Hello all and circuit design questions

[jimboyogi]

Hello everyone,

I've been reading around here for a while and picking up some great information, so thanks!

I've been toying around with electronics since childhood, mostly solid state stuff, but as I also play guitar I have long held an interest in valve/tube circuits and sounds.

After spending a chunk of life on career training and starting a family, I have returned to both guitar and electronics. I started on germanium fuzz pedals, then graduated to building my own version of a 5F2a fender princeton, which I have recently completed. So now for the next project.

My theory is very sketchy, but I am finding the internet a great place to learn, and I eventually stumbled across this here forum, and have been a regular reader since. So please forgive me when I repeatedly ask stupid questions, hopefully my questions will gradually become less stupid. Maybe one day I'll be able to answer someone's question

Recently I was cleaning up my parents shed, and came upon a few cardboard boxes of used valves . Turns out my father had bought part of an old valve powered computer, and had pulled a bunch of valves from it.
So I set myself a new project - to design and build a great sounding guitar amp just using this stash.

Before this find I had been planning to build a souped-up Vox AC10, but I am hopeful that I can get vox-ish enough sounds from this new build.

The Found Valves - mostly Brimar - are

12AU7 - twin triode

ECF82/6U8 - triode & pentode - as used in vox ac10!

12BH7 - twin triode

6AL5 - twin diode

There are at least a dozen of each of these.

I want to keep the design very simple (at first), but with 2 input channels with different tones (that can be jumpered), bright switches and volume on each channel, master top cut and master volume.

So I'm thinking to use both sides of the ECF82 for the inputs, one triode and one pentode, along the lines of the normal channel from both the vox ac10 (triode) and the ac15 (pentode - ac15 used an EF86)

Feed both of these to a long tailed PI using a 12AU7, past a top cut and master volume to a push-pull class AB 12BH7, as per the "Moonlight", giving around 5 Watts output.

So now.....

Questions

1 - I am contemplating using the 6AL5 as the cathode bypass on both the input stages. I have lots of them! But is this worth doing, or should I stick with standard R & C setup? I like bass, and want to fully bypass these stages, but I would like a tight response, and have read that diode bypassing might give the best of both worlds.

2 - How much gain do I need? I mostly use single coils, and I only want the amp to get into 'overdrive', not high gain territory. Triode stage gain is ~ 40, PI is ~ 15. Is this enough, or do I need to add another gain stage pre-PI?


Thanks in advance if any of you can help me out. Once I get these questions sorted I'll put up my circuit for advice/criticism.

Cheers

[tubeswell]

Hi jimboyogi

If you can post a schematic of what you have in mind, then you may get some specific and useful comments. Otherwise there's too many variables (which works against typing short replies).

Generally-speaking for clean sounds, the fewer gain stages, the purer the guitar tone will be able to be at cleaner settings. Otherwise more padding is required in the signal path to get the gain right. Have a look at Merlin b's Valve Wizard site for explanation about setting up triode gain stages.

[jimboyogi]

Thanks tubeswell,

I'll try and get the schematic up tomorrow.

As for Merlin's Valve Wizard site, I have been poring over that for a week, it is an amazing resource. That is actually where I found the idea for using the
6AL5 diode as the cathode bypass. Does anyone here have experience with doing this?

Cheers

[Merlinb]

Originally Posted by jimboyogi

Thanks tubeswell,

I'll try and get the schematic up tomorrow.

As for Merlin's Valve Wizard site, I have been poring over that for a week, it is an amazing resource. That is actually where I found the idea for using the
6AL5 diode as the cathode bypass. Does anyone here have experience with doing this?

Cheers

Hi Jim,
I don't think there will be many people who have tried this. It's more of a novelty than anything really useful. Not much different from using a resistor, usually.

Anyway, there is something much cleverer you can do with a 6AL5:
Connect the two diodes in anti-parallel, with a 1M (or so) pot in series, across the output valve grids. This forms a cross-line valve diode clipper, which allows you to get some pretty decent overdrive at any volume level. Poor man's power scaling if you like...!

[jimboyogi]

Thanks Merlinb,

Thanks for your reply and information, that is a cool idea about using the 6AL5 as a clipper. This time I'll leave it out of the circuit and gain myself some heater current.

I have attached my current iteration of the circuit, I have still got some values blank, esp. in the PS, as I'm not sure yet which PT i'll be using. I'll probably run the HT fairly high to get the most headroom from these valves.

I've put in a false heater CT, but I'm not sure where to connect it in a PP circuit, it went to the output tube cathode in my SE Princeton.

Any feedback welcomed

Cheers

[jimboyogi]

Any thoughts?

[jimboyogi]

I've got a copy of Merlin B's book heading towards me in the mail

But does anyone here know a good resource (ie. novice language) for explaining how to design push-pull OP stages with triodes?
All that I have found so far concerns pentodes, and because the grid curves behave so differently with triodes, am I right in thinking that a different approach is required?

I sorted my stash of tubes today, and I have s**t loads of 12AU7 and
12BH7's, so I'd better get cracking on designing some uses for them.

Also, can anyone tell me the average peak-peak voltage output from a fender single coil pup? Is it ~ 0.5V?

As always, any help appreciated.

[Robert M. Martinelli]

Originally Posted by jimboyogi

Also, can anyone tell me the average peak-peak voltage output from a fender single coil pup? Is it ~ 0.5V?

Nope, it's more in the 120-150 mV range IMHO, provided the load impedance isn't too low.

As you might already know Fender ( standard Strat ) pickups are not built to be high output pickups, they tend to be sparkling and clean as their rather low inductance and low stray capacitance make for moderate output and high resonant frequency, while their rather low winding resistance, ( together with the above L and C ) makes for a higher Q ( narrower bandwidth, so the pickup is more "focused" around its resonant frequency ). Also, don't forget the cable adds a high frequency roll-off due to its stray capacitance ( some 100-110 pF/m in a good quality cable ).

HTH

Best regards

Bob

[Steve Conner]

Designing a push-pull output stage with 12AU7s or 12BH7s is easy, because the choice of output transformers is so limited. You need a tiny push-pull transformer with a rated load impedance of about 10 to 20k a-a. There are only two that I know of, one from Hammond and another from Oxford Electronics, so you basically pick one or the other, hook it up and hope for the best.

Generally, designing with triodes is easier than with pentodes. To get maximum power, the load impedance should be roughly the same as the tube's plate resistance, and that's given on the datasheet. For push-pull you may need to multiply by 2 for Class-A, or 4 for Class-B, hence my recommendation of 10-20k above.

Mark Lavelle's Decimator used a 12BH7 with the Hammond 125A transformer:
http://www.harmonicappliances.com/de...Decimator.html

And I've experimented with the Oxford Electronics one, the N35A002F.

[jimboyogi]

Thanks guys,

So for 12BH7, plate resistance = 5.3k, so in pp AB1 i need OT impedance of ~ 21k. I've got a multi-tap OT that I could push up to a 16k primary using an 8 ohm speaker on the 4 ohm secondary tap. Would 16k be close enough?

Or if i parallel 2 * 12BH7 (like I said, i've got s**t loads of em), i'd only need ~ 10.5k impedance, and could get away with 8K?

[Steve Conner]

Yes, it should be fine. The tubes are low-powered enough that there's practically no risk of damage from mismatching. You can experiment with the OPT taps to see what gives the best volume and tone.

If you have lots of 12BH7s, why not parallel a few up. If the OPT is "too big" then several tubes in parallel will give more current to magnetize it, so you'll get more bottom end. The only risk is that it might get too loud for practice at home...

[jimboyogi]

Thanks Steve,

My biggest limit with the 12BH7 is that they draw 0.6 A of heater current each. Every PT that I've got is only rated for 2A heater current, which only leaves me (just) enough current for 2 of these, leaving 0.8 A for the preamp, PI and heater CT.
2 * 12BH7 should give me around 10W output, heaps for what this amp is intended for - clean practice at home and a bit of drive with a band.

With regards to designing the rest of a pp output stage using these tubes, it may be easy, but i don't know how! That's why I was wondering what the best resources to learn are (websites, books, etc).

[Steve Conner]

My point was that you don't really need to know how, because you only have a choice of two suitable output transformers on the market. Although if you parallel up a couple of 12BH7s, you could probably almost drive an OPT meant for two EL84s.

The classical method of finding the OT impedance is by drawing load lines on the valve's Ia vs. Va characteristic curves. This works as well for triodes as for any other kind of tube. I'm sure Merlin shows how to do it on his Valve Wizard site.

The Valve Wizard -Push-Pull

[jimboyogi]

Thanks Steve,

Got Merlin's book and doing some learning, I've already made quite a few changes to my above circuit based on advice from the book.

The output stage I'm still coming to grips with, esp. selecting values for cathode and grid resistors. I'll post when I have some quality questions.