Roast my preamp

[ECV02]

So I have this abomination of a preamp to bring to the table today. Title says it all; roast it hard. This will eventually feed into a parallel single ended 12BY7 power amp, for about 5w of power.


Some things I am trying to avoid:

Fuzzing out at high gain (hence the placement of the tone stack).

Overly bright, "jangly" sounds,

Dirty "cleans".


I know you can't tell all these from simply looking at the schematic, but there are people that are far more experienced than myself here, and I figured it would be a good idea to put it up before I solder it up to (hopefully) avoid things going sour.


Any help would be greatly appreciated.
Thanks,
E.

[sluckey]

Looks too dark to me.

[ECV02]

I was thinking that the 470p plate cap could go, but I wasn't quite sure. One of the most common complaints I hear about cold clippers is them spitting out unwanted high frequencies. I guess that's why we add a treble control to our amps in the first place.

One other thing I was wondering about is the bass response of the first stage. As of now there is very little rolloff, the input cap + biasing has a frequency cutoff of roughly 23Hz. I am slightly worried this could lead to fuzz at higher gain, so decreasing the input cap's value (say, to 22n) is an idea as well. That would lead to a rolloff at 100Hz.

Hmmmm...

[d95err]

Some thoughts:

No bass roll-off... this will sound ”farty” if pushed to a lot of clipping.

Low gain... Will that transistor have enough gain to even get the cold clipper stage to clip?

The low input impedance of the first stage could be a problem, especially if what comes before it is a guitar and long cable, or unbuffered pedals.

One option is to use an LND150 FET instead, so you can keep a more ”tube”-like input stage.

[PRR]

Looks too dark to me.

Try it the other way:

[PRR]

....roast it hard. .....

Agree the input impedance is remarkably low. Q1 bias will change significantly as the gain pot turns; does it stay operational? (If not, there is another way.) I'd have to find a pencil/sim to know where distortion happens. The 33k in the tonestack leaves little range of action, and is an unusually low Z for the 12AX7 to drive, especially cold-biased. The thrift in me says you could save three parts by omitting the V2 cathode cap and also the 2:1 pad behind it.

470p+100k is 3.5kHz, maybe 5kHz with tube in action. Not going to be shrill. Easy to adjust to taste.

[sluckey]

   I knew you would understand.   

[ECV02]

LOL! Thanks PRR... guess that's what happens when the terms we use to describe tone coincide with other words .

I've made a few changes as per your guys' words. On the topic of input gain, I'm going to lower the emitter stopper resistor to 1k, this will definitely be able to send the cold clipper into clipping. To deal with the input impedance, I'm going to double up on the biasing resistors (8M/150k). A quick sim and number crunch later, this puts the input impedance at around 60k at the highest gain setting. While not ideal, this is at the highest gain setting, and it will only get better at lower gain settings. To add some bass rolloff, the HPF between Q1 and V1 will be dropped to 14nf/100k, this will put the cutoff at 113Hz, which should be enough to prevent farting.

To answer your question PRR, Q1 should stay operational across the whole pot sweep, thanks to the roughly 200v (and thus massive headroom) I've allocated to this stage. In fact, at maximum gain, the input headroom of this stage is 1.5v peak, way more than a guitar, even with a boost pedal, can kick out. This was done purposefully to avoid first stage distortion, which sounds nasty in this context. The tone stack is ripped directly from Marshall's classic design, and because it is being driven by a high impedance source, it should become more responsive.

Feel free to correct me on anything I've said, I make a lot of mistakes 

Thanks,
E.

[Tony Bones]

Can't drive a BJT with a guitar.
Can use a FET.

[d95err]

Can't drive a BJT with a guitar.

Hundreds of well-known pedal designs says you can...

However, with the low input impedance, it will be interactive with whatever goes before.

[PRR]

> Q1 should stay operational across the whole pot sweep, thanks to the roughly 200v

The pot adjusts the AC gain. It is not needed for DC bias. Taking the bias outside the gain pot gives a far flatter collector voltage; also increases the audio input impedance except at maximum gain.