Another guitarpreamp design using an oddball tube

[Alf]

Because I am dissatisfied with current production preamp tubes I am trying to design a preamp using NOS tubes with better quality.
There must be more and cheaper NOS tubes which are not ( hardly) used in preamp guitaramps.
So I had a look at the ecc85 tube (6AQ5) which can be bought for reasonable prices.
So far I have come with this : See image above






It doesn't sound bad and is capable , when you turn up the first volume pot, of a reasonable crunch with some nice compression.

But I have come to the conclusion that it has too many midfrequencies and too little sparkle. How can give me some advice how to change this for the better ?
Could it be a matter of impedance ?

Alf

[loudthud]

I think you meant 6AQ8, a dual triode with a mu of 57. Too many mids? Adjust that 25K in the tone stack downward and the 33K upward. No sparkle? The 470K resistors on the input of the second and third stages will roll off highs, try lowering the value or a small cap of maybe 100pF across the series resistor.

[Alf]

I think you meant 6AQ8, a dual triode with a mu of 57

Stupid mistake , yes , that's what I meant !
Thanks for your remarks Loudthud , I will give them a try and see what you mean. They will defenitely add some more sparkle.

What my main mistake is I think is, trying to get two quite different sounds out of two tubes. What I would like is : a good sparkling clean channel and a fluent crunch and lead sound lead .

Maybe I have to change my approach and add another tube and tonestack for the distortion channel. And change some caps around it in order not to get a lot of fizz.

[Alf]

So I changed the 470 K resistor after 1st gainstage to 180 K but didn't like it. So I put back the 470K.
The slope resistor has now become 56 K which is much better. I think I will have to lower the output pot to 250 K ( there appeared to be a 1 meg pot which just isn't right) to make the taper more even.
Then I noticed another big mistake , I had put the 1N cap which should been on pot 1 on the output pot (2).
After having changed this the amp became much brighter and transparent. I like it a lot now but that's only my first impression. This might change when I give a few days of playing.
I now like it best with the cathode caps switched to the lowest value.

What I have proven to myself at least is that you don't necessarily need an 12ax7 to make a usuable ( good is a little too early) preamp.

[Chuck H]

Gain is gain! Most 12ax7 stages have A LOT of attenuation. Even in cascade amps. Usually earlier but certainly anything past the second stage must be attenuated to avoid the consequences of TOO MUCH gain. Enter the lower gain tube! Fewer or smaller dividers = less series resistance and noise! Generating gain you have to throw away does have some sonic consequences sometimes.

What is that 10n cap on the bottom of the volume control doing?

[Alf]

What is that 10n cap on the bottom of the volume control doing

It's a remainder of a previous version: it was an attempt to make the distorted sound a little thicker. I'll have to see if it still serves a purpose. If not I will remove it , also from the schematic.
The whole thing is in an experimental phase.
To switch between a good clean sound and a nice distorted sound I will have to make a circuit that influences the two pots at the same time.

[Chuck H]

It doesn't appear to be a channel switcher. I'll share a neat trick I have used on single channel designs. I use a dual ganged pot for the volume control. One gang is obviously volume. The other is used to change a parameter that voices for better distortion. That way the amp adjusts itself as the gain is increased.

[tubeswell]

FWIW there's a bit of discussion on the merits of an ECC85 here. Advantages of ECC85 over ECC81? - UK Vintage Radio Repair and Restoration Discussion Forum

(As an aside, its interesting to see other tube enthusiasts make distinctions between ECC83 and 12AX7, and ECC81 and 12AT7. They obviously ain't your typical geetar-amp-variety tube rollers)

[Alf]

I use a dual ganged pot for the volume control.

I had been thinking about that idea ! But aren't the dual pots always the same value ? What if you need different values , I mean e.g. 250 K for output and 500 K or 1 M for distortion ?

[Alf]

Thanks Tubeswell for that link ! It gives a lot more additional info about the ecc85.
This quote below is of much interest to me.

DangerMan is quoting from the same book as I have. It states that the input impedance is higher, then gives figures which show that it isn't! Its scales like frequency^2, so at 100MHz an ECC81 would be 3.5-6.25k, compared with ECC85 at 6k. "Higher slope" is true, but only just: 10% better? So the screen is the only significant improvement.

They are mainly talking about the use of the tube in FM (stereo) radios and that's the usage I was acquainted with. I have never seen it used as a preamp tube in guitar amps.

[cooldude666]

quad gang wired in serial or just open it up and put in a different wafer.

[Alf]

quad gang wired in serial or just open it up and put in a different wafer.

Didn't know there were quad gang pots but using a different wafer could be a good idea.
And what I realize just know is that RG Keen might have some ideas in his article about "the secret life of pots ".

[Chuck H]

Sometimes you can find dissimilar ganged pots at salvage sites. And there are certain dual pots that you can change the wafers on. Not all. Alphas are usually destroyed in the effort. Unfortunately adjusting pots for different values can't always yield a useful taper. I'm not looking real hard, but a dual gang pot I would love to have for a particular project would be a 100k/1M with both gangs in an A10 log taper. But in the event that I never find that pot or a way to make it I have redesigned the circuit so I can use a dual 1M pot instead. Not sure what the additional impedance will do otherwise. That's probably why I haven't tried it yet. But the point is that you can usually redesign for available parts. For example, if you have a voltage divider with two 220k resistors and you want to bridge a cap over the dropper as the volume is increased, You don't need a 250k/1M pot. There are a couple of ways to handle it. One would be to change the voltage divider to a 1M series and 1M dropper. Same voltage division and now you can use a 1M/1M pot. Or you could increase the dropper to 330k and parallel a 1M gang with it for a total resistance of roughly 248k. Again, you can now use a 1M/1M.

[Alf]

Or you could increase the dropper to 330k and parallel a 1M gang with it for a total resistance of roughly 248k. Again, you can now use a 1M/1M.

That is really helpful Chuck !

What I really would like to try is to make the clean and distorted setting switchable (using a relay) but don't know exactly how to proceed with that.

Maybe you have some ideas for this problem as well ?

[Chuck H]

I haven't designed much with channel switching. My limited knowledge is this... If you have extra filament current available, this is the best place to get your power. If you are actually switching channels it's best to not only change the final preamp signal path, but also lift the input to the unused path and/or ground the output to that path. Different levels of conscientiousness on this will be needed depending on the circuit to eliminate channel bleed. For quiet switching it's best to have both switched nodes at similar DC potentials. A floating capacitor, for example, will be carrying a charge that will make a sudden change in the signal path when switched into play. It helps to have a 0VDC reference at switched signal nodes to reduce this effect and avoid switching POP!s. Do include a separate volume control for the distortion channel that is activated by the switching so that the two channels may be balanced. That's about it. There are a lot of excellent examples in known circuits. And it should be mentioned that a lot can be done with SS gate type circuits that make relays seem clunky. But I've never used them for lack of experience and knowledge of these circuits. But there are also good example of this in known circuits that may be copied rote. Getting voltages for transistors and relays is often creative. If the filament supply isn't sufficient a small transformer may be added. And, of course, if you're designing from scratch there are A LOT of options now for PT's that have appropriate taps to power both tubes and transistors at the same time or, if only relays are used, two separate filament windings that may be used to isolate the relay supply for more ideal operation. I will also say that in my limited experience it's best to design such an amp from the ground up or copy a known design. Changing an existing single channel amp into a switchable two channel amp presents extra problems. You need to plan where the switching will be and design the circuits to that end. Otherwise you end up making dramatic changes to an existing amp. And it's generally harder to change something that it is to build it.