Hi,

I have recently began to copy (and modify) a guitar preamp design to build myself. I'm starting from a Fishman Loudbox Mini preamp, a rather well known company in the music business. The circuit looks like below.

There are a few things I want to change, like add a 3 knob mid (gain, freq, Q), alter the center frequency of the Low, get rid of phase, add a fixed Q notch before the Low, etc. Anyway. First step is to model this in LTSpice, something I just started learning yesterday. I got up to the GAIN section modeled and it works. See below. I'll continue on with the rest in the coming weeks.

But from a simulation standpoint, I'm a rookie. Mechanical Engineer by day, electronics nut by night. Any advice you can give I'd appreciate it.





Output of U1B: Looks like a HPF somewhere.
Maybe C5 C6 R7 R9, calculates to about 60 Hz. Peak being 20 dB, -3dB is about 60 Hz.


It would be nice if I had a dual pot model instead of 2 singles.

 

High pass frequency is generally about 1/(2π√(C5C6R7R9)) which turns into the more familiar 1/(2πRC) you use equal component values.
The Sallen & Key section here:
http://sim.okawa-denshi.jp/en/Fkeisan.htm
will calculate the exact responses for you, and calculate the component values for any frequency you might want.
R8 is superfluous - you can leave it out.
I assume that VR1 is pcb-mounted. It is a neat trick to use a stereo pot with both tracks in parallel, because a stereo pot has 6-pins to attach it to the board, whereas a mono pot only has 3 in a straight line, hence it can bend with respect to the board.
Just replace it in the simulation by a pot of half the value.

 

Any advice you can give I'd appreciate it.

Welcome to Spice-world. With simulations of complex circuits it is often worth using a divide-and-conquer approach. Identify functional stages and simulate each stage separately before combining them.
LTspice's Help, although succinct, is worth perusing.

 

I assume that VR1 is pcb-mounted. It is a neat trick to use a stereo pot with both tracks in parallel, because a stereo pot has 6-pins to attach it to the board, whereas a mono pot only has 3 in a straight line, hence it can bend with respect to the board.
Just replace it in the simulation by a pot of half the value.

Yes I'm familiar with dual gang PCB mounted pots. I've used duals pots to change a 1st order filter into a 2nd order. Good tip on using a single 50k instead for simulation. But why is a dual used for the GAIN instead of a single 50k in the original circuit?!?

 

With simulations of complex circuits it is often worth using a divide-and-conquer approach. Identify functional stages and simulate each stage separately before combining them.

Yeah a friend of mine suggested the same thing. My problem (with regards to this circuit) is I don't necessarily understand how the stuff before or after a specific section would affect the section that I might simulate. For transient stuff I can see how the simulation might take a while. But for an AC sweep or DC op, simulation speed shouldn't be a concern. But in general I understand your point. THanks.

 

Yes I'm familiar with dual gang PCB mounted pots. I've used duals pots to change a 1st order filter into a 2nd order. Good tip on using a single 50k instead for simulation. But why is a dual used for the GAIN instead of a single 50k in the original circuit?!?

Mechanical stability on the pcb.

 

Looking at the High Pot:




1/(2*Pi*C15*R18)=4200 Hz. Sure seems to start way before 4200 Hz. It's a 1st order. Having a gentle shelf like this is probably fine for a High knob, but I'll probably want to shift it a bit. Am I thinking about that calculation right?

I think ideally I'd want something like this: starting at 2k and shelving at 12k, give or take.

 

Here's the LOW pot, obviously HPF is influencing it.


Seems to reach up higher than I'd like. Ideally would look something like this (ignoring the HPF)

 

Here's the mid. +/- 6dB centered at 766 Hz = 1/(2*PI*sqrt(C13*R16*R17*C14))

Probably not a bad spot for a fixed mid. Next step is to make it a 3 knob mid.

 

Here's my first pass at a sweep-able mid, with a fixed Q of 0.9 using a dual 10k pot. Q of 0.9 = 1.5 octaves.
It sweeps from 115 < fo < 850 Hz which seems like a good range for an acoustic guitar.