EL84 / Champ Design and build

[rickyl]

Hello. This is a design I came up with combining a Champ preamp/tone stack and an EL84 output. Not sure on the grid resistors for the EL84, so I just made some educated guesses. No feedback circuit yet, I was thinking I would add that later if need be. The transformers were rescued from old equipment, so that's the deal with the over powered PT and dropping resistors. I got it wired up but wanted to get some thoughts and opinions before I started soldering. Thanks.

[Jazz P Bass]

The only thing that I see missing is the heater winding for the 5U4.

[rickyl]

The only thing that I see missing is the heater winding for the 5U4.

Yes, I omitted the heater circuits on the schematic. The 12AX7 and EL84 filaments are wired at 6.3, the 5U4 is on the 5v winding.

[eschertron]

The grid-leak resistor for the EL84 can be the traditional 220k.
Also FWIW, from what I hear the 'sweet spot' for EL84 chime is with about 300v from cathode to anode. You may find that all of the PSU dropping is not needed in your final build. I finished a PP EL84 build this year, and am putting a bigger PT (500vct vs 450vct that's in there) to get a little more B+. 260 to 270v cathode to anode voltage sounds good, but I'm looking for a little more

Please keep us posted on how it goes!

[Chuck H]

I'll mention a couple of things just 'cause you asked for it But really, looks alright to me, so...

If I were going to use a BF type tone stack I would include a middle control for versatility. Maybe just a three way mid selector switch if you wanted to save face panel space. Otherwise, sans a mid control I might choose a less lossy tone control arrangement just to get a wider range of clean/overdrive from the amp. This would be strictly my own design perspective and shouldn't be taken as any sort of critique of yours.

In the power supply you have voltages indicated. I was wondering how you arrived at your figures. 270V pretty conservative for the plate of an EL84 guitar amp. Even class A. The Pi filter arrangement works great for rectifier simulation with a considerable reduction in power supply ripple. I've used it a couple of times. Real world values in the circuits I've built make your resistance of 2.5k seem enormous. I don't know what your PT starting voltages are but I still wonder how you arrives at your PS voltages and component values. I like to use Duncan PSU. Works great and, used correctly, has always been +/- 10V for me.

[Mick Bailey]

Yes, those paralleled 5K resistors look too high to me. I'd be aiming for the 1st node to be giving 310v - 320v. I also prefer at least a 1K screen resistor for an EL84.

[rickyl]

How I got that voltage setup. I started with the OT I had on hand. It was giving me 28:1, so (28^2) * 8 = 6272 (8 ohm speaker). Then I did Va^2 / Wa = Z, and got a value of Z = 6075 when Va = 270 with Wa(max anode dissipation) of 12w. I didn't really like the dropping resistors either, but I needed to match up the tube with OT values so... Not sure if I'm exactly right with all this, but that's how I came up with it. Thanks!

[Dave H]

The input jack isn't grounded.

I'd be looking to roll off the bass somewhere by reducing the value of a coupling capacitor but the circuit only has the cap to the EL84 grid. Perhaps reduce the value of one or both cathode bypass capacitors?

[überfuzz]

The 25uF Fenderish bypass cap... well the size isn't really necessary in a guitar amplifier.

Good luck with your build. :-)

[rickyl]

The input jack isn't grounded.

It's not grounded only on the schematic. Just a typo.

[ron vogel]

You could "borrow" a few elements from the Marshall Class 5...I posted the schematics for it on this site a few years back...they might still be floating around. It does have the TMB tone controls...and I agree, they are handy to have.

aha:
http://music-electronics-forum.com/t19680/

[eschertron]

How I got that voltage setup. I started with the OT I had on hand. It was giving me 28:1, so (28^2) * 8 = 6272 (8 ohm speaker). Then I did Va^2 / Wa = Z, and got a value of Z = 6075 when Va = 270 with Wa(max anode dissipation) of 12w. I didn't really like the dropping resistors either, but I needed to match up the tube with OT values so... Not sure if I'm exactly right with all this, but that's how I came up with it. Thanks!

If the Va in the equation above is the anode-to-cathode voltage of the EL84, you'll want to add the bias voltage (somewhere between 10v to 15v) to the B+ to get the dissipation you want. So as you build, you can be looking at reducing the power resistor value in the PSU to get B+ where you need it to be.

+1 to Chuck's comments about the tone stack. If you're looking to overdrive the EL84 more than just a little, a tone stack lift or a control like the Fender 5E3 (some other early Fender models used it too) will allow the output tube to really 'sing'.

[rickyl]

If the Va in the equation above is the anode-to-cathode voltage of the EL84, you'll want to add the bias voltage (somewhere between 10v to 15v) to the B+ to get the dissipation you want. So as you build, you can be looking at reducing the power resistor value in the PSU to get B+ where you need it to be.

So, I designed the plate voltage at 270 with a cathode bias of 8 volts. You are saying I should add these values together for the B+, 278v?

[rickyl]

Yes, those paralleled 5K resistors look too high to me. I'd be aiming for the 1st node to be giving 310v - 320v. I also prefer at least a 1K screen resistor for an EL84.

Yes, when I did the math, I came up with 1K as being a more appropriate value for the screen stopper. But I found a cool reduced mass wire wound I wanted to use and it was only available at 470. I didn't think the increased dissipation would be too bad, since a lot of amp designs I see don't even have a screen stopper. I could be wrong though...

[rickyl]

The grid-leak resistor for the EL84 can be the traditional 220k.

I've seen 6V6's with the 220K. On the 6V6 datasheet it has a max value of 0.5M. So 220K is around half that.
The EL84's max value is 1M on the datasheet. So I thought, around half the max value is in the ballpark. 680K would make it a little more sensitive, but the 10K stopper will help calm that down.
If my logic is flawed let me know.