I am in the middle of designing and building a parallel 6L6GC SE amp but would value opinions on whether to use a 12ax7 as a conventional voltage amp with cathode follower output or to use a Mullard M8100 pentode (aka EF95 6AK5, 6P1) only (no separate cathode follower.)
I believe triodes tend to more even order harmonics and pentodes to more odd order.
Supply voltage can be ~300v for the tridoes and ~ 200-250 for the pentode. (main HT is ~500v)
Thanks.
I believe triodes tend to more even order harmonics and pentodes to more odd order.
Supply voltage can be ~300v for the tridoes and ~ 200-250 for the pentode. (main HT is ~500v)
Thanks.
I feel the difference will be mostly marked by using a cathode follower after the voltage gain stage or not using that, swamping the differences between triode or pentode of similar gain.
What are other tubes to which you have access?
Pentode will have big output impedance also.
What about to use some good pentode but triode strapped? E180F, 6E5P, 6J9P just to name a few.
Hybid CF
I've only have 12AX7 and M8100 in my box of bits but can always buy a different (higher current) triode for the CF. (eg 6N1, 6DJ8, ECC88, ).
What about a hybrid CF from https://www.tubecad.com/2011/08/blog0212.htm
I've only have 12AX7 and M8100 in my box of bits but can always buy a different (higher current) triode for the CF. (eg 6N1, 6DJ8, ECC88, ).
What about a hybrid CF from https://www.tubecad.com/2011/08/blog0212.htm
Attachments
Dont have experience with the hybrid design, dont want to give you advice with something Ive not tested.
I think the E88CC or 6N1P will be surely better sounding.
Or you can try the ECC83 as anode resistor loaded dc coupled to another half cathode follower - you will eliminate one signal coupling cap which is absolutely unnecessary.
I think the E88CC or 6N1P will be surely better sounding.
Or you can try the ECC83 as anode resistor loaded dc coupled to another half cathode follower - you will eliminate one signal coupling cap which is absolutely unnecessary.
Transistor in the cathode acts like a CCS?
By the way I would add protection diode, anode to grid of second tbe and cathode to cathode. For protection of the tube when turning on.
If you will buy ECC88s I would use simple gain stage or simple one SRPP as it will do the job quite nice, no need for bigger complexities.
I thought of the first picture only when you will want to use the 12ax7s.
For me, ECC88 sound far superior to any ECC83. But best will be if you will hear it for yourself and decide what you like more
By the way I would add protection diode, anode to grid of second tbe and cathode to cathode. For protection of the tube when turning on.
If you will buy ECC88s I would use simple gain stage or simple one SRPP as it will do the job quite nice, no need for bigger complexities.
I thought of the first picture only when you will want to use the 12ax7s.
For me, ECC88 sound far superior to any ECC83. But best will be if you will hear it for yourself and decide what you like more
Last edited:
I am not qualified to say whether its acting as a ccs or not. The tubecad blog doesn't say so. The distortion spectrum is very good compared to an 'ordinary' CF.
"The test was 1Vpk a@ 1kHz into 300 ohms. Yes, the hybrid CF+ offers lower distortion, a lower Zo, and higher gain (a closer approximation to to true unity gain). Yet the hybrid CF+ has no less a single-ended sonic signature, as the above graph reveals."
"The test was 1Vpk a@ 1kHz into 300 ohms. Yes, the hybrid CF+ offers lower distortion, a lower Zo, and higher gain (a closer approximation to to true unity gain). Yet the hybrid CF+ has no less a single-ended sonic signature, as the above graph reveals."
Not acting as a ccs. More like a push-pull totem pole arrangement. Notice that the signal goes to the base of the transistor, so the current is variable with the signal rather than constant.
Do you need a lot of gain, or do you have reasonably sensitive speakers? If you don't need that much gain, you could use a more beefy driver with a mu of 15-20 like EL84 or KT61 in triode. I like drivers with a fair bit of current through them.
I'll be driving a pair of 6L6GC or KT77 in parallel so need a around 40v p-p into about 50k load which under 1ma, so 10ma through the CF should be adequate.
The 12ax7 with mu =100 means about .4v p-p input probably a bit more as the CF will not be a zero loss.
My speakers are a tricky load even though I will only driving the electrostatic panel with the tube amp (Krell KSA50 clone will drive the bass unit which crosses over around 400Hz).
The 12ax7 with mu =100 means about .4v p-p input probably a bit more as the CF will not be a zero loss.
My speakers are a tricky load even though I will only driving the electrostatic panel with the tube amp (Krell KSA50 clone will drive the bass unit which crosses over around 400Hz).