Help designing a guitar amp

[teemuk]

Hi all,

I don't know whether this thread is in the wrong place or not but musical instruments forum seems a bit dead so i decided to post here...
I have been designing this low power amplifier for guitar and the design is now in the part where i'm open for suggestions of improvements. There are also some issues to which i need clearance. This circuit is mainly a test to see what kind of tone some circuit's will produce since my research has shown that most of solid state guitar amplifiers that have a reputation of having a good tone have 1. an active tone shaping stage, 2. quasi-complementary power stage, 3. smallish decoupling caps, 4. no negative feedback from output to previous stages.

Also, there seems to be a lot controversy about the amplitude range of guitar pickups. The common belief is that they produce a peak voltage of 100mV (peaks) and when the note decays the amplitude is typically about 30mV.
My research has shown that some pickups can easily do 400mV peaks and some effect pedals even put out a line level signal. 1V is quite much when compared to 30mV. When i tested my guitar's pickups by plugging them straight into mixer's input they were almost equally as loud as a line level signal. What is the truth here? To what peak voltage you people usually base your design into?

Another design issue is the frequency curve which is not flat in guitar amplifiers. Since the slope's centerpoint in the curve is usually located around 400-600hZ the amplitude in 1kHz (the typical measurement frequency) can easily be around -12dB less than the amplitude in for example 2kHz or 80Hz. Doesn't this lead into awful amount of distortion if you base your design on 1kHz sine wave, especially with 60mVpp signal? What methods you people use in designing these amplifiers?

I tried to find some discussion about this from the internet but strangely there doesn't seem to be any. I guess it's time for it now and more the merrier if it can relate to circuit i'm posting as attachment.


Thanks in advance,

Teemu K

[padamiecki]

Hello Teemu
General this amp shoul work fine
it is good that the poweramp does not have a global feedback - the distortion will be much better during overdrive it and easier for overdrive!!! Good to isolate it from the speaker, you know what I mean?
IMO using of opamps is not a good solution, try use jfets configured as tubes to get good distortion and the spectrum like in a valve amp.
One more: try to use at the input and at every stage high pass filter with the frequency~100Hz.
To get hmetal-like distortion try to overdrive signal shaped in pass filter ~1kHz.

[Ahmad_tbp]

hi teemuk
i m tryin to buil an amp too ! but for my bass guitar ...
better to take look in this site , it has several good schematics and power amp and etc .. http://www.albertkreuzer.com/start.htm
i wanna try the jfet bass guitar preamp that shown in this site .. u too better take a look at it , it may satisfy u ...
good luck

[darkfenriz]

Hi
your requirements about guitar amp assume getting distotrion from power amp. Another attitude is to add all the tube-like coloration and/or hard distortion from preamp and run power amp in 'linear' mode. This eliminates x-over dist.

Originally posted by padamiecki

.....
To get hmetal-like distortion try to overdrive signal shaped in pass filter ~1kHz.

My observations are the same, pass-boost 700-1600Hz before distortion gives nice results. £adna pogoda siê zrobi³a w tym pieprzonym kraju, nie?

[teemuk]

Hi all,

thanks for your answers but i still think there's room for more discussion.

Originally posted by padamiecki
To get hmetal-like distortion try to overdrive signal shaped in pass filter ~1kHz.

Well, i'm not really looking for distortion from this amplifier. If i'd want that i would rather use outboard effects. Distortion is nice and often associated with guitar amplifiers that are usually driven harder than "home" amplifiers but implementing a "distorter" to this amplifier would need a whole new circuit (channel) since distortion tone shaping is very much different from typical tone shaping, especially in solid state devices. And what if you want to have a clean tone?
Totally another question is how to make the distortion that WILL occur in any amplifier mellow and nice sounding. That seems to be quite tricky with solid state devices and since this is low power amp it might become an issue. A compressor stage before the final voltage amp possibly...?

Originally posted by padamiecki
try use jfets configured as tubes to get good distortion and the spectrum like in a valve amp.

What do you mean by configuring jfets as tubes? The jfets can be configured as jfets - nothing else. Solid state is solid state - tube is tube and the sonic spectrum is unique in every case.

Originally posted by padamiecki
One more: try to use at the input and at every stage high pass filter with the frequency~100Hz.

The lowest E in guitar is 80Hz. Dropped tunings put out even lower frequency and same thing happens when you play damped strings so why put low pass on 100Hz?
I have to disagree with you since i'd rather use frequncy shaping in the signal path as less as possible since this is the easiest way to get nasty sounding distortion besides overdrive. Truly some is needed to cut down lowest and highest end and to avoid oscillation. As you notice i do this with small decoupling caps and negative feedback in opamp stages. However, i don't agree with you on 100Hz.

Originally posted by padamiecki
IMO using of opamps is not a good solution

Personally i see nothing wrong in using opamps as long as they're not overdriven. That could happen in the last voltage amplifiying stage unless the input sensitivity is not set correctly. Anyone have idea of how to do it? (Frequency curve and pickup issue).
I tried jfet's in a 1/2W practice amplifier that i built few months ago. It had a chip amp otput stage (lm386) and a jfet preamp. I have to say that i liked the distortion characteristics of jfets when compared to opamps but at the end all comes down to clipping power amplifier which then should have fets too and that's a whole new story.
I didn't want to take that path this time since this is my first try in designing a discrete power stage and i wanted to keep things as simple as possible. Also, i find opamp design easier because you can avoid of using large amount of discrete components. As long as you don't overdrive them they don't seem to differ that much from jfets or any other discrete circuits.


Teemu K

[padamiecki]

Dear Teemu,
80Hz is low E maybe for the bass,
when you are using 6string better put 100Hz high pass filter, the sound would be quicker especialy during palm mute technique.
Besides this lowers a little hum and ripple from supply section.
Many people use opamps, but look at the forum - you have to know which opamp to use!
If you are going to build just an amp please ignore my posts!

A w Krakowie na wysokoœciach pod Kopcem Pi³s. pieprzona zima, dopiero ni¿ej czuæ wiosnê!!! Ale czu³emj¹ du¿o wczeœniej w koœciach i dobrym samopoczuciu!

[runebivrin]

Originally posted by padamiecki
Dear Teemu,
80Hz is low E maybe for the bass,
when you are using 6string better put 100Hz high pass filter, the sound would be quicker especialy during palm mute technique.
Besides this lowers a little hum and ripple from supply section.
!

Low E on the guitar is 82.5 Hz, trust me! Why would a 100 Hz filter affect the sound of the guitar otherwise?

But yes, turning down the bass can be beneficial to the sound, particularly if like the sound with less bass

Also, there seems to be a lot controversy about the amplitude range of guitar pickups. The common belief is that they produce a peak voltage of 100mV (peaks) and when the note decays the amplitude is typically about 30mV.
My research has shown that some pickups can easily do 400mV peaks and some effect pedals even put out a line level signal. 1V is quite much when compared to 30mV. When i tested my guitar's pickups by plugging them straight into mixer's input they were almost equally as loud as a line level signal. What is the truth here? To what peak voltage you people usually base your design into?

You really need an input sensitivity control that sets the gain of the first stage. Guitars are a varied lot, as you have discovered. Many guitar amps have Hi/Lo inputs for this reason.

Switchable resistors or a potentiometer is a good thing. I'd make U1 a non-inverting stage to make the values of resistors more manageable.

Rune

[teemuk]

Originally posted by runebivrin
Low E on the guitar is 82.5 Hz, trust me! Why would a 100 Hz filter affect the sound of the guitar otherwise?

But yes, turning down the bass can be beneficial to the sound, particularly if like the sound with less bass

Yep, i think the main question is how much the low frequncy has to be cut? Too much bass is poison for low power amplifier and damped notes easily fart out. This thing can also be compensated with a good speaker cabinet but too much is too much. Where would you put the -3dB point?

Originally posted by runebivrin
You really need an input sensitivity control that sets the gain of the first stage. Guitars are a varied lot, as you have discovered. Many guitar amps have Hi/Lo inputs for this reason.

Switchable resistors or a potentiometer is a good thing. I'd make U1 a non-inverting stage to make the values of resistors more manageable.

That´s what i figured out too. I´t´s quite easy to replace a resistor with a trimmer but still there´s a huge difference between 30mV @1kHz and 1V at, say, 82.5Hz. So what would be the starting point? 30mV on 1kHz and lowering the gain from there?

Also, has any comments of the amplifier circuit in general?


Teemu K

[runebivrin]

Originally posted by teemuk


Yep, i think the main question is how much the low frequncy has to be cut? Too much bass is poison for low power amplifier and damped notes easily fart out. This thing can also be compensated with a good speaker cabinet but too much is too much. Where would you put the -3dB point?

Teemu K

Can't say I have a general opionion. I like my bass deep. It's something I'd consider making adjustable.

Originally posted by teemuk

That´s what i figured out too. I´t´s quite easy to replace a resistor with a trimmer but still there´s a huge difference between 30mV @1kHz and 1V at, say, 82.5Hz. So what would be the starting point? 30mV on 1kHz and lowering the gain from there?

30mV is a probably not necessary. I'd start at 100mV, maybe. In my opinion the operative word is headroom, as in headroom before the master volume. That way, when it does clip you know it's the power amp that's doing it. If you're playing Sultans Of Swing style there's gonna be a lot more than a 1:3 ratio beween peak and decay levels.

Originally posted by teemuk


Also, has any comments of the amplifier circuit in general?

Well, there's a bunch of 6.8k resistors from the op-amp outputs. What are they doing there? I know som op-amps were designed to be resistor loaded, but what are you using?

As I said earlier, I'd consider a non-inverting input stage for gain adjustment.

What is R25 doing? 1miliohm is hard to buy

J1 seems a bit superflous as a whole. Surely U1 can drive the 33k impedance of U3?

U1 has no DC negative feedback.

Rune

[Giaime]

About those high-pass filter question, I have to say that in general guitar amps have a -3dB point at 60Hz, that's to cut lower frequencies to prevent them to distort, fart, or just steal power from the higher frequencies. And jes, the guitar low E is 82Hz...

However, i suggest to you to use jfets. Yes, using them "normally", like you use opamps "normally", it is the same sound. But there are jfets experts that can do about anything with those transistors, search for some on the usual fx forums (www.diystompboxes.com). Some of them are really impressive in creating a CLEAN tube sound (I understand that you want to add pedals for overdrive ). You know, a warmer clean amp is better for pedals, because they tend to give a punchier sound with more sustain. Just put a Rat in fron of an hi-fi amp and then in a clean Fender amp... you'll hear the difference, even when the THD is under 2%. Clean tube sound is better for pedals than clean hi-fi opamp sound, even with lo-pass filters and all the like. Simple lo-pass sound like lo-pass, not "warm sound".

Maybe trying some tubes in the preamp? That would be the best... Some clean 12AX7... Plate voltage took with a voltage doubler... search the net if the idea intrigues you.

Keep posting your results!