Help with valve datasheets - 12AU7s/12AX7s

[snowballfight]

Heya I'm new to this so I hope this isn't too stupid a question.

I'm designing a roughly 10-20W, all valve amp from scratch. I'm a Mechanical engineer by trade, but thought this would be a great opportunity for me to flex some design and manufacture muscles I haven't in a while! BUT I'm having problems now, starting to source components and the like.

Attached you can find a simple schematic of my preamp stage as it stands.

Found MAX and TYPICAL data for the 12AU7s i'm putting in, but can't find anywhere near the detailed tables like the one I've posted as an example. How do I use the graphs they often give? I want to know how to vary and quantify the plate, coupling and cathode resistor/cap values when I'm varying the valve voltage.

I've attached an image to clarify what I am naming there...

Anyone got any ideas as to where to look for these?

I'm planning on using JJ Electronics 12AU7s and so on for now as I get the circuitry down and then switch for something better. Heard reliable things about them.

Hope someone out there can help!

[BackwardsBoB]

You are using one cathode resistor for two tubes, and again for one tube (two triodes). Sometimes people will use two triodes in parallel with common cathode and plate resistors and common grid signal, in order to double gain and cut noise in half. This is usually on the first stage. In your case, you're better off with separate cathode resistors for each of the first two tubes and for each of the last two triodes, perhaps bypassed with a capacitor for more gain or to boost high frequencies.

Your feedback adjustment provides positive feedback, which is very bad. Both of the previous stages are inverting.

Your grid input to the second tube isn't ground referenced. 220K to GND would help.

Your driver selector floats the "A" triode's grid when that triode is bypassed.

I think it makes sense for you to start with some preamp circuit you like, and modify from there. Component value changes alone can greatly affect the result, yielding plenty of things to experiment with.

You can go to How to design valve guitar amplifiers for some basics, or buy Merlin's book, which is quite good. It'll tell you how to get useable information out of the tube performance curves.

Linear Tech gives away a great SPICE simulator, LTSpice (Linear Technology - Linear Home Page). It even comes with a full amp simulation. It's easy to use for a simulator, and you can find more tube models at Duncan's Amp Pages, along with a tone stack response calculator and power supply design program.

Tube amps are simple devices, but only experience will tell you the sonic effects of changes, and lots of things like component power dissipation, voltage rating, startup conditions, instability, failure modes, safety, etc. take a while to wrap your arms around.

[snowballfight]

Thanks BackwardsBoB,

I think the problem here is that I have referenced other designs without properly understanding what is going on! Thought I did. I'll have a good look at your website you suggested, looks to be a great help.

[mooreamps]

All the comments stated I would tend to agree with. I like the you think, but most gain stages are set up in certain ways given the nature of working with high impedance circuits.

-g

[Alf]

It even comes with a full amp simulation. It's easy to use for a simulator, and you can find more tube models at Duncan's Amp Pages, along with a tone stack response calculator and power supply design program.

Sorry for my ignorance but I downloaded the programme LTSpice but don't see the full amp simulation anywhere !
What's it called in the index ?

Tia, Alf