In looking at the EL156 data that el156 posted (silvino1.pdf) you can see that the EL156 can deliver 25W if the B+ is increased from 350V to 450V.
Since you are going to need new transformers any way, I would suggest the higher B+ as it will also result in lower THD. By going to the higher B+ you could get 50W out of two tubes in parallel.
Either way, the OPTs may be an issue. What OPTs are you planning on using? Do you have a local winder who can make them?
Since you are going to need new transformers any way, I would suggest the higher B+ as it will also result in lower THD. By going to the higher B+ you could get 50W out of two tubes in parallel.
Either way, the OPTs may be an issue. What OPTs are you planning on using? Do you have a local winder who can make them?
Look at my initial post, I plan using 420-450 Volt for +B. I'll draw the load line and see if 2 tubes will give me enough power. Should it come close to 50W, I'll rest there.
I'll look for OPT once the OP stage design will be finalized.
25W per tube in triode mode may be obtainable. I'm still fiddling with the triode curves. Both Telefunken and Shuguang data, including the silvino1.pdf, refer to this tube in pentode mode.
Joshua,
Check out this thread. http://www.diyaudio.com/forums/tubes-valves/167212-garter-cross-coupled-feedback.html
I ran into the same problem you are regarding paralleling tubes. Several people here helped me get to the root of the problem in paralleling tubes. The biggest thing is to define the situation you are working on accurately. To me the most important thing is to run equal quiescent currents through all tubes cathodes. It sounds like you want to avoid running high resistances in the cathode circuit because then you need very large caps to bypass them. That eliminates garter biasing, which is cross coupled cathode biasing.
I think you'll find that to get equal currents through all cathodes the easiest and best solution is to put BJT current mirrors in the cathode circuit of each tube. This solution will add small equivalent resistances in the cathode circuit but it won't be a lot. After all the current mirrors won't be making huge adjustments in current as long as each tube is operating normally and is within normal variation around an average for each cathode. You could still bypass them with an electrolytic but it won't be that big a deal, IMO. I think you might even be able to avoid bypassing them altogether since you are operating in SET mode. In that case at all signal levels each tube will be running equal currents that are matched based on one tube that is the "master" tube. This, it seems to me, will be the most elegant solution to your problem.
You will still need fixed biasing of tubes but you will only need one driver circuit driving all paralleled tubes. If you use a follower circuit of some kind it can bias all tubes directly without a coupling cap needed between stages. Actually a coupling cap that can produce blocking distortion in the final stage is a far worse problem than bypassing a cathode resistor, if its required.
A follower stage is also better because it is easier for a the low output impedance of a follower to drive the increased input capacitance of paralleled tubes. There are other people here that might be able to help you on the details of current mirrors. I've been reading The Art of Electronics and they have a good section on current mirrors when there are more than 2 devices that need to be mirrored.
Check out this thread. http://www.diyaudio.com/forums/tubes-valves/167212-garter-cross-coupled-feedback.html
I ran into the same problem you are regarding paralleling tubes. Several people here helped me get to the root of the problem in paralleling tubes. The biggest thing is to define the situation you are working on accurately. To me the most important thing is to run equal quiescent currents through all tubes cathodes. It sounds like you want to avoid running high resistances in the cathode circuit because then you need very large caps to bypass them. That eliminates garter biasing, which is cross coupled cathode biasing.
I think you'll find that to get equal currents through all cathodes the easiest and best solution is to put BJT current mirrors in the cathode circuit of each tube. This solution will add small equivalent resistances in the cathode circuit but it won't be a lot. After all the current mirrors won't be making huge adjustments in current as long as each tube is operating normally and is within normal variation around an average for each cathode. You could still bypass them with an electrolytic but it won't be that big a deal, IMO. I think you might even be able to avoid bypassing them altogether since you are operating in SET mode. In that case at all signal levels each tube will be running equal currents that are matched based on one tube that is the "master" tube. This, it seems to me, will be the most elegant solution to your problem.
You will still need fixed biasing of tubes but you will only need one driver circuit driving all paralleled tubes. If you use a follower circuit of some kind it can bias all tubes directly without a coupling cap needed between stages. Actually a coupling cap that can produce blocking distortion in the final stage is a far worse problem than bypassing a cathode resistor, if its required.
A follower stage is also better because it is easier for a the low output impedance of a follower to drive the increased input capacitance of paralleled tubes. There are other people here that might be able to help you on the details of current mirrors. I've been reading The Art of Electronics and they have a good section on current mirrors when there are more than 2 devices that need to be mirrored.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.