Silvia, a 2 channel MOSFET preamp. New version.

[marcelomd]

Hello!

New version of my guitar preamp I've build around a Zvex SHO and a Box of Rock. The first version was an excuse to try a PCB factory. Sounded great, but had a few problems I tracked to faulty components and my soldering skill.

Now I've made a new version, just to try the SMD assembly service. It is beautiful:


Schematic:


I've included a few extra components to try some mods:
- 10k or 5k in the drain of first stage, to account for a 5kC or 10kC gain potentiometer;
- 2x10n capacitors, 510k and 270r resistors to make the '59 Sound version of the Box of Rock;
- 220p capacitors, 33k and 56k resistors to play with the tonestacks.

[marcelomd]

I've also made a satellite board for the 3PDT with a 7809 regulator.

The idea is to use it as a standalone preamp with a regular 9V PSU or include a TPA3118 power amp that takes up to 24V.

TODO: Better pictures tomorrow morning


Schematic:

[marcelomd]

It all sounds really good for my tastes. Pure rock tones.

I have a design question: Is there a better way to handle the switching of the channels? The 3PDT leaves one channel flying. I'd like to ground the side I'm not using, but can't come up with a solution that uses only a 3PDT.

[Steben]

Fender clean stack, marshall post drive stack. Makes sense.
Although I am still in doubt whether a fender stack is the best for my SS amp idea in which power drive follows the stack. A clean DNA yet cranked Fender is not the most beautiful concept.

[marcelomd]

I did a super through scientific research on solid state amps before building this (i.e. googled it).

What the two clean SS amps that receive the most praise, Roland Jazz Chorus and Lab Series L5, have in common is the tonestack right at the front. Gain stage -> tonestack -> gain stage -> maybe some filtering.

A bunch of tube amps with good clean sounds also follow this idea. Fenders, Mesa Boogie Mark series, etc. Bass amps too.

So this is what I went with. The clean channel goes up to a slight overdrive that I like if I crank it. Takes pedals really well.

Actually, the Mark Series looks, to me, like a Fender Bassman -> few more gain stages -> graphic tonestack. Maybe it is something to be explored. Gain channel being inserted after the clean channel. I think Dumble does that.

[POTL]

Soundclips or video on YouTube?

[marcelomd]

Soundclips or video on YouTube?

My friend recorded a demo. Let's see if he let's me post it here.

[antonis]

@Marcelo: Just a delayed note..

Shouldn't be better to scale down and up R29 and C19 values (1k & 22nF say) to make a more "stiff" 2^nd order LPF..??
(in the mean of getting more close to 7.2 kHz theoretical corner frequency..)

[marcelomd]

@Marcelo: Just a delayed note..

Shouldn't be better to scale down and up R29 and C19 values (1k & 22nF say) to make a more "stiff" 2^nd order LPF..??
(in the mean of getting more close to 7.2 kHz theoretical corner frequency..)

I don't really know. This filter is present in several MIAB, including the Box of Rock that inspired this. There's no need for more volume so I just used the same values.

[antonis]

 
I'm not talking about "volume"..

Two directly cascaded 1^st order (-6dB/octave) LPFs (or HPFs or MPFs) of identical R/C values DON'T result into a 2^nd order (-12dB/octave) LPF of the same corner frequency due to first filter loaded by the second one..

But, if it sounds good as it is, let it be..

[marcelomd]

Finally got a chance to test this thing in a real environment.

Good news and bad news:

Tested it against a Mesa Boogie Mark V, a Marshall Jubilee, Mesa Boogie V Twin, Blackstar HT Dual and a Tech21 Tri AC. Sounded so good and easy to set up.

...And I totally forgot to record anything =\ *facepalm*

[antonis]

...And I totally forgot to record anything =\ *facepalm*

Bad news, I presume..

[aron]

Where can you order these populated circuit boards? What is the cost? Looks great!

[lcv]

What could dislike about these  BOR-like input stages is they have a  quite low, variable  input impedance,
about 15kohm  @gain max,  300kohm @gain min,
too much interacting w/pickups IMO for  a generic preamp

[antonis]

What could dislike about these  BOR-like input stages is they have a  quite low, variable  input impedance,
about 15kohm  @gain max,  300kohm @gain min,

Only for Zener in "active" (reverse-biased) mode..
(which shouldn't be the case here..)

[marcelomd]

Where can you order these populated circuit boards? What is the cost? Looks great!

I ordered them from https://jlcpcb.com/. They have a huge library of available parts here: https://jlcpcb.com/parts

I don't remember the exact costs because these weren't the only boards in that particular order, and I used a discount code (they have that regularly) but they were something like  $5 for the boards and less than $20 for the components+assembly.

What could dislike about these  BOR-like input stages is they have a  quite low, variable  input impedance,
about 15kohm  @gain max,  300kohm @gain min,
too much interacting w/pickups IMO for  a generic preamp

Not ideal, but they sounded good anyway. On par with the big boys =)

[lcv]

What could dislike about these  BOR-like input stages is they have a  quite low, variable  input impedance,
about 15kohm  @gain max,  300kohm @gain min,

Only for Zener in "active" (reverse-biased) mode..
(which shouldn't be the case here..)

no, just plain ac small signal simulation, zener omitted,
transient  analysis confirms those values ( up to output clipping , when feeedback disengages)
would be nice if someone could doublecheck

[antonis]

Quite right..!!

Didn't realize 1M feedback resistor..
(thought of been tied up to Vcc)