Triode comparison - EL34, 6L6, 5881, KT88, 6550 etc

[neodymium]

6A3sUMMER,

I said: basically.

If the driving impedance is very high (pentode without any feedback), the load impedance (+ series sec. winding res.) determines near totally the low frequency corner - at the OT: at half R, will near halved low corner freq. If the driving impedance (+ series primary resistance) is 1/3-1/4 compared the reflected nominal secondary load, the load influence of corner freqency decreases. Of course, your argument is true: the influence of load isn't nulled, only reduced. That is the reason, I said "basically".
You and me have consent basically, only we walked around the problem well. 🙂

 

[WhiteDragon]

Beam tetrode always.
Being common but very good tubes.
If playing with lower power triode then 6V6.
More power not triode use 6550
Only 2 tubes you need.
Straight plain bottles.
Need more romance with wavy bottles for
looks. Play with modern KT66 not triode.
KT88 just forget about this tube all together
regardless of power ratings/ specs.

 

[toshiba_nz]

I've used Svetlana EL34 in triode. Pretty nice.

 

[benchbaron]

I'd like to try EL86 in triode........(I´m tempted too as I have a quite a few)......Quick glance in triode mode, the tube with one of the lowest Ra of the B9A group (right/wrong??) , a bandolier of 6 or more tubes in parallel would really simplify the output transformer with vastly improved bandwidth and an easy B+ voltage. The caveat is the summed higher tube Cin capacitance, and driver stage designed accordingly. An EL86 would rob only 1/2 the heater current to a KT88 and two parallel EL86´s would halve the ra of the KT series.
Shall give this a whorl. and burn some midnight oil over it.
Bench Baron

 

[stephe]

I am not sure what the problem with triode connected 6550/KT88 comes from, but I just found them very "uninvolving" and slow sounding. Maybe the greatly increased miller capacitance was just too much for the wimpy driver stages I used at the time, both SE and PP.

Fuling said: This sounds plausible to me, wimpy driver stages is a classic way to ruin the sound from triodes.

That does make sense. I agree taking a UL or pentode KT88 amp (usually has some sort of a triode driver) and simply triode strapping the KT88 ends up sounding mushy and just bad IMHO. And my admitted limited experience working with the 300B tube has lead me to believe driving it with a pentode or a cascode sounds WAY better than any triode based driver, even a 2 stage triode capable of the needed voltage swing. Maybe there is something to the adage "Drive a pentode with a triode and drive a triode with a pentode"?

I do really love the sound of a beam pentode UL SE with plate to plate "Schade" type feedback. Plenty of second order harmonics and almost no third order. It also gives you the ability to tune the sound signature of the amp by simply adjusting this resistor value to balance the THD vs "liveliness" you find enjoyable to listen to. I've tried other SE UL designs and always come back to this.

But this has got me thinking about trying my cascode 300B SE design, but with a triode strapped KT88 or KT120 instead of the 300B tube.

A minor correction to an earlier post in this thread: you need to tie pins 1 and 8 with an EL34, not pins 1 and 2.

 

[Fuling]

In a situation like this, where the sonic merits of different (but equivalent) tubes are being compared, I imagine it could be wise to use a buffered driver stage. By doing we should be able to get a clearer picture of the differences between the actual tubes as their input capacitances etc. no longer can interfere with the driver stage.

 

[6A3sUMMER]

Stephe,

Thanks for correcting me in your Post # 45 !

In my Post # 5, I made a major mistake. Too late, so I can not change or edit my post.

For an EL34, I said to tie pin 1 to pin 2. . . . Wrong!
I should have said, tie pin 1 to pin 8. . . . Correct! That ties the Suppressor Grid to the Cathode.