An externally hosted image should be here but it was not working when we last tested it.
Asides from the 60J in the coupling capacitor and a bleeder to ground, am I missing anything? Have I not seen something like this before because its crazy, or is there something wrong with it? Do I need a de Qing diode, and could the choke be omitted if it were present? Speaker is 144ohm midbass line array. Its not truly direct drive, but there is no output iron.
Circuit is largely derived from tubelab.com's 845/833A set.
I can't quite read your values, but I don't know why you have a resistor in series with the speaker? The parafeed capacitor will need to be quite large to get any bass. The supply voltage on the 833A will need to be MUCH lower than I use, a few hundred volts at most. The parafeed choke will need to be capable of handling at least the quiescent current of the tube. It should be higher, and you will need a fair amount of inductance to get good bass. I have some military surplus chokes that I am saving for a circuit like this, although I was thinking parallelled 6336's. The chokes are 2 henrys at 1 amp. they are about 10 inches cubed and weigh about 70 pounds each. This is going to be a BIG inefficient amp.
I think that you want about 400 to 500 volts on the plate. Run the tube at max current, 500 mA or higher. This will require +50 volts of grid bias. You will get 20 to 30 watts of output power. Total input power will be 400 to 500 watts. Welcome to the world of OTL.
If you build it let us know how it works.
I think that you want about 400 to 500 volts on the plate. Run the tube at max current, 500 mA or higher. This will require +50 volts of grid bias. You will get 20 to 30 watts of output power. Total input power will be 400 to 500 watts. Welcome to the world of OTL.
If you build it let us know how it works.
It looks like a very inefficient and expensive design.
The max current of a 833A is ~500mA which means that the maximum power in 144 ohm would be ~18W but the efficiency would be very low as you would need a very high anode voltage.
The 500 ohm resistor in series with the output doesn't help either as you will loose quite a lot of power there.
It would be much better to use another tube for instance a 6C33C that has lower output impedance and doesn't need that high current, it also have much lower drive requirements.
The choke can be eliminated by using an active current source but a better alternative in my opinion is to build a SEPP type OTL where you have 2 tubes connected in Single Ended Push-Pull, this gives higher efficiency and higher output power.
Regards Hans
Resistor changes where bass roll off occurs, its there to raise speaker q and flatten impedance as much as anything else. Figured Id be current limited even with the resistors. Going below 800V I could actually use 92 or 184uF of capacitance, I have a stock of oil powerfactor correction caps. I dont have any >500ma choke though, they are real hard to find, I keep getting outbid. Id probably have to cut up transformers to make chokes like that. Id might cut up a 480V to 240V transformer as a choke, those seem plentiful. You dont think Ill get much power even with 18 speakers in series?
Though I was offered a 1500 pound (dry!) 2H choke, too bad I dont have a trailer. This concept came about largely because I have most of the parts on hand from other projects, I think I could find the choke too. Enlarged and revised schematic:
Though I was offered a 1500 pound (dry!) 2H choke, too bad I dont have a trailer. This concept came about largely because I have most of the parts on hand from other projects, I think I could find the choke too. Enlarged and revised schematic:
An externally hosted image should be here but it was not working when we last tested it.
"It would be much better to use another tube for instance a 6C33C that has lower output impedance and doesn't need that high current"
The current requirement is as you pointed out a requirement to generate power into the speakers. In an OTL there is no transformer to turn voltage swing into current swing. A 144 ohm speaker helps greatly, but 1/2 amp of average current translates to about 20 watts. To get more power you must have more cutrrent. Since a SE amp must run class A the peak speaker current will never be more than twice the tube's quiescent current. The plate voltage must be high enough to force this current into the required load, with some headroom. In this case the plate voltage would only need to be 250 volts or so, but the 833A is pretty non - linear at this voltage, that is why I suggested 400 or more.
The resistor in series with the speaker will suck up 3/4 of the power generated by the tube. It would be far better to find a good 600 to 8 ohm transformer. I am using UTC LS-33's that I got on ebay.
As for the chokeless parafeed or the SEPP output stage, the efficiency of these is also poor, because the power supply voltage must be almost doubled to provide the positive going swing that was provided by the inductance of the choke. In this case it might be a wash though. This is a path that is being explored on another thread, athough OTL has not been considered yet. As this progresses I plan to build a chokeless parafeed with an 845 on the bottom and an 813 on the top. Hey I like BIG amps.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=67437
The current requirement is as you pointed out a requirement to generate power into the speakers. In an OTL there is no transformer to turn voltage swing into current swing. A 144 ohm speaker helps greatly, but 1/2 amp of average current translates to about 20 watts. To get more power you must have more cutrrent. Since a SE amp must run class A the peak speaker current will never be more than twice the tube's quiescent current. The plate voltage must be high enough to force this current into the required load, with some headroom. In this case the plate voltage would only need to be 250 volts or so, but the 833A is pretty non - linear at this voltage, that is why I suggested 400 or more.
The resistor in series with the speaker will suck up 3/4 of the power generated by the tube. It would be far better to find a good 600 to 8 ohm transformer. I am using UTC LS-33's that I got on ebay.
As for the chokeless parafeed or the SEPP output stage, the efficiency of these is also poor, because the power supply voltage must be almost doubled to provide the positive going swing that was provided by the inductance of the choke. In this case it might be a wash though. This is a path that is being explored on another thread, athough OTL has not been considered yet. As this progresses I plan to build a chokeless parafeed with an 845 on the bottom and an 813 on the top. Hey I like BIG amps.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=67437
Sorry, this should of course read "It would be much better to use another tube for instance a 6C33C that has lower output impedance and doesn't need that high voltage"
Yes, agreed but the advantage is that you can run in AB1 which allow for much higher output power, a 6C33C can easily give a peak current of 2.5A while still providing very long life, Note! this is achieved without even drawing any grid current, i.e class AB1 so no special driver is necessary.
2.5A peak current give very useful power levels even with 8 ohm speakers. Another advantage with 6C33C is, (as was intending to say earlier) that you don't need that high voltage to give useful output power, +/- 150V is enough to give 25W in 8 ohm.
A 833A tube doesn't seem like a good choice for an OTL however you build it, there are better tubes, for instance the 6336 you mentioned earlier, (even though 6C33C is still much better).
Regards Hans
On the resistors, if my tube is dissipating 350 watts, does wasting 60w+ here really matter much? Being that the tube is in class A, is it not better to work the tube harder, even if your throwing most of it away? Part of why I liked this concept was to give somewhat of a current source amp though this is a rather inefficient way of going about it. It is definitely a beast that I would only run in wintertime.
I mentioned 6336's because I have some. I don't have any 6C33's although they may be a better, and more generally available choice. They are certainly more popular in the OTL world.
I think that any OTL is by definition inefficient. Puting a resistor in series with the speaker is highly inefficient, since you have the losses associated with generating a high power output, and the losses associated with losing 3/4 of it on the way to the speaker.
If the tube is operated at a high enough voltage to cover the losses, then you should be able to put the same current into the speaker. Many people strive to make amplifiers that are good voltage sources, because most speakers are designed to be driven by a voltage source. If you believe that your speakers would benefit from a current source, then try it. Be sure to let us know what happens. We might all learn something. There have been some recent interest in current source drive. It has been stated that this works with full range (single driver) speaker systems, or requires a re design of the crossover network.
I would plan to try different resistors though.
I think that any OTL is by definition inefficient. Puting a resistor in series with the speaker is highly inefficient, since you have the losses associated with generating a high power output, and the losses associated with losing 3/4 of it on the way to the speaker.
If the tube is operated at a high enough voltage to cover the losses, then you should be able to put the same current into the speaker. Many people strive to make amplifiers that are good voltage sources, because most speakers are designed to be driven by a voltage source. If you believe that your speakers would benefit from a current source, then try it. Be sure to let us know what happens. We might all learn something. There have been some recent interest in current source drive. It has been stated that this works with full range (single driver) speaker systems, or requires a re design of the crossover network.
I would plan to try different resistors though.
I plan to use these for an OTL. I have 6 or 8 of these, I don't have a large quantity, and I don't know the condition of them. So I will hang on to them until the amp is done.
6C33C-B
Hi Tubelab-
Im currently working on a design for an OTL with the 6C33C-B based on a Circlotron with 100% feedback round the O/P stages and with your Power-Drive scheme driving them, as I am so impressed with the way the PD deals with the 811A!. (Ill post the scheme when its finallised and in a workable condition..........)
I have more 6C33c-b than I really need, and once Ive got the thing running, Ill have a couple left over, and they are fairly cheap in bulk off Fleabay.........
When Ive finished my design and prototype bread-board, I would value your input on it, maybe in exchange for a couple of these Russian beasts--Ill try and pick some fairly well matched for your OTL experiments, They say they are more robust than the 6336 type tubes...........
Hi Tubelab-
Im currently working on a design for an OTL with the 6C33C-B based on a Circlotron with 100% feedback round the O/P stages and with your Power-Drive scheme driving them, as I am so impressed with the way the PD deals with the 811A!. (Ill post the scheme when its finallised and in a workable condition..........)
I have more 6C33c-b than I really need, and once Ive got the thing running, Ill have a couple left over, and they are fairly cheap in bulk off Fleabay.........
When Ive finished my design and prototype bread-board, I would value your input on it, maybe in exchange for a couple of these Russian beasts--Ill try and pick some fairly well matched for your OTL experiments, They say they are more robust than the 6336 type tubes...........
- Status
- Not open for further replies.