Author Topic: Building a small guitar amp (Read 26282 times)

Torrentula · « **on:** February 17, 2012, 07:05:24 pm »

Hey guys!

I want to build a little guitar amp (nothing fancy, just for the hack of it). I do have a circuit in my mind (and already actually drawn a basic schematic) but I'm still not quite sure about it. Do you think the output RC-filter is not interfering with the output signal of the opamp?

I haven't yet chosen on an opamp so I'd appreciate any suggestions

I have searched the forum and the internet but I can't find any information on how to set an opamps gain with a microcontroller. Do I have to use a digital potentiometer to set the gain or is it possible to set it only with the built-in peripherals of the microcontoller?

Thanks for your help

Torrentula

DrGeoff · « **Reply #1 on:** February 17, 2012, 08:44:57 pm »

An op amp is not going to drive an 8 ohm load directly.
For a guitar amp you need a high-gain high-impedance input for regular passive pickups. Following the preamp stage you can then drive a small power amp into the 8 ohm speaker.
What supply rails are you planning to use? The OPA743 is a +/-6V or +12V device. If you are planning to run the whole thing of 12V then you can always look at car radio power amp chips (eg TDA1519 etc) to make the output drive section easy.
If it is a simple project, why would you even bother with a microcontroller in it? Just use pots to adjust the input and output gains. A microcontroller only complicates it and is unnecessary.

DaveHardy · « **Reply #2 on:** February 17, 2012, 10:05:34 pm »

I've been in the music business for years and have been on a few guitar amplifier design teams. First off, this is just my opinion, monolithic operational amplifiers sound very unmusical for guitar applications - especially in the very critical first gain stage. I suggest looking into vacuum tube technology if you really want to learn about guitar amps; guitar amp designers have been trying to get solid state designs to sound like tubes from day one.

If you are simply having fun and using parts that you have lying around, there is a virtual plethora of guitar related designs that you could borrow from on the net.

Regarding your design, C1 is a bit large, you should make it closer to .01uf

C2 is also very large, and should be closer to 1uf.

Never!! put a gain pot in the feedback loop because if the wiper jumps off of the track, your op amp will peg to a rail.

Not really sure why R3 is in parallel with an 8 ohm load.

To power 8 ohm headphones, I suggest an emitter follower discrete transistor configuration with a very large 500uf dc blocking cap. You could prob get away with 32 ohm headphones without an emitter follower.

741, 1458, 4558, TLO72 are basically the preferred operational amplifiers to employ on guitar designs because of their distortion effects. The older the better!!

Here's a great place to borrow some designs from. Notice how lots of these pedals employ what is called LED clipping and other creative techniques for distortion effects.

http://www.webphix.com/schematic%20heaven/www.schematicheaven.com/effects.html

DaveHardy · « **Reply #3 on:** February 17, 2012, 10:12:26 pm »

I forgot to mention that you can't load you input with anything less than 1M or you won't hear anything:(

electronwaster · « **Reply #4 on:** February 17, 2012, 10:59:51 pm »

It doesn't answer your question directly, but I searched around quite a bit a few months ago looking for a very simple guitar amp schematic, and came across the noisy cricket.

I built it and it's great - I really recommend it. If you have all the soldering equipment and don't mind a plastic peanut butter jar for an enclosure (for example), then you can build it for just a few dollars. It's based around the LM386 IC. The project is very well explained and there are PCB layouts, veroboard layouts (off-site) and so on. I think it's aimed at (and is ideal for) beginners.

I built mine in a hammond-type case the size of a stompbox.

Here's the page: http://www.beavisaudio.com/projects/NoisyCricket/

If you're interested in this kind of thing in general, there is loads more interesting stuff at that site.

Cheers,
electronwaster

amspire · « **Reply #5 on:** February 18, 2012, 01:09:17 am »

I would use an audio amp power amp IC instead of a power opamp. It is designed for the job so you get a soft start-up - no ugly thump from the speaker as the power comes on, and it will be completely stable with the reactive speaker load. Pick a speaker that is more powerful then you need so it can handle the low notes without hitting its limits.

Now small guitar amplifiers will be regularly driven to overload, and solid state clipping sounds really bad. I had always heard how it was the odd harmonics from clipping in particular that were bad, but I thought I would do some calculations to see why this is. So say we take a C note and clip it generating 2nd, 3rd, 4th, 5th, etc harmonics. What notes are these harmonics? (Hope I get the numbers right

)

Fundamental: C
2nd Harmonic: C Pretty good - always sounds great
3rd Harmonic: G Not too bad.
4th Harmonic: C excellent
5th Harmonic: Out of tune E Doesn't sound good
6th Harmonic: G OK
7th Harmonic: Very out of tune A# Yuck
8th Harmonic: C
9th Harmonic: Slightly pitchy D Not good
10th harmonic: Out of tune E again like the 5th harmonic but an octave higher, so for the first time, the even harmonics are going bad
11th Harmonic: Half way between F and F#
12th Harmonic: G

So I guess the story is that even harmonics in distortion are great, as long as there is not much 10th harmonic and above. With odd harmonics, anything above the 3rd is a disaster.

What this means is it is really worth adding a soft limiting circuit before the power amp to make sure that there is never any power amp clipping and to limit the amplitude in a way that boost the lower even harmonics, and minimizes odd harmonics. If you are interested, I can suggest some ideas. Something reasonable can be done with a few cheap parts that may not give the true "valve" sound, but will sound decent.

Richard.

DrGeoff · « **Reply #6 on:** February 18, 2012, 02:25:34 am »

There's been a lot of work done on distortion characteristics. Some sound good and some sound OK and some sound terrible. The worst one is crossover distortion. Avoid it at all costs. Edward Cherry did a demo at the 1984 AES convention (http://www.aes.org/e-lib/browse.cfm?elib=11640)with varying amounts of different distortion types on different types of music.
Harmonic distortion is related to the input signal, the music, and as such it can add to the sound, providing additional texture. The harmonic degrees and amplitudes will determine the type of texture. Crossover distortion, on the other hand, is not related in any way to the music and just sounds awful on all signals.

DaveHardy · « **Reply #7 on:** February 18, 2012, 04:39:17 am »

Quote from: DrGeoff on February 18, 2012, 02:25:34 am

There's been a lot of work done on distortion characteristics. Some sound good and some sound OK and some sound terrible. The worst one is crossover distortion. Avoid it at all costs. Edward Cherry did a demo at the 1984 AES convention (http://www.aes.org/e-lib/browse.cfm?elib=11640)with varying amounts of different distortion types on different types of music.
Harmonic distortion is related to the input signal, the music, and as such it can add to the sound, providing additional texture. The harmonic degrees and amplitudes will determine the type of texture. Crossover distortion, on the other hand, is not related in any way to the music and just sounds awful on all signals.

I disagree. Awful is a matter of opinion:) Some folks like what you might call awful and vice versa. There are several guitar pedals that create crossover distortion in their signal paths on purpose! Personally, I bias guitar amps with a tiny bit of crossover distortion in the output tubes; this is also a matter of opinion.

DrGeoff · « **Reply #8 on:** February 18, 2012, 05:49:30 am »

Of course. From an instrument amplifier and musical signal source perspective any form of colouration is fair game. You can't get that heavy metal sound with a pure amplifier!

amspire · « **Reply #9 on:** February 18, 2012, 06:03:46 am »

Quote from: DrGeoff on February 18, 2012, 05:49:30 am

Of course. From an instrument amplifier and musical signal source perspective any form of colouration is fair game. You can't get that heavy metal sound with a pure amplifier!

But you can with a pure amplifier with a foot switch controlled distortion box connected to the guitar. Any distortion any power amplifier can add can also be done at preamp stage followed by a low distortion power amp. It is not unknown for big bands to be using fake gigantic speaker boxes and valve amps on stage while they are actually using off-stage solid state amps.

DrGeoff · « **Reply #10 on:** February 18, 2012, 06:13:37 am »

Or just use a good power amp and a DSP box like the line6 pods, NI guitar rig etc, that will model a vast combination of guitars, amps, cabinets and effects.

Torrentula · « **Reply #11 on:** February 18, 2012, 10:26:03 am »

Wow that's a lot of information

Well let's see... so how can I achieve an input impedance of 1MOhm on my preamp? Can I use an LM386 as a preamp and then feed it's output into the TDA2040 directly?

Sorry for my noob questions but I'm pretty new to analog electronics.

Thanks for your advice

Torrentula

electronwaster · « **Reply #12 on:** February 18, 2012, 10:35:42 am »

The input impedance of the LM386 is 50K (http://www.ti.com/lit/ds/symlink/lm386.pdf), so it will load down the instrument a fair amount.

You can get around this by having a unity-gain amplifier, such as a jfet transistor in common-drain mode (also known as voltage buffer). The noisy cricket uses a MPF102 in common-drain mode with a 1M resistor to ground to set up input impedance (http://en.wikipedia.org/wiki/Common_drain).

Here is an interesting page on buffers specifically for guitars (http://www.muzique.com/lab/buffers.htm)

Cheers,
electronwaster

Zero999 · « **Reply #13 on:** February 18, 2012, 03:44:45 pm »

I think the OPA743 would be perfect for building a buffer for a power amplifier such as the LM386 or TDA2040.

To achieve lower distortion and noise, you may also want to add some gain to the buffer stage and reduce the gain of the power amplifier as much as the IC you choose will allow you. Power amplifier ICs tend to have better THD and noise when operated at lower gains. Don't go too low though as it may cause oscillation, for example the LM386 has a minimum gain of 20. I don't know about the TDA2040, I haven't read the datasheet.

Torrentula · « **Reply #14 on:** February 18, 2012, 04:56:55 pm »

So taking all you advice into account, I now come to the problem of adjusting the gain, if I shouldn't put a pot into the feedback loop of the amp and if I set the gain of the power amp (LM386) as low as possible to keep the distortion low.

Quote

you may also want to add some gain to the buffer stage and reduce the gain of the power amplifier

How do I make the gain adjustable without using a pot? Doesn't the buffer lose it's infinitely high input impedance when I add some gain with the help of a voltage divider?

Zero999 · « **Reply #15 on:** February 18, 2012, 06:54:06 pm »

Quote from: Torrentula on February 18, 2012, 04:56:55 pm

So taking all you advice in to account, I now come to the problem of adjusting the gain, if I shouldn't put a pot into the feedback loop of the amp and if I set the gain of the power amp (LM386) as low as possible to keep the distortion low.

How do I make the gain adjustable without using a pot?

For the audio amplifier IC you read the data sheet. For the op-amp you look for a tutorial on calculation the gain you require.

Quote

Doesn't the buffer loose it's infinitely high input impedance when I add some gain with the help of a voltage divider?

An op-amp keeps its high impedance when connected as a non-inverting amplifier.

Torrentula · « **Reply #16 on:** February 18, 2012, 08:29:28 pm »

Ok so now I've redrawn the schematic.

R5 and C3 are the tone control and the LM386 is set to a gain of 20 (lowest possible).

I'm still not sure about R4 as the gain control for the buffer

DrGeoff · « **Reply #17 on:** February 18, 2012, 09:00:48 pm »

I don't think you want a pot directly on the input. Most guitars have level controls on them already.
Perhaps you should use a FET input op amp like the TL072/LF353 and make the preamp 2-stage. The first stage is high impedance input with a fixed gain of around 20dB. This is followed with an input level control which feeds the next stage which has a 40dB fixed gain, giving a total of 60dB gain through the preamp when the input level control is set on max. The output of the second stage would then go through an output level control into the power amp stage.

DaveHardy · « **Reply #18 on:** February 18, 2012, 09:18:20 pm »

You're getting there. You've input impedance is still less than 100K - get rid of R2 and replace R1 with 1M. You might also want to put a small cap in parallel with R1 or an ac integrator cap in the feedback loop of the first op amp, probably both for stability. Replace R3 with the same setup as R5 where you've got the wiper tied to one of the pins. R3 should be 500K in series with a 10k resistor. Replace R4 with a 10K resistor. R5 wiper needs to go to pin 3 of the second op amp; this will give you a fixed load on IC2. Also, I wouldn't bother with the fancy op amps. Just use a 4558 A and B section. Pin 2 of IC1 should have a resistor there to eliminate bias current effects. I'm not sure why IC1 doesn't have a feedback loop and don't know much about it.

Zero999 · « **Reply #19 on:** February 19, 2012, 02:39:19 am »

What does R5 and C3 do? Nothing, get rid of them.

C2 is so small the cut-off frequency will only be 20kHz which is above the audible frequency range which is useless, make it 1000uF for a lower cut-off of 20Hz.

Torrentula · « **Reply #20 on:** February 19, 2012, 11:03:48 am »

So here is another revision of the schematic

Quote

C2 is so small the cut-off frequency will only be 20kHz which is above the audible frequency range which is useless, make it 1000uF for a lower cut-off of 20Hz.

Added a 1M resistor on the input:

1 / 2*PI*10⁶*10^-8 = 15.91Hz

So I'm now using the dual stage preamp as suggested, I hope I got everything right with that one.

I'm not sure about what value to choose for R7, in the Datasheet for the LM386 the inverting input is not grounded through a resistor.

Zero999 · « **Reply #21 on:** February 19, 2012, 12:43:27 pm »

You've still forgotten to chance the value of C2.

Torrentula · « **Reply #22 on:** February 19, 2012, 03:33:59 pm »

Okay so now all suggested improvements are made. Still the question remains what value should be chosen for R7?

Tube_Dude · « **Reply #23 on:** February 19, 2012, 04:04:38 pm »

Hi

No need to use R7, it will just increase noise...

Tube_Dude · « **Reply #24 on:** February 19, 2012, 06:52:41 pm »

The pin 4 of the LM386, must be connected to ground, and not to a negative voltage...


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Building a small guitar amp (Read 26282 times)

Share me