There are variants of each to be sure. For example there are Futterman amps and non-Futterman totem poles. In the Circlotron side there is the Atma-Sphere concept (zero feedback, one gain stage), there is the Cecil Hall amp which includes the driver in the Circlotron (at least one of the Graaf OTLs used this concept) and there is the Alan Kimmel amplifier.
Of the totem-poles, the circuit that appeals to me the most is
The OTL5998L amplifier
I like this one for its simplicity. One 6SN7 plus power tubes- not bad.
You can build a Circlotron amp with the driver section of the latter quite easily. I have also built Circlotron amps using a driver transformer to convert from single-ended to push pull, similar to this:
http://www.diyaudio.com/forums/tube...-amplifier-only-3-components-signal-path.html
bingo - need to add that I have yet to do a circ otl but its coming most likely a Graaf 20 copy
Interesting design this Graaf GM20. This is the kind of thing I'm looking for - designs showing feedback paths that I can expand on.
This is a re-draw of the amp that was posted by someone on another thread in this forum in 2009:
GM20 schematic.pdf - Windows Live
It's a 6C33C based circlotron. It looks like it has the potential to sound good. The reviews I read said it's stable on the DC bias and sounds great.
Based on what I see in simulations on PSPICE, the circlotron has more potential to create a low distortion amp with less NFB, although the output impedance is slightly higher than the totem pole. Then again, there's no reason a totem pole can't sound great.
This is a re-draw of the amp that was posted by someone on another thread in this forum in 2009:
GM20 schematic.pdf - Windows Live
It's a 6C33C based circlotron. It looks like it has the potential to sound good. The reviews I read said it's stable on the DC bias and sounds great.
Based on what I see in simulations on PSPICE, the circlotron has more potential to create a low distortion amp with less NFB, although the output impedance is slightly higher than the totem pole. Then again, there's no reason a totem pole can't sound great.
This is the schematic I had located as well. Will look it over soon to see if its accurite.
If you have a second source to check the accuracy of the schematic and you find any errors please let me know. I'll have as accurate a model as I can get a schematic for in PSPICE within a few days.
What kind of diodes are those? There are some undefined parts in the schmatic - I don't think there's a parts list.
Any help with the remainig items would be greatly appreciated.
Here's a pic of my amp staging area. Those are 24" x 24" x 1/8" aluminum plates for the top and bottom and the 8" x 3/4" oak will form the sides. The idea is to construct the amp in the same room where it will be used and never have to move it.
The red stuff is GPO-3. I bought 0.062" thick panels. The stuff is great for insulating parts from the chassis and its cheap. And of course the tubes, Russian 6H13C. They finally arrived, at least the batch of 25 from Ukraine at about $8 a pop, not bad for a 26 watt tube. Now if my Romanian tubes don't get destroyed by homeland security as mini-bombs I'll have more than enough output tubes. That's a roll of copper foil to shield the wood parts of the chassis and any other sensitive areas of the amp. The GPO-3 can also be cut to form small enclosures or shields inside the amp and covered with copper foil.
The power (isolation 1:1) trannys are in the styrofoam - one sits on top. Each is 100VA and they are close enough in tolerance that they can be paralleled to get as much power as I need for each supply.
An externally hosted image should be here but it was not working when we last tested it.
The power tranny needed for 150 - 160 volts DC when rectified is about 120 VAC (RMS), the voltage used in the US for household power. Some of the old totem pole amps simply used the line voltage and rectified it with no power transformer!
An isolation transformer is nothing more than a power tranny that doesn't change the voltage, they just provide a margin of safety that all power trannys provide. The great thing about them is that they are CHEAP since they are mass produced. You can also put them in series on the secondary side to obtain higher voltages.
I could have purchased a large 120 volt primary / 120 volt secondary power transformer. But wait - that's called an isolation transformer!!! Plus, they make them in bulk and dump surplus frequently on places like ebay.
An isolation transformer is nothing more than a power tranny that doesn't change the voltage, they just provide a margin of safety that all power trannys provide. The great thing about them is that they are CHEAP since they are mass produced. You can also put them in series on the secondary side to obtain higher voltages.
I could have purchased a large 120 volt primary / 120 volt secondary power transformer. But wait - that's called an isolation transformer!!! Plus, they make them in bulk and dump surplus frequently on places like ebay.
So the isolation transformers are 120 primary 120 secondary and you use 2 of them to get the desired sets of voltages (+150 -150)? Or they each have dual secondaries?
A wide variety are made. It is not uncommon to have two 120 volt primaries and two 120 volt secondaries. The windings can be connected to provide the voltages needed. A standard 120V circuit in the US has he primaries parallel connected. If you were to series connect the primaries, you could connect them to a 240 volt circuit. The advantage would be isolation from common noise on the premises common or "neutral" connection. Of course it would be a major pain to connect anywhere else and all of the transformers used in the amp would have to be 240 volt primary.
Look for dry type distribution transformers. Acme.
This is already understood but thx away.
I was wondering about the ones he was using since he got them on the Bay and I see many out there so I had a specific question for the ones he had retrieved.
Just wrote a long response and it got erased when I acidentally hit the wrong key while typing. WHY does Microsoft do that???
As the previous post indicated, they are very flexible. The only caveat I'd add is that I'm using them with the secondaries in parallel. You can ONLY do this as a general rule if the transformers are wound by machine and are from the same manufacturer and from the same production run. Otherwise chances are the voltages (or turns ratio) won't be exactly the same and there will be large circulating currents at no load and the load won't distribute equally when under load.
I bought 16 100VA trannys on ebay for $10 each ($0.10 per VA) and they all match so close that parallel operation is accepable. In my case I may parallel up to four for each of the four supplys in a stereo circlotron, or up to eight on each side (4 for the pos and 4 for the neg) of a split supply on each channel of a totem pole output amp. In either case I'll have 1600VA of capacity for $160. Try to find four - 400VA transformers of any voltage (or two - 800VA trannys) for $160 total. Good luck.
You can also put these in series (mix and match as long as the VA rating is the same - any manufacturer) for 120, 240, 360VAC supplies, etc.
Does anyone here know how the 6C33's failed in his amp? I would have first probably gone to Six - 6C33's per channel before going to the 6AS7. I built a pair of the Atmosphere M60 MK-2's way back when and was generally unimpressed with them. Atmosphere also made an OTL with Six - 6C33s per channel and it sounded incredible. I like his design and I would think the extra set of 6C33s in his amp would make for greater reliability than just the4.... I think he is asking a little much from Four 6C33 toobes!!
Also FWIW: The BAT VK-60 is easily the finest power amp I've ever owned over the years. It does 60 wpc and uses torodial output transformers. Absolutely amazing amp! I sold mine quite a while back but have been tempted to get another.
Mark
Hey Mark,
I think, and I may be wrong (you can look it up on thew "Wayback Machine) that Mr. Covi experienced repeated failures of 1) Sockets due to heater current; 2) tube shorts due to mismatching or perhaps because he didn't precondition per Ralph (I really don't know) and 3) grid failure due to excess grid current. From what I understand, the grids of these tubes were not intended to dissipate as much current as he was doing in his amps.
That said, I'm repeating what I've heard, not what I know. And my memory is way less than perfect. What was I talking about? Oh yeah, 6C33C failures.
So if Ralph is spot on, then some of the failures may be cured by the preconditioning. The socket failures may be helped by vigorously cleaning the tube pins and maybe coating with Silclear or some similar product. Your suggestion of two more tubes per channel could alleviate the grid problem.
I have 12 and I'll take Ralph's preconditioning advice, breadboard the whole thing before boxing it up and see what happens. If the outputs fail, then I'll have to start buying 6AS7Gs.
Stuart
Yea, I like the bread board idea alot on that one. Unfortunately there are not a lot of socket options available for that tube but I find it hard to believe they used those very same sockets in Airplanes. There must be some sort of higher end socket made in Russia for the 6C33!! I also wonder if something like Silver plating the contacts might help... It'll dissipate the heat much better then most other platings will. I never had any problems with the BAT VK-60 that I had in that regard... but there may well be people that have had problem!! I think I may give that amp a go using six 6C33s and see how it behaves. I might also try Tim's amp expanded out to 6 or 8 tubes as well but I just don't know if the driver stage could handle that many 6C33's or not.
BTW: I do need about 70 watts or so into 6 ohms... I do have relatively inefficient speakers.
Mark
BTW: I do need about 70 watts or so into 6 ohms... I do have relatively inefficient speakers.
Mark
This page shows 2 socket types for the 6C33S:
Tube sockets - Vacuum radio tubes, capacitors, socket for tubes, nixie, ferrites from military Russia. Datasheets tubes and help. Worldwide shipping
Tube sockets - Vacuum radio tubes, capacitors, socket for tubes, nixie, ferrites from military Russia. Datasheets tubes and help. Worldwide shipping
I read on another thread that the 6C33C tubes were originally sold in the USSR with new sockets. Apparently it was expected that at each tube change interval, the socket would also be replaced. No doubt that was a response to field failures.
I'm in the same boat as you, Mark, with inefficient speakers. Covi was using the OTL to drive his Maggies full range. I biamp mine, but still need ~80+ WPC on the midrange/tweeter panels, so I'm still looking at your 6 tube per channel solution.
A better driver than the 6922 that Covi used might be either a 6BQ5 or a Mosfet. The 6922 has some linearity issues and he actually is operating it outside of it's B+ limits. Need to be able to provide that needed grid current on peaks.
Stuart
I'm in the same boat as you, Mark, with inefficient speakers. Covi was using the OTL to drive his Maggies full range. I biamp mine, but still need ~80+ WPC on the midrange/tweeter panels, so I'm still looking at your 6 tube per channel solution.
A better driver than the 6922 that Covi used might be either a 6BQ5 or a Mosfet. The 6922 has some linearity issues and he actually is operating it outside of it's B+ limits. Need to be able to provide that needed grid current on peaks.
Stuart
http://cgim.audiogon.com/i/vs/i/f/1246841453.jpg
Find all you can about the Atma-Sphere's MA240 there is a guy on OTL Forum
driving electrostatic speakers with them. I believe it has 3 pairs of 6c33c for 120 watts in circlotron output. Each 6c33c has it own bias.
My goal is to build with a Diy m60 frontend (4 6Sn7) wih 2 pairs of 6C33c with individual bias for about 60 watts. Similar to Atma-Sphere's Novacron
You must also follow Mr.Karsten caveats about building with 6c33c.
I started this thread and thats my conclusion. Just have to figure out the individual bias scheme.
Darrell/dwhitf
The M240 in question is driving a set of Sound Labs.
You will need a cathode follower driver tube for each 6C33. This requires an separate coupling capacitor for each grid of each cathode follower. That is how the M-240 is set up ideally. There were some built that relied on matched triplets of 6C33s. This version had only one bias control and one DC Offset control, and as you can imagine, was harder to live with.
Before using any 6C33 they should be pre-conditioned for at least 3 days and nights by running the heaters with the plates, cathodes and grids tied to one side of the heater circuit. This will minimize arcing and double the life of the tubes. BTW this is a good idea for 6AS7Gs too!
After looking at those sockets that brown one just flat out looks crappy. The ceramic socket would hold up much better but who knows what the contacts are made of or the quality. The ultimate way to operate the 6C33's would be as is done in many transmitters... where each tube is enclosed in a glass chimney and air is pumped up from under the socket itself and past the toobe's pins and then up and out a chimney surrounding the tube. Interrestingly, many new transmitter toobes also have to be preconditioned before hi-z is applied or arc over is much more likely to occur. Or you couold do as we do in motion picture projectors with high temperature xenon lamps where we enclose them inside a plenum and force a good amount of air though that plenum and then exhaust it. I think both toobe and socket life would be greatly extended either way. I will have to give all this some more thought and do alot more investigation before I proceed.
Mark
Mark
Last edited:
I have been buying NOS Johnson 247 ceramic sockets off ebay
sometimes called RCA UT-106.
A .125 drill deck,fan on each socket and a heater mod like the Graff GM20 otl (thick runs of metal then
attach to heavy duty wire, metal is a heat sink i guess).
Sockets lowered so air can pass by,aluminum spacers to transfer some heat to the deck also.
Any anymore ideas?
Comments appreciated
Darrell/dwhitf