Well, there you go! 🙂
I built my audio tube based Modular with KT88 and cathode bias. Works well, and you get about 60W UL or 30W RMS in triode using 4k3 OPT (Hammond 1650N) with 560V B+. I paid 200$CAD for the four KT88 Shuguang tubes. Transformer set ~500$CAD
To run fixed bias, I would have needed 1650R to get 100W UL and ~35W triode. Extra cost for not much gain IMHO.
I'm not sure where 5W wirewound pots are 7$ though, but I would suggest using at least 10W pots there - 12½ Watts even better - however, just as I thought, those aren't 7$ - at least for the Ohmite I had in mind... Only 80$USD at Digi-key though 😛
https://octopart.com/res500e-ohmite-153557?r=sp
Consider making the 500R pot 200R so a 5 Watt pot will last a while 🙂 Adjust the other resistors to compensate?
Another idea is to use TL783 or another CCS for cathode bias - set it and forget it and have matched current per tube. Unfortuneately TL783 is backordered until the Cardassian Empire conquers Earth 🙁
You'd also need decent heatsinks for them.
I built my audio tube based Modular with KT88 and cathode bias. Works well, and you get about 60W UL or 30W RMS in triode using 4k3 OPT (Hammond 1650N) with 560V B+. I paid 200$CAD for the four KT88 Shuguang tubes. Transformer set ~500$CAD
To run fixed bias, I would have needed 1650R to get 100W UL and ~35W triode. Extra cost for not much gain IMHO.
I'm not sure where 5W wirewound pots are 7$ though, but I would suggest using at least 10W pots there - 12½ Watts even better - however, just as I thought, those aren't 7$ - at least for the Ohmite I had in mind... Only 80$USD at Digi-key though 😛
https://octopart.com/res500e-ohmite-153557?r=sp
Consider making the 500R pot 200R so a 5 Watt pot will last a while 🙂 Adjust the other resistors to compensate?
Another idea is to use TL783 or another CCS for cathode bias - set it and forget it and have matched current per tube. Unfortuneately TL783 is backordered until the Cardassian Empire conquers Earth 🙁
You'd also need decent heatsinks for them.
Last edited:
Let me suggest to do yours first project with a cheaper tube for earning purposes. A 6V6, 6F6, or any horizontal or vertical deflection pentode.
The 100 ohm balance pot should only be passing, at most, 10VDC to ground, at 130mA max. It's a very small portion of the total cathode voltage. That's less than 1.5 watts. You don't need a 12 watt pot there, and you don't want more than 100 ohms or fine-tuning the grids of the two tubes will be difficult. The tubes still need to be well-matched. The pot only compensates for a few milliamps between the tubes.
This pot works just fine:
https://www.digikey.com/en/products/detail/cts-electrocomponents/026TB32R101B1A1/203775
Another thing to remember about the balance pot in this sort of design is that it's sized to float the grids of the output tubes at a certain percentage of the cathode voltage--a very low percentage, like 5-6vdc. If you increase the value of that pot to, say 500 ohms, the grids will be floating at a much higher percentage of the cathode voltage, more than halfway, thereby reducing the differential between the grids and the cathodes and restricting the operation of the output tubes. Again, the purpose of the pot is to balance the grids, not to provide any signficant cathode resistance. The Williamson output stage cathode arrangement was carefully designed.
Curious about the Williamson balance pot. What would happen if the pot fail and there is a disconnect between the sweep and the trace? The output tubes lose their ground reference. Hypothetically there is no longer a current path, but that's not always what happens in reality when it comes to tubes. What's the failure mode?
It seems to me that if the balance pot failed, both the grids and the cathodes would lose reference to ground. The tubes would fail to conduct and the plate voltage would increase accordingly, but it would not be a disaster. But I could be wrong...!
Pot wiper open? . . . .
The cathode voltage would increase (it is now a zero bias tube), and the cathode voltage would increase, it might go high enough to destroy the cathode bypass capacitors.
I only use pots if I have to, I usually use a circuit that does not need them.
Just my preference.
The cathode voltage would increase (it is now a zero bias tube), and the cathode voltage would increase, it might go high enough to destroy the cathode bypass capacitors.
I only use pots if I have to, I usually use a circuit that does not need them.
Just my preference.
With the grids floating as well? In other words, the only thing in play is the plate voltage. Essentially the tube is not operating at all. It seems to me that if you disconnected grid and cathode, and left only plate voltage, the tube is drawing no current at all. If I applied plate votage to a tube with both the grid and cathode disconnected, what would happen? Nothing. Am I wrong?
I have a simple and cheap solution to the bias problem.
1. Use a common cathode resistor of 300 ohms 10w
2. Put 2 lineair pots of 50k across it
3. connect the 200k grid resistors with the wipers.
4. Start with wipers at ground position when you switch on the amp; put a voltmeter across the 10 ohm resistors and slowly turn up the wipers until you get 0,5v across them. Each tube is now drawing 50 ma. There's no need for a balance pot because the cathode currents are adjusted separately.
1. Use a common cathode resistor of 300 ohms 10w
2. Put 2 lineair pots of 50k across it
3. connect the 200k grid resistors with the wipers.
4. Start with wipers at ground position when you switch on the amp; put a voltmeter across the 10 ohm resistors and slowly turn up the wipers until you get 0,5v across them. Each tube is now drawing 50 ma. There's no need for a balance pot because the cathode currents are adjusted separately.
The output and power suply transformers i already have.
And that are the most expensive parts, from my point of view.
I am aiming for Ia=45mA en Ug=-36V.
Therefor i need 800 ohm kathode resistor.
And that is where the resistor values in my schematic come from.
As you use a common Rk for 2 tubes that resistor should be halved so 400 ohms. So instead of using 300 ohms as I suggested earlier, take 500 ohms.
To prevent damage to the tubes when turning the wiper the wrong way connect a resistor of 50k in series with the pots and connect those with the 10 ohm
test resistors. Now it's still not safe to turn the pots all the way up but at least you wont destroy the EL34's right away.
To prevent damage to the tubes when turning the wiper the wrong way connect a resistor of 50k in series with the pots and connect those with the 10 ohm
test resistors. Now it's still not safe to turn the pots all the way up but at least you wont destroy the EL34's right away.
No not like that. You must connect the now open end of both pots to ground and not the wiper! If connected correctly the voltage across the Rk (which is now relatively big) provides the grids with a small positive voltage depending on the position of the wiper and that determines the amount of current through the tubes.
The way you have drawn it the wiper is shorted to ground and that's dangerous because in its highest position it shorts the cathodes to ground with foreseeable results.
The way you have drawn it the wiper is shorted to ground and that's dangerous because in its highest position it shorts the cathodes to ground with foreseeable results.
Leemhuis01,
Your Post # 35 . . .
Be sure to use very well matched EL34 tubes.
Really well matched ones are available at eurotubes.com
They ship to many countries worldwide.
Then with very well matched tubes, your output transformer will love you because they hate un-balanced plate currents.
Your Post # 35 . . .
Be sure to use very well matched EL34 tubes.
Really well matched ones are available at eurotubes.com
They ship to many countries worldwide.
Then with very well matched tubes, your output transformer will love you because they hate un-balanced plate currents.
That's better. To return to my earlier post add 50k resistors from the hot side of the pot(s) and connect those with 10 ohm resistors. This prevents an accidental short of the cathodes to ground in case the wipers are all turned up thereby destroying your power tubes.
Well,
Now you see why I pay for very well matched tubes, and why I use self bias.
Simple.
In recent years, I used individual self bias resistors, and individual bypass caps. Of course, a loud long section of music does shift the self bias a little.
I Finally built a design that I have been thinking about for a long time.
My most recent amplifier is balanced, I use a single resistor for both cathodes, no bypass cap, and it sounds wonderful.
But of course I do use very well matched output tubes so that the plate currents are equal even with the cathodes tied together to the single common self bias resistor.
Now you see why I pay for very well matched tubes, and why I use self bias.
Simple.
In recent years, I used individual self bias resistors, and individual bypass caps. Of course, a loud long section of music does shift the self bias a little.
I Finally built a design that I have been thinking about for a long time.
My most recent amplifier is balanced, I use a single resistor for both cathodes, no bypass cap, and it sounds wonderful.
But of course I do use very well matched output tubes so that the plate currents are equal even with the cathodes tied together to the single common self bias resistor.
Okay, some time have past.
And I have spend some time designing my amp.
I came up with the following schematic, see PDF file.
I build this schematic on a piece of wood.
And the result is a working amp.
But Since this is my first experince building a tube amp.
Do you guys have some tips and tricks or something.
Or do you see things that i have not done right.
I would love some feedback on my project.
And I have spend some time designing my amp.
I came up with the following schematic, see PDF file.
I build this schematic on a piece of wood.
And the result is a working amp.
But Since this is my first experince building a tube amp.
Do you guys have some tips and tricks or something.
Or do you see things that i have not done right.
I would love some feedback on my project.
