Tube choice for tubed crossover

[ashok]

I am building a tubed active 12db/octave crossover.
The configuration uses the tube in a cathode follower mode.
If I use an ECC type tube , which is better - an ECC82 or an ECC81 ? The load will be over 22Kohms.

Are the 6SL7 and 6SN7 useful here. I am concerned about the distortion levels at about 2 volts and the output impedance.
Thanks for any help.
Cheers,
Ashok.

[EC8010]

22k is a heavy load. Just because a cathode follower has low output resistance does not mean it will take kindly to such a load. A cathode follower is a special case of an ordinary common cathode amplifier, but with the load moved to the cathode to give 100% feedback. Plot a 22k loadline on these valves, and you will see that it is a heavy load.

Since the gain stage is a cathode follower, the more mu you have, the more distortion-reducing negative feedback you have. However, Sallen & Key filters use positive feedback to achieve their slope, so they increase distortion. 6SN7 is probably the best compromise because it is low distortion. ECC82 is low mu and high distortion. 6SL7 is high mu and quite low distortion, but cannot drive significant current, and that's important because the input impedance of S&K is lower than you would expect from just glancing at circuit values. ECC81 can drive current, and has reasonably high mu and gm, but is not particularly low distortion.

[SY]

My favorite tube for xovers is sadly no longer available. But what I would probably do is configure the innards with something like a 12AX7 or 6SL7 (you need gain to get the filter functions to follow properly), then buffer the output with something like a 6SN7 or 7119.

[pedroskova]

Might be a good place for an ECC99.

[corbato]

Hi Ashok,

E88CC. I built the Steve Bench xo and remain quite happy with the results.

[planet10]

I have speculated about active tube XOs, which seem to end up being pretty much filters sandwiched between cathode followers (what i'd like is a tubed subtractive XO). Since a cathode follower works a lot better with a CCS on the bottom i have wondered about the feasibility of using something like 6AN8 (this tube seems to get trashed even more then the 'AU7) or a 6X8 or somesuch with the pentode section used as the CSS?

dave

Tube XO Schematics

[EC8010]

Oh yes. Almost any cathode follower with a CCS load is in with a chance, and some of the little triode pentodes are ideal. 6FG8 looks good. None have the linearity of a 6SN7, but they're dirt cheap. The only things that you need to watch out for is to check how low an anode voltage the pentode section will tolerate whilst still being securely on the flat part of the curves, and to check the permissible heater/cathode voltage for the triode section. The PCF types tend to have better heater/cathode insulation. If you pull the pentode's control grid up to 10 or 20V, this allows a much bigger cathode resistor, making ra' very large indeed, and the triode's loadline horizontal.

[DougL]

My favorite tube for xovers is sadly no longer available. But what I would probably do is configure the innards with something like a 12AX7 or 6SL7 (you need gain to get the filter functions to follow properly),

Would a Pentode cathode follower be appropriate, or just overkill?
Tube Cad recomended 6DJ8, the 6SL7 seems cool, and the 12ax7 is the triode gain champ. Interesting choices.

I am just starting to research building the Linkwitz Phoenix or PMTM1 & PW1.
If I'm going to build crossovers, tubes seem the natural choice for most of the functions. The exception is the gyrator. I got to admit, a tube gyrator is a bit much. But I could simply use an inductor for that.

Anyone with experience on these type of filters?
A 3 way stereo L-R 4'th order would have 14 dual triodes.
And that doesn't include the baffle compensation, the all-pass phase shifter, etc. I would really like your collective insight on this.

Doug