This afternoon I’ve been working on another amplifier project. This particular one is the 6L6 SE-UL for which some readers have been looking since I published my Optimization of the 6L6 SE-UL page back in March of 2019. I recently realized that it had been three years since I published that data set and that it was probably time for me to get the project started. So this afternoon I’ve been investigating driver stages for the amp.
The 6L6 operating point I’ve selected for the amp is Vp=350v and Vk=32v for a peak power output of just over 7W per channel. This power level should make it a nice amp for filling a little larger rooms with lots of music. What I’ve found thus far is that the -32v bias point is in a weird spot from a driver perspective. Especially because I like to use current production tubes whenever possible.
My personal preference is for an amplifier to be able to get to full power with an input signal of about 1v-rms (1.4v-peak). This puts the input sensitivity at a point where the amp can be driven by virtually any modern signal source. This means that I’m looking for a driver stage amplification factor of about 32v/1.4v ≈ 23v/v (27.2dB). This spot is out of reach of the lower µ dual triodes like the 12AU7, 12BH7A, and the 6SN7/6CG7/6FQ7. I could specify a 12AV7 (µ≈35) or a 12AY7 (µ≈44) but I am hesitant to specify a driver tube which is so specific and generally more difficult to obtain. Although JJ does have a current production 12AY7 in their lineup.
However, the other thing for which I’m looking in this project is more glass presence up top (i.e bigger tubes). There are some good looking 6L6GC and 5881 tubes available. I am planning on an octal rectifier and I thought that making it an “all octal” amp would be a nice change. So this brought me back to the 6SL7. After checking a few different operating points on paper, I thought I’d try out a few tubes and see how they performed. Here’s my workbench this afternoon as I played with some 6SL7 driver stages.
This was just a gain study so I didn’t bother with the oscilloscope using a dual channel AC voltmeter instead.
I think I’ve settled on a bias point for the 6SL7 driver at Vp≈230v, Ip≈1.2mA with a 100kΩ plate load and a B+ of about 350v. At his point Av is about 42v/v (≈32.5dB). This may sound large but it is only about 5dB above target. A standard volume control has about 40dB of workable range so a 5dB offset shouldn’t be too bad. It’s a little more sensitive than I typically like my amps but still acceptable. Now that I’ve settled on a driver, I need to begin working on the overall schematic design.
And one more thing. I did make some progress this morning getting the third coat of oil on the 6AS7 SET chassis.
It’s starting to take on a nice finished look. I had thought that I would dull the finish before final assembly so that it had a more aged look. However, now I’m not so sure. I’ll have to give it some more thought. Especially with the copper top in place.
As always, comments and questions are welcome.
Pingback: 6L6 SE-UL Initial Power Up | Cascade Tubes
Pingback: Volume Controls and High Frequency Response | Cascade Tubes
Marblewood is even better than Lacewood. Are you thinking to tube roll 6550, 6n3c, 5881, or something bigger than a 6L6GC?
Yes the 5881 and 6П3С (or 6P3S) will drop in with no issues. I think the KT-66/7581A should also work but I’ll have to check to be sure. What I’m really hoping to make work is having the KT88 and 6550 drop in with no adjustments as well. The voltages won’t be an issue, I just need to check the big tubes with 470Ω of cathode bias and 400v B+ to ensure that total cathode current remains below 100mA (the current limit of my Edcor transformer). I will also have to make sure that the output transformers I use are OK with 100mA of bias current. So lots of work to accomplish as yet.
You used a 5U4 rectifier for the Lacewood. Would you use that tube again since you want to keep it octal? Your description sounded a little like a 6L6 Lacewood.
The choice of rectifier will depend on how the power supply works out. I have a 400v-0-400v power transformer (Edcor XPWR011-120) and the power stage requires a B+ of about 400v. Even with a 5U4 I think I’ll need a significant dropping resistor. Although that leaves me the possibility of using a 5AU4, 5V4, or 5AR4/GZ34 if I want to increase the B+ without circuit changes.
I am also exploring the possibility of making some more power tubes possible without changing the circuit. Kind of like a Marblewood on steroids.