Just thinking. If you have say 4xKT88 in a push pull output stage could you just turn 2 off (by removing the heater supply) for low volume. I know it will affect the output impedance etc. From data sheets it appears a valve may actually tolerate more voltage when cold.
The volume won't change until you're at high power where it will begin to be dominated by distortion. The Zout would be 1/2 of normal WRT to the tubes that are still functioning.
You could just unplug two of the tubes and connect the 4 ohm speaker to the 8 ohm tap or the 8 ohm speaker to the 16 own tap to reflect the correct impedance. If you get creative with some 4PDT relays, you could do it all with a switch...
I've done it on my monoblocs, but I use 1k3 for output and one pair of 6P45S can work into 1k3 (triode) almost as well as two (80W vs 110W).
To save power, you might bias the output stage colder than usual, too. I once used 10mA on EL34. Except for the bouncy B+ rail (no regulation) it worked well enough.
You could just unplug two of the tubes and connect the 4 ohm speaker to the 8 ohm tap or the 8 ohm speaker to the 16 own tap to reflect the correct impedance. If you get creative with some 4PDT relays, you could do it all with a switch...
I've done it on my monoblocs, but I use 1k3 for output and one pair of 6P45S can work into 1k3 (triode) almost as well as two (80W vs 110W).
To save power, you might bias the output stage colder than usual, too. I once used 10mA on EL34. Except for the bouncy B+ rail (no regulation) it worked well enough.
Last edited:
I am contemplating something similar for the "big one" a tube amp with about 500 WPC. I'm still waffling between two or three pair of output tubes per channel, but the ability to run on one pair for most of the time will be available.
Relays will be used to select the best OPT tap for the chosen mode of operation, and the ability to shut down pairs of tubes as needed by bias and / or heater power will be provided. There will be several possible operating modes based on power, and "sound profile" needs. All of this as well as supervisory, safety, and even tube testing in the amp will be under microprocessor control.
A tube amp can run away if a tube gets gassy or something goes wrong. Things can go wrong very quickly in a 1 KW tube amp, so monitoring of all the "vital signs" is necessary.
One thing you don't want to do is to shut the heater off on a set of tubes while the amp is up and running at power. If you are still trying to get full power from a tube when it's cathode is cooling a tube arc can occur.
Ideally tube heater turn on - off should occur without current flow, or at least at idle. I plan to cut the tubes off by bias or screen voltage before changing the heater supply.
Tubes can withstand considerably more voltage when cold than when hot, provided that they are not gassy.
Relays will be used to select the best OPT tap for the chosen mode of operation, and the ability to shut down pairs of tubes as needed by bias and / or heater power will be provided. There will be several possible operating modes based on power, and "sound profile" needs. All of this as well as supervisory, safety, and even tube testing in the amp will be under microprocessor control.
A tube amp can run away if a tube gets gassy or something goes wrong. Things can go wrong very quickly in a 1 KW tube amp, so monitoring of all the "vital signs" is necessary.
One thing you don't want to do is to shut the heater off on a set of tubes while the amp is up and running at power. If you are still trying to get full power from a tube when it's cathode is cooling a tube arc can occur.
Ideally tube heater turn on - off should occur without current flow, or at least at idle. I plan to cut the tubes off by bias or screen voltage before changing the heater supply.
Tubes can withstand considerably more voltage when cold than when hot, provided that they are not gassy.
I am just reading this. As the cathode cools hot spots can be created when current flows to the anode, causing this part of cathode to remain hot, channeling the current into a smaller and hotter area. This can damage the cathode. So yes the grid needs to be pulled negative before heater is removed.
A high vacuum is normally an excellent electrical insulator. However, if the voltage is high enough between two electrodes in a high vacuum, electrons can literally be pulled out of the negatively-polarized electrode (called the cathode) in a process called field emission. Field emission typically occurs as explosion-like bursts as large numbers of electrons are ejected from microscopic projections on the cathode. These sudden flows of electrons cause localized heating and evaporation of the surrounding cathode material, forming small glowing spots on the cathode ("cathode spots"). The newly-freed electrons are then accelerated by the electric field between the electrodes until they slam into the positive electrode (anode).
A high vacuum is normally an excellent electrical insulator. However, if the voltage is high enough between two electrodes in a high vacuum, electrons can literally be pulled out of the negatively-polarized electrode (called the cathode) in a process called field emission. Field emission typically occurs as explosion-like bursts as large numbers of electrons are ejected from microscopic projections on the cathode. These sudden flows of electrons cause localized heating and evaporation of the surrounding cathode material, forming small glowing spots on the cathode ("cathode spots"). The newly-freed electrons are then accelerated by the electric field between the electrodes until they slam into the positive electrode (anode).
If those KT88 run as pentodes you could cut the G2 voltage ( plate current will drop to zero ) and the heater .
Anyway with some relays and a switch you can do anything , including full negative bias and heater off , or multiple other possibilities .
To be volume dependent is doable too ...
Anyway with some relays and a switch you can do anything , including full negative bias and heater off , or multiple other possibilities .
To be volume dependent is doable too ...
Better to power the OP valve heaters with a SMPSU if possible. Shutting down a pair of OP valves - ish was done on the extended Class A amp, there are four OP valves per Ch, where one pair are biased more or less to cut off, the other pair in Class A. As more power OP is needed the biased off pair are biased on, so, best of both worlds.
Andy.
Andy.
Umm interesting and does not change the impedance as much. Looking at simulation the higher bias valve should have a slightly smaller signal on the grid so that it does not clip first for large signals when the grid voltage gets to 0V or just above.
Last edited:
I think its the other way round ...
1/2 the number of tubes = 2x Zout
to make up for this the 8 ohm spkr should go to the 4 ohm tap then
may not be necessary at all if the amp has feedback ....
ON/OFF is a safe , obvious option and more "eco" ... who knows what effect on the life of tubes and sound have a full conducting tube in parallel with one less biased 😀
Well you have to stay within the tubes rating. And yes the one's on all the time/higher power dissipation will have a shorter life. As for distortion that's something that can be analyzed and compared with both having the same conduction. It could go the other way - you could actually save on the tube costs as well as the electric.
Last edited:
They did that in Fender guitar amplifiers. I don't recommend it as one pair will become unmatched over time. Not what you want!
I guess the basic question is .. do you really need a higher power tube amplifier? Build / use a smaller one. I guess one possibility might be to reduce the B+ for a lower power mode. Retain the outputs in circuit so they age together and more slowly compared to resting it all on one pair.
Hey George, 500 watts output power!! Holy crap man! I guess if anyone can do it, you can. I don't want to have to lift it though!
-Chris
I guess the basic question is .. do you really need a higher power tube amplifier? Build / use a smaller one. I guess one possibility might be to reduce the B+ for a lower power mode. Retain the outputs in circuit so they age together and more slowly compared to resting it all on one pair.
Hey George, 500 watts output power!! Holy crap man! I guess if anyone can do it, you can. I don't want to have to lift it though!
-Chris
I turn 69 years old in a few days. I need to build the amplifier in such a way that I can still move it several years down the road. Do I need a 500 WPC amp? No, but I have had the OPT's in my collection for a dozen or more years, so I need to build it while I can.
To minimize the mismatch in wear on the output tubes I plan to use a different pair every time the amp is powered up.
Rotating tubes like tires, on every power up? I guess you’re going to have some servo auto bias circuit?
He’s going to cheat and use a switch mode power supply (or several in series).
Fixed bias monitored and adjusted as needed by a microcontroller. The output tubes will always stay paired. Different pairs will be powered as needed. I have already seen one pair of tubes crank out 250 watts for about 10 minutes without glowing too much. Two pairs should suffice in any sane operating scenario, but I may use three if I can find a suitable chassis and packaging plan. The Plitron toroidal OPT's are rated for "400 watts @ 20 Hz." They will eat 500 watts of typical music. Again, the microcontroller will watch the screen and cathode currents and turn things down if needed.
Yes, 13 X 48 volt 3 amp supplies in series. This gives the ability to change the plate voltage in 48 volt steps to suit the intended power requirements. Why turn 625 volts of plate voltage into heat when you only need 50 WPC?
NOW I see what’s taking so long on this project - not just a “500 watt per channel amp”, but a computer controlled completely scaleable one. Different animal altogether. I’ve seen my own projects stalled for years because the required supervisor circuit ran into snags and took more time than anticipated to develop. This is going to be one for the record books.
If you are going to put in tubes on demand, I’d find a way to use 3 pairs. I like the idea of the scaleable voltage - solves the impedance mismatch issues when yanking out tubes. It can also compensate speaker impedance, but I’ll bet your output toroids already have 2,4 and 8 ohm taps. Power toroids are missing a tap, can only switch 4:1.
I SAW those for sale years and years ago on Plitron’s surplus page. At the time I had no idea I’d even be messing with anything bigger than a single pair of EL 34’s - even class H solid state was a dream back then (I couldn’t get one to not oscillate when the rails switch). Things have moved on since then. Should have just bought a pair anyway, and figured out what to do with them *later*.
If you are going to put in tubes on demand, I’d find a way to use 3 pairs. I like the idea of the scaleable voltage - solves the impedance mismatch issues when yanking out tubes. It can also compensate speaker impedance, but I’ll bet your output toroids already have 2,4 and 8 ohm taps. Power toroids are missing a tap, can only switch 4:1.
I SAW those for sale years and years ago on Plitron’s surplus page. At the time I had no idea I’d even be messing with anything bigger than a single pair of EL 34’s - even class H solid state was a dream back then (I couldn’t get one to not oscillate when the rails switch). Things have moved on since then. Should have just bought a pair anyway, and figured out what to do with them *later*.
Hi George,
That makes a lot of sense, but I wouldn't bother switching outputs in and out. That throws off the match with your output tubes. Dropping B+ to the levels as required greatly extends the life of the outputs while maintaining your impedance load. I assume that at lower powers, you would be listening more closely and enjoying the music quality. When you crank it up, you're more into the overall experience of feeling your chest move in time with the beat.
Anyway, great project! Can't wait to see what you do and how you do it!
-Chris
That makes a lot of sense, but I wouldn't bother switching outputs in and out. That throws off the match with your output tubes. Dropping B+ to the levels as required greatly extends the life of the outputs while maintaining your impedance load. I assume that at lower powers, you would be listening more closely and enjoying the music quality. When you crank it up, you're more into the overall experience of feeling your chest move in time with the beat.
Anyway, great project! Can't wait to see what you do and how you do it!
-Chris