What's all this Impedance Stuff About?

I'll keep the math to a bare minimum, but you can't describe impedance without a little math. When I was in college, the professor in an introductory course on circuit theory shared this anecdote: "As part of their oral exam, PhD electrical engineering students were asked to define the term 'impedance'. The review board was dismayed to find that many of them were unable to provide an adequate answer." He went on to explain that impedance is a very simple concept. By definition, impedance is the slope of the V-I (voltage/current) curve. Now that you know that, you're smarter than some PhD students.

For the purposes of pedal building, what is impedance and why do we care about it?
Impedance is the AC equivalent of resistance. It governs how electrical energy, in the case of effects pedals the guitar signal and DC power, move thru the circuit. Every electrical component, including wire, has impedance. The input impedance of the first pedal in the chain can influence the tone because if the impedance is low, it will load the pickups. Pedals like the Tube Screamer have a high input impedance and allow all of the signal to get from the pickup to the pedal's input circuit. The classic Fuzz Face, on the other hand, has a relatively low input impedance that loads the pickups, absorbs energy and dampens the pickup's resonant peak. For the first pedal in the chain, input impedance strongly influences a pedal's response to the guitar's Volume & Tone controls because the pedal loads those things too. Lower impedance pedals are generally more responsive to guitar control settings. The output impedance of a pedal influences how the guitar signal passes to the next pedal in the chain or the amp. Lower output impedance pedals can drive long cables without losing brightness. They can also drive the shit outta the next pedal in the chain. The reason some pedals don't get along is mostly about input and output impedance.

Impedance is expressed in Ohms and is given the symbol Z. Impedance showed up late and was last in line. By the time it got its symbol assigned all of the other letters had been taken.

Resistors
Resistors have an impedance equal to their resistance. Zr = R. Simple enough.

Reactive Components
Capacitors and inductors are called reactive because they store energy and their impedance is dependent on frequency. Their impedance equations presented below are slightly simplified, but are adequate for this discussion.

Inductors
Inductor impedance: ZL = 2 x π x f x L

where Z is in Ohms, f is in Hz and L is in Henries. The 500mH choke we use in wah-wahs and Conqueror pedals has an impedance of 691 Ohms at 220Hz, the frequency of the A string at the 12th fret. For inductors, impedance goes up as frequency goes up.

Capacitors
Capacitor impedance: Zc = 1 / (2 x π x f x C)

where Z is in Ohms, f is in Hz and C is in Farads. A 47nF capacitor has an impedance of 41K at 82.4Hz, the frequency of the open bottom E string. With capacitors, impedance goes down as frequency goes up.

Semiconductors
Semiconductors (diodes & transistors) are non-linear. Their impedance varies with current (or voltage). If we plot voltage vs. current, at any point on the curve, the impedance is the slope of the curve. The convention is to plot current on the vertical axis and voltage on the horizontal axis. In that case, the slope of the curve is the admittance, which is the inverse of impedance. Take a look at the v-i curve for a silicon rectifier diode.
generic diode curve.jpg
Below 0.56V the curve is essentially flat, which means that the impedance is very high, could be megohms. At 0.82V, the slope is about 10mA per 200mV. 200mV / 10mA = 20 Ohms. At 0.9V, the slope is around 60mA per 100mV. 100mV / 60mA = 1.67 Ohms. People like to simplify diodes and think of them as either OFF or ON. But there is a smooth transition from OFF (high impedance) to ON (low impedance) and that transition is what gives diode clippers their signature sound.

Filters
Getting back to inductors and capacitors, it is their impedance, and how it varies with frequency, that makes it possible to build electronic filters. The calculations used to design and analyze filters take into account the impedance of the capacitor (or inductor) and how it interacts with the other parts (resistors, diodes, transistors) to determine the frequency response.

Example: A simple low-pass filter can be realized with a resistor and a capacitor. We see that all the time in pedal circuits. The corner frequency is given by:

Fc = 1 / (2 x π x R x C)

At that frequency, the impedances of the resistor and capacitor are equal. Half of the signal power is lost in the resistor. Below Fc, very little signal is lost and the frequency response is flat. Above Fc, more and more signal is lost as frequency goes up. For every doubling of frequency, the voltage is cut in half.

There is another layer of complexity that I completely glossed over, but that's the quick tour. Any questions?
 
Man, Chuck, I've said it before and I'll say it many times again me thinks, but this is fabulous stuff!
Thanks so much to putting it in digestible words and examples and in the context of our practical pedal building and gaping at schematics...
Your igniting light bulbs here, the questions will follow when I try to wrap my head around examples and circuits.
When I have been scooting and trying to research stuff like this on the net I usually get bogged down in too basic or (mostly) way above my head and not well related to pedal building stuff and then I realize that I run out of time in my day trying to still do other stuff and haven't understood a thing, play guitar, talk to me wifey and stuff.

Thanks for putting it together.
There'... I said it again...
 
Thanks again for providing this information for us.

The first question that I have is the relationship between impedance and 'how easily a signal can flow'. Specifically, why is it that a high impedance at the input of a circuit is better at accepting a guitar's signal? Second, regarding the relationship between series circuits, can you explain why low output impedance is optimal? The characterization of impedance as AC resistance makes those concepts a bit difficult.

Further, would you mind exploring impedance and transistors a bit more. Definitely provide more math if necessary.
 
Damn Chuck, I feel like we owe you a nice Christmas ?. Thank you again!

I do this becasuse I like to and I don't solicit donations. But anytime someone wants to toss a circuitboard or PedalPCB gift card my way, I won't refuse it. ;)
 
  • Like
Reactions: fig
Thanks again for providing this information for us.

The first question that I have is the relationship between impedance and 'how easily a signal can flow'. Specifically, why is it that a high impedance at the input of a circuit is better at accepting a guitar's signal? Second, regarding the relationship between series circuits, can you explain why low output impedance is optimal?
I can answer both of those questions in terms of voltage dividers. Here's a simple resistor voltage divider:
1607555567435.png
The relationship between Vin and Vout is:
1607561870099.png

Now we'll replace the resistors with impedances.

1607555816086.png

And the relationship is:
1607561754522.png

Those impedances can be any kind of part or combination of parts, doesn't matter.

Make sense so far?

Now let's say that Z1 is the bridge pickup on your Strat and Z2 is the input impedance of a Big Muff. Z1 varies from a few KΩ at low freq up to around 100KΩ at the resonant peak of the pickup, somewhere around 6KHz. These numbers are all ballpark and vary from pickup to pickup.

Many Big Muffs have an input impedance around 40KΩ. At low freq, that voltage divider won't eat very much signal at all. At the resonant peak, it will shunt most of the signal to ground, making the peak less pronounced and the pickup less bright. Which is probably fine for a distortion pedal.

Fuzz Face pedals have an input impedance that is much lower and non-linear. The way they interact with pickups is more complicated, so I won't get into it here.

We can apply the same voltage divider concept to the output side of a pedal. Let's look at an extreme example. The Hoof has a 1M Volume pot at the output. The output impedance is determined mainly by that pot and can be as high as 250KΩ when the pot is in the middle of its resistance range. Now let's suppose we have 25 ft of cable from the Hoof to the amp. Shielded cable runs somewhere around 30pF per foot, but it can be much higher or lower depending on the dimensions of the cable. 25ft x 30pF/ft = 750pF. Therefore, in our voltage divider example Z1 is the pedal's output impedance, 250KΩ and Z2 is the cable impedance, 750pF. That pedal together with the cable have made a low-pass filter with a corner freq of... (let's not see all the same hands, class)... 849Hz.

The characterization of impedance as AC resistance makes those concepts a bit difficult.
It should make it simpler. Impedance is not exclusively AC resistance, it's resistance at all frequencies, including DC. The thing to remember is that resistance is constant with frequency, but impedance can, and probably does, vary with frequency.
 
Last edited:
Thanks for the response! All of that makes sense. I think why I was getting tripped up is that I wasn’t envisioning the relationship of output and input impedances as a voltage divider within the circuit.

On a slightly tangential note, why are inductors / LC filters much less common than RC filters in pedal circuits?
 
Good question! Inductors flourish in high-end audio, but they rarely show up in pedals. The reason is that inductors are larger and more expensive than capacitors. It is cheaper to synthesize an inductor using an opamp or transistor and a handful of Rs & Cs than it is to fabricate an inductor. Several pedals sold here contain synthetic inductors, including the Promethium and Nobleman.

I tried breadboarding a Crybaby with a synthetic inductor and it did not work out. High-Q synthetic inductors are noisy and there's not much that can be done about it. Crybabys have real inductors in them because with that circuit, there is no other way. There are inductorless wah-wahs, like the Tychobrahe Parapedal, but they don't sound or feel anything like a Crybaby.
 
Hey Chuck. Sorry to dig up an old thread, but what's the difference between the output impedance and the load in a given circuit? In other words, why do we need a resistor to ground at the end of a circuit when there's plenty that comes before it?
 
Depending on the circuit, they might be the same thing. When you say "load," I think you're referring to the the resistor between collector (or drain) and Vcc. There may be some stuff after that, like a Volume control, which changes the output impedance. In fact, if the last component in a circuit is the Volume control, then the output impedance is determined by the Volume control and it varies as you rotate the Volune control.

The strict definition of output impedance (Zout) is this:

Zout = Change in output voltage divided by change in output current. You can measure it in a real circuit with a scope or you can simulate it and evaluate it in LTSpice.

A long answer to a short question, but I just finished a large Margarita, so cut me a little slack.
 
Speaking of impedances, when used as a voltage divider in a digital circuit, such as the digital control inputs on an FV-1, what difference does the pot value make? For example, the FV-1 example datasheet shows 50k pots for each of the pot inputs, but the PedalPCB effects (and I think the Madbean one, as well as most that you see elsewhere) use 100k.

Since it’s just dividing a control voltage that gets interpreted in 9-bits, it seems that impedance shouldn’t be a factor or concern. Is it just a matter of current consumption, or is there more to it that that? For example what would the difference be if I used a 10k pot instead, or a 500k pot for that matter? It’s completely isolated from the audio path after all. Just total current draw?
 
Good question. If you look at the FV-1 datasheet, Spin Semiconductor uses B50K pots in their example circuits. As long as the input current to the control pin (10, 21 & 22) does not generate too much error, the pot value doesn't matter. The datasheet does not specify input current, but this is a CMOS device, so input current is measured in picoamps. The datasheet says that the input impedance of the control pins is between 10M and 20M, so pretty much any pot value would work. If you don't have the B100K called out in the Build Docs, feel free to sub any B-taper pot you have.
 
This is great learning material as always, Chuck. In this article, Mr Black says ideal values for pedals are 500kΩ-1MΩ for input impedance, with 1KΩ-10KΩ preferred for output impedance: Buffers, impedance and internet lore

Simulating a circuit I'm working on in LTSpice, I get an input impedance of 520KΩ and an output impedance of 26KΩ. It's 2.6x the ideal mentioned, so is that an issue? If so, what do I do to lower the output impedance?

The output section looks like this:
1665482250712.png
It's a JFET amplifier stage, similar to the one in at the end of a BSIAB2 or the Wampler Pinnacle. In the circuit (a boost/drive I'm fiddling with), it's following a pair of mu-amps and a James tonestack, to recover insertion loss from the James. The values here are from one of the circuits I was using as a template. Is there something that could/should be changed to improve output impedance?

I appreciate the things I am learning on this forum so much 🙏

Edit: R19 is actually a 50K trimpot for biasing the JFET, I set it at an average value as I'm not sure how to calculate and set it properly in LTSpice yet.
 
If you want a lower output impedance, you can add on a unity gain buffer. An emitter follower after the VOL control will do the job nicely. Or you could use another JFET and build a source follower. Look at any of the Boss pedals for examples of input and output buffers.

The ideal values cited in the article above can be used as a guideline, but there are exceptions.
There are plenty of good sounding pedals have a low input impedance; FuzzFace being a prime example. Personally, I prefer high input impedance pedals, but it is a preference, not a rule. Guitar pedals and amplifiers are supposed to alter the signal one way or another, so some of the rules we would normally follow in electrical engineering do not always apply. Same goes for output impedance. A majority of the pedal circuits I've seen have output impedances that are higher than 10K. The Boss & Ibanez pedals use soft switching and buffered bypass. Their input impedance is high (<100K) and their output impedance is low (~1K). They do this to accommodate the FET switching.
 
Back
Top