Moving my response to this thread too:
Resistive losses don't change the frequency that much.  Both exist, arc capacitance and arc resistance.  In my limited experience, long thin DRSSTC arcs tend to have less loss relative to capacitance, so capacitance dominates.  Thicker arcs from QCW coils or continuous (not interrupted) SSTC coils tend to have more loss compared to capacitance.

Hi all,

:^) Yeah, the 833 is a weird tube with grid and plate pins both on top. Bad thing is several people in the past have accidentally mixed the connections up when they are adjusting/changing things since pins are same size.
Very bad outcome of course!

I also feel that this high voltage negative peak may also be very possible, but overall driving power low.

Several years ago, in the first tube coil of mine that I tried a tall primary I experienced the first time of noticable strange feedback winding behavior. Performance was sub par this way also despite a few changes I tried. Any changes I tried seemed to not be very responsive at all. Very little change, when it should be much more. I then immediately realized since the feedback is up so high now on the secondary that it was now influenced by the voltage rising on the secondary, instead of the primary winding that we want. And since the feedback is grounded...arc over problems too...and I see the arc shield done here, which others have resorted to also when having this problem.

Me, I knew we want to stick to feedback off of our primary. So I just took the whole primary and feedback form and I literally with no changes, just turned it upside down as is so feedback is now on the bottom. Since bottom of secondary is mainly all current, arcs to feedback problems gone. And now we always have the wanted feedback from primary only. I feel this is key to best performance possible.

If you flip this around, experiment slowly to get things tuned up correctly for the feedback. Also feedback leads may need to be switched around for correct phasing after this change.

Ever since I have run into this problem my first time, I now always build my tube coils with the feedback on the bottom below the primary, whether a short primary or tall one. My experience shows that it works much better this way once adjusted correctly of course.

I agree a variable rheostat is the way to go on the grid leak... being able to adjust on the fly is key also to best performance. I sometimes also in conjunction with the rheostat, have a fixed smaller ohm value "safety" resistor with it. That way you do not accidentally turn down to zero resistance. But these days...I mostly run just rheostat only and remember about not doing this.

Chris