OTLs in the trolling article linked (to which is referred above) are false- its pretty clear this individual hasn't got direct experience. This article has been debunked before. But let's explore a statement here:
[/SIZE]
[SIZE=+1]This statement is so false as to be entirely apocryphal. There is nothing anywhere to suggest something special about low impedance triodes like the 6AS7G that they need 40dB of feedback! Independent measurements of our amps have shown a THD of our earlier gear at about 3%THD which is not bad for a zero feedback ampilier and between 0.5% and 1% with our newer stuff (depending on how nicely the tubes are set up).
[/SIZE]
[SIZE=+1]
[/SIZE]
[SIZE=+1]Now I can show were the errors are in other parts of this article; it might be a good topic for another thread.
[/SIZE]
One thing about 7241s (other than their high price and extreme rarity) is that they need a great deal of care and feeding. The preconditioning must be at least 5 days and nights. There must be sufficient cathode current limiting resistors for each cathode. I think running in excess of about 135volts on the plate to be foolhardy. But if you are patient with them, I have seen them give good service. A CF driver is recommended, with a separate coupling cap and CF driver (with attendant biasing networks) will be needed.
In a nutshell, the 6AS7G is far easier to use and far easier to live with!! Plus, the sound you get is the same.
I was finally able to get Ralph out of the cave, I think brother, that you should watch what is said about this from the OTL more closely. We know that you are the Guru of the Circlotron adapted to the valves in this configuration. It is sure that after more than forty years of producing the prestigious brand of amplifiers that I dreamed of from the first day I knew they existed. Valves, and without transformers, since I lived in Argentina and read the Glass Audio article in the year 92 by a certain Clive Lock, I fell in love with that circuit which later you saw as a copy of one of yours. He had a servo at the exit to protect the class of work. It is true that what you do is wonderful and I cannot even discuss it, a person with your background is indisputable and I understand that you feel annoyed by my comments. These comments have originated in my humble experience and tried from my capacity of explanation to give it a scientific foundation. In short, audio is pure physics. The problem is that it is the means to impress the sense of hearing, and here a number of variables originated by the psychic appear. Regarding my experience, I have tried to be as objective as I can, and after living more than 60 years, I may draw useful conclusions. The first is that the key to a Hifi chain is the transducers. In this I can tell you that I have tried a number of speakers, Lowther, Feastrex, hornloaded, OB, Dynamic, multipath, bi-amplified, tri-amplified, I have consistently tried to change crossovers in frequency and slope. Each one has a different taste, this is where you can verify that a well-matched amplifier in impedances to a speaker is quite indifferent. And here they come to carve the amplifiers, obviously I have never reached an Atma in my life, in another time if I had had the money I would have bought it, I would also have bought an original Firstwatt instead of making a clone, but today, with the aplomb that time grants us, It is possible that you invest more time in listening to the disco that I have and stop trying to improve the unbeatable. Regarding your amps which have been the ideal compromise of what an OTL fan is looking for. I think of all the valves the best is the Russian 6H13. Cheap, linear and there are many. It is stable and does not require much polarization correction. I know there are others who have an ideal figure, more mu, less plate resistance, but this is the ideal compromise. Now, regarding your circuit, it is a circlotron, cathodic follower, and the soul of its amplification is the driver in which you have a series of three 6SN7 in Cascodo that excite a follower with 6SN7 to be able to work in class A2 or AB2 with G1 current at the output stage. However, if we analyze the transfer curve of the triodes, despite having four in the driver and eight in the output, I do not have any triode voltage amplification curve. Ergo and with the same vision, it is possible to make a follower with SIT or with Lateral Mosfet, which can feel much better adapted to the impedance of the speakers. And for much less complexity and care to have the same or better efficiency and I would even say with triode curves and better linearity. I did not want to offend you dear Ralph, on the contrary, I knew that if I made an acid comment I would bring you back to the thread so that you defended your design with more passion but without less scientism. So far it is my humble opinion and I await a response that will enrich the forum.
It is basically a Atmasphere m60 input but with drivers for each 6c33 so it resembles , I guess, the Novacron.
And the speakers are a IDS 25 clone and the speaker impedance is nominally 8 Ohms 5 parallel groups of 5 series coupled drivers.I f I were to put them all in series I would have to redesign the amp as it is not optimized for such a high impedance, and then I could not use any other amp for driving the speakers.
atmasphere otl amps sound great, if you heard you know, the circlotron design seems to have a lot of upside...
I am trying to get my head around the m60 schematic you posted much earlier in this thread. The power section looks like a bit of a lego tube goodness basically 2 tubes, 11 resistors and a cap per ‘module’, the driver section is a little more confusing.
Is V4 (which is a 6sn7 running power section voltage) part of the power section or the preamp section?
Why is V4 being used ‘backwards’ with the signal going to the cathode and the power section driven off the B- side.
Can I use the power section like legos and replace the gain section with a simple 12au7 or 12ax7 preamp section or even some blasphemy like putting a BA-3 gain section on the front end (somehow the BA-3 front section driving otl ‘modules’ sounds good to me, I’m going’ to tube helll for sure). If this is possible what happens to V4?
Btw thanks to everyone for IDing the tubes in the Aitos amp
I am trying to get my head around the m60 schematic you posted much earlier in this thread. The power section looks like a bit of a lego tube goodness basically 2 tubes, 11 resistors and a cap per ‘module’, the driver section is a little more confusing.
Is V4 (which is a 6sn7 running power section voltage) part of the power section or the preamp section?
Why is V4 being used ‘backwards’ with the signal going to the cathode and the power section driven off the B- side.
Can I use the power section like legos and replace the gain section with a simple 12au7 or 12ax7 preamp section or even some blasphemy like putting a BA-3 gain section on the front end (somehow the BA-3 front section driving otl ‘modules’ sounds good to me, I’m going’ to tube helll for sure). If this is possible what happens to V4?
Btw thanks to everyone for IDing the tubes in the Aitos amp
This is about the only sentence in that post that made sense to me.
V4 is a cathode follower. It takes the high-impedance output of the input stage and converts it to a low-impedance output to drive the output tubes.
This is complete nonsense. If it were true I'm pretty sure I'd not have been able to make a business of the last 44 years
But 5 6AS7G tubes could push 100W into 8R using a transformer, no? Plus a transformer makes a great speaker protector
Last edited:
Thanks for the advice, I really like this circuit design and I think I am gonna build it as soon as I have a better understanding and get some idea how I will power it. The power section seems so modular by using the ba-3 gain stage I can see ways I might power this with a single commercially available transformer.
Thanks for the knowledge, that makes sense. While I am asking basic questions, I think V3 is being used to add gain in a place where a simpler circuit would have a resistor/bypass cap. What is the advantage of using a tube here? Amount of gain? Sound quality? Something else?
V3 looks like a current source that allows V1 and V2 to operate in a more linear way.
V1 and V2 form two cascodes, which increases the gain of the input stage. The high gain allows using (global) feedback, which brings the gain back down to the right level, increases the linearity of the whole amp, and reduces the output resistance.
The two cascodes (V1 and V2) form a long tailed pair (LTP), which is used to get nicely symmetric signals for the driver/output stage.
To learn about cascodes, LTPs and cathode followers, take a look at the valvewizard website.
As per advice from Audiowise, I measured the cathode voltage of each tube in the input
section.
The bottom tube (V3) is being fed over -300V and the cathode to heater ratings for the
12SN7GTA was exceeded. It was the only tube out of spec.
My HV transformer has a spare 6V heater winding. I made a couple pcb's that
contains a voltage doubler, dropping resistor and voltage divider.
The heater now sits at -130V well below the -200V max voltage rating.
Thanks Audiowize!
section.
The bottom tube (V3) is being fed over -300V and the cathode to heater ratings for the
12SN7GTA was exceeded. It was the only tube out of spec.
My HV transformer has a spare 6V heater winding. I made a couple pcb's that
contains a voltage doubler, dropping resistor and voltage divider.
The heater now sits at -130V well below the -200V max voltage rating.
Thanks Audiowize!
Come in guys, it should be other tube choice beside 6c33, 7241, 6sn7 ... you mention above.
There are thousands tubes out there, germany, usa, russion, japanese made.
What is other choice that has low Rp less than 100ohm ??
There should be some ... out there ...
I wish there is a Tubes database I can search by “Rp”
There are thousands tubes out there, germany, usa, russion, japanese made.
What is other choice that has low Rp less than 100ohm ??
There should be some ... out there ...
I wish there is a Tubes database I can search by “Rp”
But 5 6AS7G tubes could push 100W into 8R using a transformer, no? Plus a transformer makes a great speaker protector
One 6AS7G is good for about 10 watts, so no, couldn't do that. If the amps had a tendency to damage speakers, we'd have been out of business a long time ago! If the S-30 is driving a 16 ohm load it can make 45 watts, so its doing pretty well in that regard- that is the sort of load for which it was intended- a lot of older horn systems are 16 ohms.
The only time a speaker might be at risk is if it can't handle the power of the amplifier. So usually this means full range high efficiency drivers, which might handle 20 watts on a good day. For such speakers I recommend a fuse. Normally the B+ fuse of the amplifier (in the case of the M-60, 3Amps slow blow) is sufficient to protect the speaker, as it will blow if a tube shorts.
This is one of the reasons we run separate power transformers. But its not the only one; we also found out quickly that most transformer vendors had serious problems getting everything right (including but not limited to noise and reliability) when presented with multiple secondaries!
One thing that I find odd is the push for lower and lower speaker impedances. Its a simple fact that regardless of the amplifier technology, it will make more distortion into a 4 ohm load as opposed to 8 or better yet 16 ohms. You can see this in the specs of any amplifier! Its audible too- as distortion in any form is converted to tonality by the ear, and usually the result of using a lower impedance load on any amp is greater harshness on account of the higher ordered harmonics thus created.
So if you really want high performance audio reproduction as opposed to maybe (if you have a solid state amp) maybe 3dB more sound pressure, then your amplifier investment dollar will be better served by a speaker of higher impedance.
A fuse is surely a good idea. However, I find it odd that people don't seem to use any DC protection relays at the speaker terminals. Most seriously engineered solid state amps have one, so why not OTL tube amps?
That's where the money is. The world is dominated by solid state amps (class D and chip amps these days). These amps will happily output loads of current when they have to go loud, but not as happy with driving to higher voltages. That's why modern speaker drivers tend to be low impedance designs.
Money, as opposed to the push to make sound better. That's the classic deal, which has been with audio for a very long time.
Relays aren't needed; unlike solid state amps, tube amps do need tubes replaced on occasion. a fuse failure tells the operator the amp needs attention; its not that weird for a power tube to fail (hence the use of sockets). A solid state amp that had frequent output section failure would not be a viable product! So if a B+ fuse fails, its common sense to suspect an output tube has had a failure. Fortunately they are robust enough that for the last 44 years we've been able to place a 1 year warranty on the power tubes and its not put us out of business.
Hi, i've just finished one Chanel of the m60, unfortunately the 6N13S tubes I bought from aboard return back to seller because there were no flights to Greece ☹️
I decide to give a try with 6080 from GE until the things go well and I will be able to buy some 6N13S.
I make exact the same schematic as atmosphere post here. Except A-AB switch and feedback switch.
Power supply voltages:
For 6sn7 B1+- 310VDC, B2+- 300VDC
For 6as7 2x B+ 135VDC
I let the tubes burn for 24 Hours then I set the bias for 480 mA per rail. I didn't use any A-AB switch as per schematic
I set the trim for zero DC offset
I plug my signal generator, with 1kHz sine wave before clipping i measure 17V at 8ohm load about 36W
My first listening test wasn't so impressive because the sound is little harsh
What may couse this harsh sound?
The voltage for 6sn7 is ok or high?
6080 tubes?
Need some feedback?
My speakers are 8ohms, I also tried different speakers but same.
For 6sn7 I used the Russian equivalent
I decide to give a try with 6080 from GE until the things go well and I will be able to buy some 6N13S.
I make exact the same schematic as atmosphere post here. Except A-AB switch and feedback switch.
Power supply voltages:
For 6sn7 B1+- 310VDC, B2+- 300VDC
For 6as7 2x B+ 135VDC
I let the tubes burn for 24 Hours then I set the bias for 480 mA per rail. I didn't use any A-AB switch as per schematic
I set the trim for zero DC offset
I plug my signal generator, with 1kHz sine wave before clipping i measure 17V at 8ohm load about 36W
My first listening test wasn't so impressive because the sound is little harsh
What may couse this harsh sound?
The voltage for 6sn7 is ok or high?
6080 tubes?
Need some feedback?
My speakers are 8ohms, I also tried different speakers but same.
For 6sn7 I used the Russian equivalent
Last edited: