Build Your Own Clone Message Board

Please give me some ideas for a good book for a guy who:

Has built several pedals from kits, some with minor mods

Wants to understand what all the op amps, transistors, etc are actually doing

Wants to maybe design my own overdrive/dist or be able to predict what types of mods might be right for my style

Is very good with math formulas

Knew/was good with all the basics from HS physics of current, resistors, etc, but has gotten rusty

I have been to dozens of web pages and learned a lot, but I need one good physical resource instead of a spaghetti of web pages. I want to wrap my head around how biasing works, impedance, where the circuit has DC and/or AC, etc... some resources I have seen are way over my head and some are so basic that they are a waste of my time.... some have nothing (seemingly) to do with guitar pedals....

So, something basic, but not too basic... something that will help me grow. Any suggestions are appreciated! THANKS!

Unfortunately this seems to be a problem concerning electronics more generally, not just for guitar pedals. My associates who taught "college-level intro electronics for science majors" over the last few decades continually complained that there just wasn't a "good enough" textbook and they never could settle on one.

A book that sometimes infuriates me with its typos and dumb-@$$ "water flow analogies" but which you might find useful is Practical Electronics for Inventors by Scherz (1st. ed.) and Scherz & Monk (2nd, 3rd, 4th editions). It's pretty comprehensive (and includes a lot about digital electronics that may not be of interest) so might be much more than you are looking for, but it has the advantage of being available used for pretty cheap--even new it doesn't cost that much. It's possible that you could find it in a local library or large book store and flip through it to see if it's something you'd want to buy. It's not geared toward guitar effects pedals but does have some explanations for some of the building blocks (transistor circuits, op amp configurations) if you look in the right chapters.

(The non-answer: I have a great fondness for Anderton's Electronic Projects for Musicians, and also The Art of Electronics by Horowitz and Hill but I don't think either of them is really what you are looking for. The former doesn't really explain in depth about the electronics, and the latter is more like a reference than a textbook or study guide. )

I found taking a few tech classes at the local community college to be absolutely necessary for me to understand what pedal circuits were doing... it just got to the point where I knew I needed the math, or I'd never move forward.

Pop on to the website of your local community college and see if you can access the required texts for a class like "AC/DC Electronics 101", then hit the used market for an older version. The new text will probably be $100, but last year's revision will have all the same info for about $10. Read up on Ohm's Law, Kirchoff's voltage and current laws, and maybe refresh your memory with info on resistors, capacitors, inductors. If you're good with math, do some of the sample test questions to make sure you understand those circuit basics. Then, skip ahead to the chapters on semiconductors. You'll want to take some time looking at BJT transistor circuits to learn various biasing techniques, but you won't be able to grasp that too well until you have a firm handle on Ohm's and Kirchoff's laws. Then give yourself some homework. An analysis of the big muff Pi circuit would be a great way to start! Work your way thru a circuit schematic diagram (just pick your preferred version) and solve for DC voltages at all points along the signal path. You'll need to pull datasheets for the transistors to get some nominal values for their performance. That will teach you what the circuit is doing (mostly).

The next step would be to grab the text from the second semester of the class (AC/DC 102 or equivalent) and start learning how AC voltage works, and how it's different from DC voltage. Then, returning to the Big Muff Pi schematic, solve for AC voltage at all points in the signal path. This gets tricky, as you'll see, because of clipping either from the clipping diodes and/or the transistors (depending on bias). It may be helpful to visualize the waveform of the AC (guitar) signal, which will enter the pedal's front end at around 1-2Vp-p (peak-to-peak).

And check in here and ask for help whenever you get stuck, we'll try to talk you thru it! If you have the gear and opportunity, build yourself a copy of the circuit on a breadboard and use a voltmeter to test DC voltages to check your math and discover just how much variance there really is in component tolerances! And if you can get your hands on an oscilloscope, probing the AC voltage (signal) will show you in seconds what it can take 20 minutes of math to discover... plus the visualization might be REALLY helpful (it was for me).

_________________

Man, this is fantastic information. Thank you.


Sent from my iPhone using Tapatalk

_________________
Pedal building is like the opposite of sex. All the fun stuff happens before you get in the box.

Excellent ideas!

So far, I have gone through 2 HS physics books to find some review of voltage, current, resistance, capacitance...

Practical Electronics for Inventors can be found in 1st edition pretty cheap, it is in my shopping cart.

Good advice about seeking a college text, I will look into that, too.

I have my eye on Brian Wampler's book, but it is a little more expensive and although his videos are great, I have seen mixed reviews about the book...

Thanks for the suggestions... if anyone else has an eye-opening reference, whether printed or on the web, please share.

Unclesixer I saw your post in that other thread about the guv'nor/chancellor schematic. If you do end up getting one of the editions of Scherz(/Monk) I can point you to some specific locations in there (I have 'em all). In the mean time, I can tell you what to look for in whatever references you *do* have:

(1) For op amps that are set up with negative feedback and serve as inverting or non-inverting amplifiers, the gain depends on a ratio between the "resistance" between the output and the inverting input (numerator) and the "resistance" between the inverting input and the thing it attaches to (either the signal being amplified, or else "ground" or something similar) (denominator). If those are in fact just resistors, then you're almost done.

(2) If those paths also have capacitance in them, either in parallel or in series, then the gain becomes frequency-dependent and for "resistance" you probably can get away with "magnitude of impedance" if you aren't concerned quite yet with how phase is affected. So now you need to know about capacitive reactance, which can play a role analogous to resistance in some ways; and how to combine a resistance and a reactance together (in parallel or in series, as needed) to get the size of the impedance, and use that in the gain formula. The reactance is frequency-dependent, so the gain becomes frequency-dependent.

(3) You can skip ahead to a *qualitative* assessment if you are just trying to determine whether something is high-pass or low-pass by using the following crude rules (which follow from the formula for how a capacitive reactance depends on frequency and on the capacitance C): at higher and higher frequencies, the capacitor acts like lower and lower resistance; at lower frequencies, the capacitor acts like a larger resistance.

So for example, referring to your other message, think of the R2-C2 series connection of IC1a as getting a bigger number in the denominator at low frequencies, because C2 contributes even more to the "resistance-like" thing in a series combination with an actual resistor. On the other hand, the R5-C6 parallel combination at IC1b gets a smaller number in the numerator at high frequencies, because the C6 "resistance-like" thing gets small and is in parallel with the fixed R5--parallel, so the smaller one dominates what is going on.

So pretty soon you start to recognize that parallel cap in the feedback arm as something that suppresses high frequency.

(Then you need to go back to some formulas to figure out at what frequencies the suppression is starting. We can help you with that too.)

OK, you are speaking a language I can understand...

I suddenly realized today that I have been stuck thinking that R2, C2 were "bleeding treble away from the negative input", but it is more proper to think that R2, C2 allow the negative input to "see" certain frequencies "from ground" and block others... I know my jargon is not proper, but I think I am understanding that piece a little better. Then you can apply the ol' F=1/2*pi*RC, which I kinda understand from you tells you which frequencies the resistor is the only resistor and which frequencies the resistor AND capacitor together act like a resistor..... and that effective resistance plays out with the gain (which I kinda get)

Then these caps in the neg fdbk loop (C3, C6)... they BOTH cut treble even though one is attached to a non-inv amp and the other is inv? I think I got that now, since they are parallel, there is less resistance to high freq in the fdbk loop, in turn, less gain on the treble... is that still F=1/2*pi*RC for the cutoff?

Man, I have to go back and stare at that schematic a while more now

But also, you mention phase in relation to the capacitors... this is because they have a charge/discharge turnaround time, I guess... but I am "not concerned with that quite yet"... is that because the phase doesn't matter (or is negligible) for overdrive and distortion (but may be very important, I assume in modulation, time-based, and/or wet/dry effects???) or am I just not smart enough yet to worry about it

Master (1) the voltage divider and (2) the way a capacitor deals with an applied sinusoidal voltage and you'll have some valuable tools. Then (3) apply them to the conventional op amp topologies (neg-feedback leading to inverting or non-inverting amplifiers) and many chunks of pedal schematics will start to make sense.

(1)

Ohm's law + Kirchhoff's loop law (esp in this form: two components connected in parallel have the same voltage across them) + Kirchhoff's node law (sum of currents going into a connection = sum of currents going out) lets you derive the result for the two-rung voltage divider made up of a voltage draped across a series connection of two resistors. Let that be your starting place. The two individual resistor voltages are in proportion to the individual resistances, and must sum to the applied voltage.

Then, note that the formula for the resulting "division" (or perhaps better, "partitioning") applies even if the applied voltage changes with time. Note in this case that if the applied voltage is a simple, zero-centered sine function, then each resistor voltage has that same frequency and phase, but with smaller amplitude, as dictated by the voltage divider formula.

Well guess what? We can build voltage dividers that have OTHER things (rather than just resistors) as the two rungs. Kirchhoff's laws still apply. For example, if you replace one of the resistors with a capacitor, and still drape a sinusoidal voltage across the series pair, you get something that can act as a low-pass filter (if you pay attention to what the resultant capacitor voltage is) or a high-pass filter (if you pay attention to what the resultant resistor voltage is). The "crude rules" from my previous post about how capacitors work can then be used in the voltage-divider results to confirm this. But you want more ... for example, you want to know at what frequency is the knee of the filter, and how fast the response falls off with frequency, and what is the phase relationship between the applied ("draped") voltage and the output voltage (taken at either just the capacitor, or just the resistor). So now you need to know more about what the capacitor is doing.

(2)

I'll just give a bit of motivation rather than the full results, since I don't want this installment to get to be too long.

Consider these things/ideas:
(a) A sinusoidally-varying driving voltage V(t) of known frequency f and amplitude V_o. Let us assume that it is zero-centered (has zero average value; "has no DC component")
(b) A two-lead electrical component to which that voltage is applied.
(c) The resultant current I(t) in the electrical component. I call this the "response."

For "linear" components (resistors, capacitors, inductors; combinations of them; and some other things) the resultant response I(t) is also (co-)sinusoidal, and with the same frequency f. But it might not be in phase with V(t), and we also would like to know what is the amplitude I_o of that response.

For a resistor, we already know the answers. The response is in phase with the driving voltage, and its amplitude follows from Ohm's law: I_o = V_o/R.

What about for a capacitor? First of all, the response is 90˚ (= pi/2) out of phase with the driving voltage. Note: it's not IN phase, but neither is it exactly OUT of phase. It's half-way in between. Keep this in mind, especially when you come to adding together (K. loop law) the voltages of a resistor and a capacitor that must have the exact same current running through them if they are in series (K. node law). Secondly, the amplitude goes like this: I_o = V_o/X_C where the denominator has the same units (ohms) as does a resistance, and plays the same kind of role ("for this much applied voltage, how much current response do I get?"). X_C is the capacitive reactance, and is X_C = 1/(2pi*f*C). (See that? The higher the frequency, the lower the reactance ... the lower the reactance, the bigger the size of the response I(t) ... same sort of thing a resistance does or doesn't do.)

I hope I don't have too many typos so far as I'm doing this off the top of my head.

Now the next thing to consider is ... What happens if you have two (co-)sinusoids that have the same frequency but are out of phase by 90˚ and you add them together? The result depends on their two amplitudes, which may or may not be equal. But what you get is another (co-)sinusoid of that very same frequency again, with some intermediate phase dominated by the stronger of the two, and some amplitude that's bigger than either of the two you started with.

(Special case: If the two that you add have the same amplitude, the resultant phase is 45˚ from each of them, and the overall amplitude is larger than either, by the factor square-root-of-two.)

That SUM is the sinusoidal voltage that you applied to the voltage divider that has a resistor as one rung and a capacitor as the other. The two 90˚-out-of-phase (co-)sinusoids you added together are the individual voltages across the resistor and the capacitor. So the idea is to think about how the sinusoidal voltage you apply gets partitioned into two (co-)sinusoids, one for the resistor and one for the capacitor.

I better stop for now. But here are a teaser or three.
(1) The explanations for the gain formulas for those op-amp topologies can sort of be understood as applications of a voltage divider, with the inverting input as the junction between the two rungs. Note that the rungs can be even more complicated than what I described above.
(2) The parallel R and C in the feedback loop of an op amp (essentially one rung of a voltage divider) can be understood as a current divider ("partitioner"). The same kinds of mathematical ideas I pointed at above can be applied there. It's not that the R,C parallel pair act as a filter so much; they need the other rung to help them.
(3) The RC filter "knee" is the frequency for which R = X_C. At that frequency, you have the appearance of 45˚ and the square root of 2 in various results of interest.

WMP1... did somebody tell you that when I am not playing guitar or being a dad... I teach HS PreCalc.... with a LOT of Trig?

I have shown my kids that exact graph (y=sinx+cosx and/or y=sinx+sin(x+pi/2)) with the brief explanation that adding periodic functions can give surprising results!

I printed a copy of your post, leaving myself VERY WIDE margins to recopy formulas and make notes.... I think I am seeing a lot (for me) of the big picture, but I will need to reread your post about 10 more times to be sure.

As a teacher, I commend you for sharing this the way you did... I had not been considering the negative op amp input as being in the middle of a voltage divider! Because I just wasn't "seeing" it that way, but it suddenly makes more sense...and the two filters make more sense...including their relation to the gain of diff frequencies the op amp!

I look forward to my new book arriving... and I will prob hit you up again as I proceed!

Yes, well, in this case we're going to ramp it up a bit in the next go ... maybe some derivatives and complex numbers ...



I am not sure that I've ever seen it described that way anywhere, but at some point in the recent past it occurred to me that that is how it is. The validity of that interpretation depends on the "ideal" op amp feature that its inputs draw/source zero current. Usually we think of being presented with the extreme potentials at the ends of the ladder and want to find the intermediate potential. In this interpretation, we're given the potential in the center (the non-inverting op amp input, which the negative-feedback op amp tries to match to the inverting op amp input) and one end (the non-feedback connection to the inv input) and are trying to find the other end (the op amp output). Usually however we just use the gain formulas for the two topologies.

This has 75 pages on pedals and gets a little geeky (in a good way)

https://www.amazon.com/Electronics-Guitarists-Denton-J-Dailey/dp/1461440866

_________________
Twisting and tinning is for chumps.

The book arrived in the mail. But before it arrived, some other things have popped up "in real life"... but the table of contents is promising and sure to keep me up at night once in a while.

In the meantime, I am also trying to get my buddy's wedding present built and my deadline is 9/25.... the chancellor with a few mods that (hopefully) give it the right flavor of growl for my friend it will be named "Appetite for Distortion"

Cute!

_________________
“My favorite programming language is SOLDER” - Bob Pease (RIP)

My Website * My Musical Gear * My DIY Pedals: Pg.1 - Pg.2

Thanks, lefty. I am still working on several things and my first book has arrived (and I found it super cheap!), but that recommendation from you looks good, too!