Bass amp; everybodies favorite 4x K88 driver tube

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Just wondering what's the favorite PI, driver, or PI/driver tube is in a quad KT88 bass guitar PPP power amp (120W). I'm going back and forth between the 12BH7, ECC99 and 6SN7. They all have their strenghts and weaknesses. Any other tube suggestion is welcome as well. The 12AU7 didn't make the cut though.
 
Do you intend to follow the data sheet recommendation for maximum grid circuit resistance of the power tubes? (50K IIRC) Most musical instrument amps don't. If you do, you usually need a cathode follower driving the grids of the KT88s, an LTPI usually won't get the job done.
 
Do you intend to follow the data sheet recommendation for maximum grid circuit resistance of the power tubes? (50K IIRC) Most musical instrument amps don't. If you do, you usually need a cathode follower driving the grids of the KT88s, an LTPI usually won't get the job done.
Thanks for the tip, and I indeed plan to use a cathode follower (hopefully DC coupled to the LTP). For 4x the KT88 some (not me) consider it overkill, but by next project will be a 250W 6x KT88, where a CF is definitely needed, and I like to use the same front-end.
 
The other more complicated but technically superior option is to use a pretty much standard PI with a 12AX7 or 12AT7. Then buffer the output of this with 4 off mosfet source followers (one per output tube) with the load resistors on these returned to a negative rail of at least 3 times the bias voltage. These resistors PLUS the grid stop must should meet the Max grid resistance spec. That is teh output tube grids are direct coupled to the siource followers. AC couple the followers. Apply the bias to the mosfet gates. The mosfets need gate stoppers of say 1K and protection zeners (15V) with their cathode to teh gat and their anode to the source. Resistors from the other side of teh moosfte gate stops to the bias can be 1M or even 2.2 M. That means you don't need big coupling caps (10nF should be fine) and the load on the PI is just 2 of these in parallel (at least 500K).
This scheme has the advantage that you can run the KT88 at higher idle currents ( e.g. bias for 90% of dissipation instaed of being restricted to 70% by the high grid resistance) and hence screw a bit more power out of teh amp and get better damping without the need for lots of global feedabck. (At higher idle current the KT88 internal resistance "rp" is lower resulting in a lower output impedaance rlected to teh secondary of the output tranny).

If you want to get real fancy then use a "ring of two" transistor current source in place of each mosfet source load resistor.

Lots of text here - to demonstrate what I'm talking about see the circuit below for the output tube drive/bias arrangement (this is my HiFi Amp design, Ignore the UL connections etc. this is just to show the output tube drive/bias arrangement discussed).

Cheers,
Ian
 

Attachments

  • BH_Fixed_bias_A.jpg
    BH_Fixed_bias_A.jpg
    79.9 KB · Views: 78
Thanks for the extensive reply! I'm learning a lot here, but I have to be honest, that's still way above my expertise. It's just my tube power amp, so I'm trying to limit it to a design I can understand at a fundamental level ;). But again, great stuff and I will study it for every bit of knowledge I can gather.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.