schematic up for scrutiny

[makepeace]

Hi tube wizards

I've drawn up a draft schem for an amp that I might want to build soon.



The triodes in the preamp will either be 12AX7's or 6SL7's and the pentode will be an EF86. Morph is a triode/pentode switch. Haven't attributed any values to the preamp stages as I haven't worked them out yet!

The intention is to get as good as I can cleans out of the 12AX7 dominated stage, with some grit pulled in with a boost switch that'll add another appropriately biased triode gain-stage into the chain. The EF86 containing stages will be the main drivey channel, also with a switchable triode on the end.

There is also a switchable and adjustable cathode bias/fixed bias in the power stage (need to find some pots with a high power rating).

Would love it if anyone could have a glance and give some criticism, especially as regards the preamp and PI stages. As mentioned in some of my previous posts, I have a VERY limited knowledge of the theory involved, so please don't be too harsh! I am busy reading and trying to get to grips with Merlin's tutorial on designing triode gain stages

Thanks in advance,

Jono

[Zozobra]

Its hard to comment without values but I noticed that you have bypassed a lot of the stages which is going to be quite gainy! You might want to consider limiting the boost in places if you dont want lots of gain or perhaps omit the bypass.

[soundguruman]

There is going to be one heck of a POP when you flip those selector switches.
Suggest bypassing the switches with large value resistors to de POP the POP.

[makepeace]

Thanks for the responses Zozobra and soundguruman!

Get you both. Zozobra, do you think I should go as far as removing all bypass caps on the TMB channel, and maybe the one on the last triode of the EF86 channel? Also, could you elaborate on how to limit the boost gain-stages?

[Zozobra]

Hard to say. Sometimes its best to try it and adjust as necessary with one eye on the maths to make sure nothing dumb happens. I'm the first to admit guilt of handwringing and more often then not I should just try stuff. I have just bought the stuff for a tinker box to play with though to rectify this.

To limit the boost from the bypass cap you add a resistor in series with the cap to ground. This is useful if you want a slight boost as just banging a fully bypassed stage in will make the signal much bigger and thus louder if you've not hit the limit. Also you can use this as a gain boost in its own right if you take an unbypassed stage and then add a limited boost bypass to it.

What sort of sounds do you want from the clean channel? Also you might need to add a gain pot after the first stage as otherwise it will be pretty dirty. My apologies if there is one and I missed it.

[makepeace]

Ooh yes, read about that last night actually! Thanks! I'll first try it with and without cathode bypass caps, and will then fiddle around with partial bypassing depending on how things go. You're right, there's no gain pot.. I'll put one in and then see which resistances work well and maybe replace with fixed resistors once I'm happy.

To be honest, I'm not actually sure what sounds I want from the clean channel.. 1st prize would be probably to have the versatility for smooth, jazzy, big upper midrange tone as well as a scooped fenderish bite with single coils. I'm more or less mucking about with this though, which is why I've included so many adjustments.. I There's a lot options in the design that should have a big bearing on the tone.

Re the PI, I've had a look at the 18W Marshall schem, there the two channels are fed into the grids of each of the long-tail PI triodes. What would be the difference between that and what I've drawn up? Also, is my volume control in an appropriate position? (just before the PI?) Or maybe I should diss the Boost in the TMB channel and use that triode as a separate input for the EF86 stage, and then have separate vol controls.

Thanks for your help!

[J M Fahey]

Sorry but besides actual parts values, which neither we nor you know yet, it won´t work because you omitted *lots* of parts which are indispensable to make any tube circuit work.
For one, grid bias resistors which establish a DC ground path.
I congratulate you on your enthusiasm, but think (others may disagree, of course) that you´d reach your goals better by going step by step.
I mean , say, building an output stage first and debugging it.
Why?
Because in our trade, besides the very important Lab measurements, *sound* is paramount.
So if you build a working output stage first (any power, 15W will be ample), hook it to a good guitar speaker (your choice) and *then* add, say, a single triode stage, plug your guitar and play.
Experiment at will, and only after you are happy with it, add another stage (triode, pentode), tone controls, switching, and so on.
Step by step you will build a very complex amp, but where each time you will have only one new variable added, and earlier ones already worked out.
If you build something new, complex, untested, you will get crazy trying to solve any problem because you will never know where it came from, too many unknowns at the same time.
Good luck.

Note: PLEASE don´t refer to us as "Gurus" and such names; I at least feel it slimy, and I think I´m not the only one.
Thanks.
PS: if in doubt, I use the British definition, not the American one.

[makepeace]

Thanks for your words J M Fahey. I still haven't replied to your pm, as I would like to give it some proper time, but I will. And I'm sorry that you've taken my tongue in cheek greeting to heart, absolutely no malice and nothing but respect intended. I don't particularly care for the British vs American definition, firstly as I'm from South Africa and secondly I wasn't expressing the explicit definition, which was obviously not easy enough to distinguish. But sure, I'll refrain from such behavior in the future.

I've been slightly confused about your comment regarding grid bias resistors. Do you maintain then, that a circuit such as this won't work either?



Are you talking about grid stoppers? Or pull down grid resistors? For eg, the 2nd triode of channel 2 does not have either..

As far as I know, and its quite plain to see, all I've really done is borrowed here and there from established designs and put something together, and I was pretty sure I got it down right, but obviously not.

As far as your suggestion, I can see the merits in breaking this build down like you've described, and I'm definitely going to consider it. First though, I'd like to have a plan of what I'll be doing, and as part of that I need a circuit diagram representing an incarnation of the finished product that I can reference and change as I go along.

Here is v0.1:



Changed the whole thing around, this is probably better. Also, took some stabs at the dark and hastily added some values in.

edit: I see what you mean now re the grid resistors Mr Fahey. Elementary error and thanks for helping me learn something.