Ambitious New Compilation

[Howlin' Mad Mac]

Just finished drawing this up. It's a more ambitious design for me & I could really use a few more sets of eyes looking it over before I go any further. It's a dual channel setup where channel 1 is supposed to be akin to a jumpered plexi & channel 2 a BF Bassman/Bandmaster kind of deal into a typical Fender/Marshall PI with an 18 watt EL84 cathode biased backend for now although if I'm happy with the end results it should be simple to alter design to accomodate any P-P output pair depending on transformers used. Same if I decide to change to a fixed bias design. Also haven't designed the PS yet so voltages listed are desired approximations. Let me know what y'all think... can't wait to read your comments & suggestions.

Dual Channel.pdf

[loudthud]

Have you tried the arrangement on the tweed front end with the partially bypassed common cathode? Might be too much crosstalk between the two channels. You might want to leave room for split resistors/caps in the layout. The one thing I question is the way the channels are mixed at the PI. You may have to tweek things there to get good balance between the channels. You won't really know until you build it. There should be plenty of gain with EL84s. R21 probably needs to be 100K or higher.

[es345]

Split R32 and C18 so that that you have a separate Cathode bias for each EL84
(R32a,R32b= 270Ohm - very hot - or 330 ohm -a little bit colder, C18a,18b=25uF) .

This helps you in using non matched EL84.
In addition if one El84 is dying the 2nd is not killed by this fault.

[Chuck H]

You have approx 5:1 voltage division on the NFB loop AND your driving the loop from the 16 ohm tap. That's A LOT of NFB. This is where loudthud is going with the value of R21.

Also a lot of swing for a pair of EL84's. Be prepared to change the PI circuit values in case it's problematic. Maybe R24 at 1k and R22 at 47k. That being the possibility I never like having a 5k NFB shunt when I'm already using a higher value tail. You could change the presence pot to 500R, swap the NFB connection to the 8 ohm tap and change the presence cap to .33uf. If you do this just leave R21 at the value it is.

When ordering parts, pick up a pair of 1W 15V zeners, a 13V 1W zener and two 1n4007 diodes. That way if (when) you get that hideous hashy crossover distortion that plagues 2XEL84 amps you can try the Paul Ruby mod (with my addition) to smooth it out. Cheap insurance in case it comes up.