First-Time Tube Amp Design

[Theashe]

Hey all,

I'm just entering the world of designing and building tube amps. I've been spending months digging through forums and Couch's and Merlin's works to learn what I can, and I think I'm finally ready to start designing.

The design I'm working on is a 30W cathode biased push-pull amp using two 6L6GCs as output tubes. I just finished the first iteration of the power stage using Couch's book, and I'd like to get some feedback from you guys here at the Amp Garage before I move onto the next stage of the design.

While understanding the theory is essential to making a good amplifier, I hear everywhere that experience and intuition are just as, if not more, important. That's why I want to post this here (see schematic in the attachment). Those of you that are more seasoned and intuitive than I, please take a look at this and tell me if there are any glaring discrepancies or obvious no-no's that jump out at you.

The specifications I wrote at the bottom were calculated in a spreadsheet as I went through the design process, and hopefully reflect something close to reality.

Thanks for taking the time to look at this, and sorry about the quality of the schematic - I hacked it together in PSpice. I'm looking forward to sharing my progress on the forum as I complete my first build.

Dave



EDIT 4/12/2014 - I've redesigned the power amp, and it's not similar to the attachment anymore. However, I'll leave it in the post for relevance.

[matt h]

(deleted)

[Theashe]

Thanks for the feedback, Matt. Could you elaborate on what you mean by 'unusual screen arrangement'? What is a more typical/common arrangement?

The grid leaks were chosen as the max from the GE data sheet for 6L6GC tubes, and the grid stoppers were more or less arbitrarily chosen to be larger with the hope of making a better low pass filter with the tube capacitances.

When I'm home from work I'll share some of the math/rationale behind this design.

[matt h]

(deleted)

[Theashe]

You're right - by tying the screen to the plate supply without any extra filtering, a fair amount of power supply ripple would reach the screen. That would certainly make for an interesting sounding amp, but not quite what I'm looking for. I made a few tweaks to the schematic thanks to your advice, Matt - but I'm realizing that I'm going to have to design and test my power supply before some of these values become permanent.

Which brings me to another beginner question - I want to make this amp as easy to tweak as possible to adjust the sound to what I like without replacing a lot of parts, which means using potentiometers instead of resistors in some important places. Eventually, when I finish more of the amp design, these are resistors I'm thinking of replacing with potentiometers:

- Cathode bias resistors
- Cathode follower grid resistor
- Screen bias resistor
- Global negative feedback adjustment
- Phase inverter plate resistor (for adding or removing imbalance)
- Phase inverter cathode resistance (for shifting balance toward power tube or preamp distortion)

Do any of these jump out at you as bad ideas for replacing with potentiometers? I don't mind buying higher power potentiometers since this is a one-off personal project and cost isn't as important as sound. However, I'm curious - do potentiometers have worse noise artifacts than resistors?

Dave

[matt h]

(deleted)

[johnnyreece]

You might consider a decade box rather than a pot for higher voltage resistors. You can roll your own, or I think Weber still has them for sale.

[d95err]

As it is your first build, I strongly recommend sticking to old prooven designs. You don't have to copy a particular design, just use bits and pieces of them as building blocks.

Old Fender schematics from the 50s and 60s is a good place to start.

[Theashe]

PI cathode-- check outthe "watts" control, as its often called. Essentially, it's a variable tail resistor, leaving the bias resistor alone.

I haven't found this one yet, but I didn't do a lot of digging. I'll check into it more tomorrow. Why is it better to vary the tail resistor instead of the bias resistor?

My idea was to have a range of bias from say -1V to -2V by putting a fixed bias resistor that brings the bias to -1V in series with a pot to the drop the rest of the way.

[matt h]

(deleted)

[Theashe]

Hey all,

Attached are the simplified diagrams of two ideas I've had about how to lay out the gain stages of my preamp.

The sound I'm aiming for in this amp is a rich, scooped clean that breaks up when you attack notes. I'd still like to leave the possibility of achieving overdrive when I want it.

However, I have this itching suspicion that I might have too much gain for a good clean sound. I'm mostly using 12AX7s for the preamp, but I have a ton of NOS GE 12AU7A tubes lying around that might be a good low-gain alternative.

My first gain stage is a 12AX7 with a DC-coupled cathode follower as described in Merlin's book. The cathode follower feeds an EQ that closely resembles the EQ from a Dumble.

Now, in Design 1, the output of the EQ feeds into a cathode follower that drives the effects loop, which is mixed back together with an anode mixer, also described in Merlin's book. The output of this is divided and sent in parallel through a final amplifier and reverb driver/tank/recovery to the PI. The problem with this, is that when I look at my theoretical gain stages, the only tube that ever has the possibility of getting overdriven is the PI, which might not sound good. Also, there might not be enough gain in the reverb recovery circuit to bring the reverb to a good level. So I came up with Design 2 instead.

In Design 2, the amplifier that went after the anode mixer is moved in front of the cathode follower that drives the effects loop, to get more gain earlier in the circuit. This means it may be possible to overdrive the anode mixer as well as the PI. An FX send control is necessary to bring the signal down for any connected effects. As well, the amplifier I used before the reverb driver has been moved into the recovery circuit to offer additional gain. However, after doing all this and looking at my notes, I don't think I'll be able to get a clean tone unless all my volume dials are down almost to nothing. I have a "pre" volume in the EQ circuit, "wet" and "dry" volumes in the anode mixer, and I plan on having a final volume control just before the PI. If my math is right, the PI can handle about 6Vpp before overdrive.

Could anyone with experience offer some insights into these designs? Bear in mind that I'm only talking about how to organize gain stages to get my desired sound.

[Theashe]

Also, I have a beginner's question about where I've positioned my effects loop and reverb. I've found schematics online that placed the effects after the reverb, right before the PI. I thought it would be desirable to go the other way around - have the sound of effects be included in the reverb. Is that the case?

As well, I've seen circuits with reverb send/reverb return. Can anyone shed some light on the differences, and why you might want that instead of just one reverb level (in the return position)?

As always, I appreciate how patient you guys are to beginners like me.

[EtherealWidow]

I'd like to just say that you're right in that first-hand experience is king. The technical knowledge is extremely useful for getting you in the sonic ballpark initially for the build and very useful for modding yourself in the right direction. I recently just finished my first build and the component values and arrangements are much different than anything I was initially planning for.

[Theashe]

What was your first build like? Design is one thing - parts are a whole other story. I'm on the fence between looking for a chassis or buying an undrilled one and tools to drill it myself. There's a metal shop in my town that might be able to do it, they're not expensive. Getting/making turret board is another story too.

[matt h]

(deleted)