Preamp resistors calculation for http://www.shine7.com

[southernoise]

I'm new to tubes and all that jazz and would appreciate some help in my calculations.

I'll be adapting Shine7 Audio DIY Page preamp design into my Sinovt kit amp. B+ voltage is 276V for the preamp section.

Currently it is a parallel common cathode pre and I thought it would be a waste to parallel the triodes, originally 12AX7 and now 12AU7.

Attached are my calculation sheets for 12AU7/12BH7 and the original Sinovt kit schematics. Thanks!

[southernoise]

uhm.... anyone? Please help a noob not blow any tubes!

[Miles Prower]

Loadlines look OK, but you're running 'em awfully hot. The spec sheet calls for Pd= 3.5W (max) for the 12BH7, and Pd= 2.5W (max) for the 12AU7. If using a 12AU7, you're over spec on Pd. For the 12BH7, you're awfully close to the recommended Pd= 0.8Pd(max).

You don't need to bias that hot, especially not a gain stage driving a cathode follower. A shallower loadline will help increase gain, and improve linearity.

[southernoise]

Thanks Miles for your insight.

So does this mean I should redraw the load lines to a lower Ia and recalculate especially for the follower? For the 12BH7 follower, if I choose Ia = 25ma, cathode idle voltage will be 128V and Pd = 3.2W which is within specs but a little hot.

[Miles Prower]

Originally Posted by southernoise

So does this mean I should redraw the load lines to a lower Ia and recalculate especially for the follower? For the 12BH7 follower, if I choose Ia = 25ma, cathode idle voltage will be 128V and Pd = 3.2W which is within specs but a little hot.

Yes, definitely redraw the loadlines to reduce Ia, and increase Rp for the gain stage as well as the cathode follower.

From the EL34 specs, I get:

Ci= Cgk + Cmiller + Cstray= ~40.0pF

At 30KHz:

Ic= 18.9(2pi X 30E3 X 40E-12)= 142uA

5 X 142E-6= 0.71mA

This isn't a very hard load, and by the Rule of Five borrowed from solid state design practice, you wouldn't need even a milliamp of Q-Point current from that follower.

For considerations of slew limiting, clip on overdrive performance, incidental grid current as Vgk approaches 0, I would pick a Q-Point current of ~5.0mA for the follower. That'll pretty much take care of all of that, and keep it nicely away from cutoff conditions. You just don't need any more than that.

As for the gain stage, definitely get that Rp up. Triodes work better, the lighter the load they operate into: more gain, more output swing, less distortion. A cathode follower makes for a very "friendly" load for any triode gain stage.

[southernoise]

Thanks for the detailed explanation. Let me digest this all and look for the necessary equations.