As I said in the beginning, impedance is made up of three things that impede or restrict AC current flow: resistance, inductance and capacitance.

Inductance is generated by AC signals flowing through coils so guitar pickups add impedance to the guitar and amp input circuits.

Capacitance in a circuit acts as a drag on AC signals because the signal looses energy moving electrons on and off the opposing capacitor "plate." The "plate" can be the actual plate of a capacitor, the ground shield in a guitar cable or a nearby wire in an amp. There's even capacitance between the electrodes in a tube.

Tube amp gain stages typically put out high impedance signals (high voltage, low current, "thin") but a cathode follower can put out much more current to create a low impedance signal.

It took me a long time to understand why input and output impedance matters in amplifier circuits. Here's my layman's explanation:

For simplicity sake you can think of impedance as AC resistance. The maximum power-transfer theorem says to transfer the maximum amount of power from a source (guitar) to a load (amplifier), the load impedance should match the source impedance (a.k.a. Impedance Matching).

Impedance Matching is used for maximum power in the power amp circuit where the power tube output impedance is matched to the speaker impedance through the output transformer. But at the amp's guitar signal input we're not concerned with maximum power transfer, we want maximum voltage transfer because a guitar amplifier amplifies a signal voltage--the voltage changes are the audio signal. We really don't care about the current generated by a guitar's pickup coil, it's the pickup's voltage changes that will be amplified by the preamp tubes. Since we favor voltage over current we can use an impedance mismatch to trade current for voltage. This intentional impedance mismatch reduces the current from the guitar but boosts the signal voltage the amplifier receives.

Guitar Output Impedance & Amplifier Input Impedance

Guitar circuit on left, amplifier input on the right. The current generated by the guitar pickup coil must be allowed to loop back to the guitar pickup. Typical guitar impedance is around 10k to 20k ohms. The "Amplifier Input Load Impedance" is typically a 1 megaohm input resistor mounted on the amp's input jack.

In the diagram above the guitar's pickup coil generates a signal voltage. The guitar's output impedance is made up of the resistance of the pickup coil, volume and tone controls + the guitar circuit's capacitance and inductance. Since we can't control the output impedance of the guitar we can maximize the signal voltage by making the amplifier input impedance as large as possible, preferably 10 or more times greater than the guitar's output impedance (Rule of 10).

A high input impedance reduces the current flowing through the amplifier but increases the signal voltage level. It also reduces distortion because the guitar's coil can output less current. This is called high impedance bridging and is used extensively in electronic audio circuits. The Input Resistor R1 on the 5F1 amplifier's guitar Jack 1 adds input impedance to boost the signal voltage from the guitar.

The same principal applies between amplifier stages. The output stage driver tube V1B sends the guitar signal voltage to the power tube V2. Low output impedance from the driver tube and high input impedance from the power tube boosts the signal voltage. This is what resistor R9 (power tube grid leak) does, it sets a high input impedance to the driver tube V2.  Back to Signal Input.

As mentioned earlier to get the most power out of the amplifier the power tube and speaker should match their impedance. But what happens when they don't match?

Notice how when plate load resistance (bottom of graph) is increased beyond 4k ohms power output begins to drop off (left side of graph), total distortion increases, 2nd order harmonic distortion increases (2nd H) but nasty sounding 3rd order harmonic distortion (3rd H) increases much more. Also notice how when plate load is decreased power output again decreases, total distortion again increases but sweet sounding 2nd order harmonic distortion increases and and nasty sounding 3rd order decreases. This is why a low impedance mismatch can sound better than a perfect amp-to-speaker impedance match. Also note that when running a 6L6 amp with an 8 ohm output transformer hooked up to a 4 ohm speaker the load resistance is cut in half from 4k to 2k and output power drops from 7.3 watts to 5.6 for a 23% power loss.

High Mismatch

If you connect a 16 ohm speaker to your 8 ohm output transformer (16 ohm load and 100% increase) the impedance as seen by the power tube plate increases and plate current decreases which can lengthen the lifespan of your power tubes, especially in Class A amps. This decrease in plate current will decrease demands on the power transformer and it will run cooler. Power filtering effectiveness is increased as current demand decreases so hum and line noise should decrease, especially in Class A amps. Decreased hum can help prevent ghost notes which are caused by hum interacting with musical notes to create false harmonic tones.

Since the power tube and transformer are not coupled at maximum efficiency some power is turned into transformer heat and the amp's power output is reduced.

We get the opposite of the sonic improvements of a low impedance mismatch: Sweet sounding 2nd order harmonic distortion in the power tube and nasty sounding 3rd harmonic distortion increase but the 3rd harmonic dominates.

The main problems with a high impedance mismatch are increased screen current and flyback voltage generated in the output transformer. A higher plate load impedance can rotate the plate load line to below the knee and cause excessive screen current which can melt the screens and kill the tubes. Flyback voltage can damage the power tubes and the output transformer itself. Flyback voltage spikes can cause the power tube to arc between pins or burn through the thin lacquer wire insulation used in the transformer windings. This is normally not a concern when going "one step" away from a match such as running a 4 ohm output transformer with an 8 ohm speaker unless the output transformer is cheaply made or really old. But running a 4 ohm transformer with a 16 ohm speaker can generate very high flyback voltages when running the amp hard near max volume.

For you Hi-Fi guys, increasing a speaker's impedance will change a passive crossover's crossover frequency. Changing out a two-way speaker's 4 ohm bass driver with an 8 ohm will shift the low pass filter's crossover point higher a full octave which will feed the bass driver more mid freqs and create an overlap where both the tweeter and bass driver are active. The bass driver impedance mismatch won't directly affect the tweeter crossover frequency.

Low Mismatch

If your output transformer is designed to match your power tubes to an 8 ohm speaker and you connect a 4 ohm speaker then the impedance as seen by the power tube plate decreases by 50% too. Less impedance will cause plate current to increase. The power tubes are stressed by this increased plate current so the power tube lifespan can be shortened. This is especially true of Class A amps because they idle near max current flow. Since the plate current idles near maximum, the entire power supply also runs at maximum output so the rectifier tube and power transformer will run hotter. Power filtering effectiveness is reduced as current demand increases so hum and noise may increase, especially in Class A amps. Increased hum can cause ghost notes.

The increase in primary current will cause the output transformer to run hotter. This generation of excess heat reduces power output at the speaker.

Many vintage Fender tube combo amps have an aux speaker jack that's wired in parallel with the built-in speaker. Plugging in a speaker rated at the same ohms as the built-in speaker will cut the speaker load in half for a "one step" low mismatch. Fender designed their output transformers to handle this and is considered safe.

But there are two possible sonic improvements with a power tube and speaker low mismatch. Sweet sounding 2nd order harmonic distortion in the power tube increases and nasty sounding 3rd harmonic distortion decreases dramatically. This is why it's worth experimenting with different speaker loads, you may like the tone you get. This is also why I prefer a 6.6k:8 ohm output transformer for 6V6 power tubes like the Deluxe Reverb uses instead of the typical 6V6 8k:8 ohm transformer.

If your power and/or output transformers run hot with a matched speaker load then mismatching them is more of a risk. The bottom line is the greater the low impedance mismatch, the harder your amp works, and the greater the high impedance mismatch, the more likely you are to burn out your output transformer and/or power tubes. For tube amps a low mismatch is safer than a high mismatch, that's why most speaker jacks have a shorting switch--a short is better than an open circuit. The opposite is generally true for solid state amps--a short will burn out their output transistors. An output transformer open circuit must be avoided, that's why I recommend soldering speaker wire to the speaker and why I recommend you shut down your amp before changing the speaker impedance switch (and avoid cheap impedance switches in your builds).

For two-way Hi-Fi speakers, decreasing a speaker's impedance will change a passive crossover's crossover frequency. Changing out a two-way speaker's 8 ohm bass driver with a 4 ohm will shift the low pass filter's crossover point lower a full octave which will feed the bass driver less mid freqs and create a "hole" in the speaker system's frequency response. The bass driver impedance mismatch won't directly affect the tweeter crossover frequency.

Running 2 Power Tubes in a 4 Power Tube Amp

Many push-pull amps designed to run 4 power tubes can be run with 2 power tubes to cut the output power in half. But since the output transformer was designed to load the current put out by 4 power tubes we need to make an adjustment on the speaker end to load 2 tubes properly. Since 2 power tubes put out half the current of 4 tubes we need to double the speaker impedance so 2 tubes feel the same load as when 4 tubes are used. If your amp is designed to run an 8 ohm speaker with 4 tubes like the AB763 Blackface Single Showman, it will need a 16 ohm speaker when run with 2 power tubes.

Amplifier and speaker phase has to do with which way a speaker moves, in or out, with a positive voltage at the amp input (such as a 1.5v battery + terminal to input jack tip). For multiple speakers playing the same source material such as a stereo audio amp or a multi speaker guitar cab, phase does matter. You want all the speakers moving in concert--all of them moving out together, then in together. If not there will be interactions between the speakers that change the apparent volume of certain frequencies. A loss of low end is usually what is noticed but the volume loss and frequencies depend upon the position of the listener.

For single speaker amps however, phase does not matter. I have had guitar players argue with me that it does matter, especially when they are working with guitar/amp feedback. Why would it matter if the speaker moves outward when the guitar string moves upward versus downward? If the listener (or guitar player if he is working with amp feedback) moves a foot or two toward or away from the amp speaker the phase flips at his location so phase does not matter.

Many multi-channel guitar amplifiers have different phase for different channels. If a channel has 1 or 3 extra gain stages compared to another channel then the amp/speaker phase will be 100% different between those two channels at the speakers. Again, this does not matter due to the material in the previous paragraph. But, if you jumper two out of phase channels they will sound weak and funky due to the phase cancellation when the two channels meet inside the amp.

What about two guitar players playing the exact same song? Does it matter if their two amps and speakers are in phase? No, because their picking would have to be incredibly synchronized to make a difference. At 1000Hz there is three hundredths of a second difference between the signal and 100% out of phase. The placement of the players' amp speakers is also important. A 1000Hz signal has a wavelength of 13.5 inches so if one of the speakers is 6.75 inches closer to the listener the phase will change by 100%. Don't worry about your band mates being out of phase with you or your amp.

Changing an amp's phase can help when recording live performances. Standing acoustic waves can cause tonal anomalies that can sometimes be improved by changing the phase of the guitar or bass amp. The bass amp open E is 41Hz which has a wavelength of 25 feet so flipping the polarity of a bass amp can change a standing pressure wave at the microphone to a standing wave trough which may solve the guitar and bass interaction at the microphone. Live concert sound board operators sometimes run into the same situation. At their location there is a funky interaction and by flipping the phase of the guitar or bass signal using the sound board can improve the tone at his location. The problem is if he moved to a new location he could find another spot with similar funky interaction.

Let's say we take a new 325-0-325v rated power transformer out of the box and connect the primary to 125 volts AC from the wall. The power transformer center tap and high voltage secondary wires aren't connected to anything (they're floating) and we measure 650 volts AC from one high voltage secondary wire to the other. We would also measure around 325 volts AC from either hot wire to the center tap.

We can touch either hot wire or center tap and not get shocked because the voltage isn't referenced to ground--the voltage is only between the three secondary wires. If we measure the voltage between a metal ground rod sunk into the earth and any secondary wire we'll see only random low voltage caused by transformer leakage.

But if we touch any two secondary wires we will get shocked.

If we connect the power transformer center tap to an un-grounded metal chassis we would measure around 325 VAC between either high voltage wire and the chassis but we can still touch a single hot wire or the chassis (not both) and we won't get shocked.

If we ground the chassis  we now have a modern, standard, powered amplifier. We and the grounded chassis have a common path literally through the ground. If we touch an amp hot wire electrons can flow between the ground and the hot wire through our body and shock us.

If we rest one hand on the chassis and touch a hot wire with the other then 325 volts AC will push current from one hand to the other across our chest--that's why we have the one-hand rule for working on hot amplifiers.

If we stand on an insulated plastic box so our body has no path to earth we can touch either hot wire and not get shocked.

We can hold the grounded chassis in our hands and connect a hot wire (or failed short capacitor) to the chassis and not get shocked because the lowest impedance path to earth is through the chassis ground wire. Now if we were standing in a pool of water in our basement and did that we would probably feel a tingle.

The early Gibsonette (mid-late 1940's) has an unusual design. It uses a single pentode (6SJ7 tube) stage for the preamp, a self-split push-pull power amp (2x6V6) which requires no phase inverter. The self-split power amp works like this: The guitar signal on the upper power tube grid changes the current flowing through the tube which causes a voltage drop across the cathode resistor. This voltage drop is also present on the lower power tube cathode and since the lower grid is grounded (common grid) the voltage changes on the cathode affect the current flowing through the lower power tube. The speaker field coil is an electromagnet that takes the place of the speaker permanent magnet. The field coil also acts as the choke.

The GA-40 V3 uses one 5879 pentode per channel as a single preamp stage. The paraphase phase inverter lower triode gets its signal to its grid through the upper 470k grid leak resistor. The tremolo oscillation is fed to the channel 2 preamp plate. The "Tremolo High Pass Filter" is a fifth order high pass filter which removes tremolo thumping that can be created in the tremolo oscillation. Each RC network stage rolls off at a slope of 6dB per octave at 32Hz. Five successive filters create a very steep filter cut which allows the corner frequency to be set low enough to allow normal bass response in the tremolo channel (Channel 2) while still killing tremolo thump.

This amplifier won the Tremolo War with by far the most complex and intricate tremolo design. Each stereo channel has its own tremolo driver, harmonic pitch-shift circuit, power amp and speaker. Vibrato Channel signal path shown in blue. Normal Channel signal path shown in green. Tremolo Oscillator and Tremolo Drivers at lower left. Dual channel Harmonic Pitch-Shift circuits at upper center. Click image to see full size schematic.

Magnatone used an unusual first stage bias circuit in some of their amps including the 280. Using the upper 1A "High" input jack bias voltage is generated by the 4.7M Grid Leak/ Grid Bias resistor because it creates a voltage drop as grid current flows through it. The 470 ohm Cathode resistor generates a voltage drop when electrons flow from the cathode to plate. Since this circuit has no blocking capacitor the grid bias voltage will be passed to the guitar. When the lower 1B "Low" input jack is used the 4.7M grid leak resistor is bypassed so no grid leak bias is generated.

Rock and Roll Globe Article About Scott Totten's Magnatone 280 Rebuild

See the article here.

Humbucking pickups are really two single coil pickups connected together in a way that cancels noise and doubles the pickup output voltage. Humbuckers use common mode noise rejection to reduce hum and other electromagnetic noise picked up from surrounding sources like fluorescent lights. Any noise that gets into both pickup coils is nullified when the two signals from the two coils are combined because the noise signals from the two coils are of opposite phase.

In other words the positive noise signal is added to the negative noise signal and they wipe each other out. We get a "positive" coil and "negative" coil by simply winding the wire in the opposite direction.

Why doesn't the guitar string signal get nullified like the noise? Because of the opposite magnetic polarity of the pickups' pole pieces. One pickup has North up magnets and the other has South up magnets. This flips the signal polarity of one coil so the string signals are in phase, not out of phase like the noise. The string signals from the two coils are added together instead of being subtracted and nullified.

Just like two humbucker pickup coils, two separate pickups can be paired together to work as humbuckers and reduce noise if the two pickup coils are connected the right way.

You need two things to get hum cancelation in a humbucker pickup or between two single coil pickups: Opposite electrical current flow in the two coils--one coil's string generated current flows clockwise, the other coil flows counter-clockwise. The two pickups must also have opposite magnetic polarity (North+South).

The two coils of the bridge humbucker circulate electric current generated by the strings in opposite directions. The black arrows show the direction of current flow around the pickup coils. The same is true when you select switch position 2 which pairs the right bridge humbucker coil with the mid pickup and switch position 4 which pairs the mid and neck pickups. The magnetic poles of the left humbucker coil are South up and the right coil is North up. In switch position 2 the North bridge coil is paired with the South mid pickup. In switch position 4 the South mid pickup is paired with the North neck pickup. 5-Way switch positions are: 1 Bridge Humbucker, 2 North Humbucker coil + South Mid Pickup, 3 Mid Pickup, 4 South Mid Pickup + North Neck Pickup, 5 Neck Pickup.

It is pretty obvious when you pair two coils out of electrical phase (both coils' electrical flow running in the same direction)--the guitar tone is weak and tinny because the two string signals are subtracting and nullifying each other. The solution is easy, just swap one of the coil's two hookup wires to reverse its electrical flow.

It's much harder to detect by ear when a pair of coils are in phase electrically but out of phase magnetically--they will sound normal when paired but there will be no hum cancelation so you may notice some 60Hz electrical hum. The easiest way to check the magnetic polarity of two pickups is to carefully move the two pickups close together--top pole to top pole. If they magnetically repel each other they are the same polarity and will not hum cancel. If they attract each other they are opposite polarity and will hum cancel when paired. You can also purchase a $9 Schatten Design Magnetic Polarity Tester from Stewmac which makes checking installed pickups a breeze.

Many pickup manufacturers will change the magnetic polarity of their pickups for no charge. I had to do this on my Telecaster to get the mid and neck pickups to hum cancel. Note that if you change the magnetic polarity of a pickup the current flow direction will also reverse so you will have to swap its hookup wires to keep the same electrical polarity. Be aware that some aftermarket pickups have the opposite magnetic polarity as modern Fender pickups so when purchasing pickups that will be paired for humbucking make sure you know what magnetic polarity you're ordering.

Normal Humbucker Magnetic Polarity

Note the magnet on the bottom in contact with the Screw and Slug poles. Note how the bar magnet, slugs and screws all have a North and South pole.

NPN transistor on the left, PNP on the right. A positive Base voltage on an NPN transistor will allow current to flow through the resistor. A negative Base voltage on a PNP transistor will allow current to flow.

Here's two transistor circuits that look and function very much like tube amp circuits. The first is the transistor common emitter amplifier:

Tube triode amp circuit on the left, transistor common emitter amp on the right. Remove transistor resistor R2 and this circuit looks just like a tube amplifier stage. The two circuits serve the exact same purpose.

The big difference between the tube and transistor amp circuit is the addition of resistor R₂.  Resistors R₁ and R₂ form a voltage divider which sets the bias voltage on the transistor's base (similar to a tube's control grid). Resistors R_C and R_E determine the amplifier gain and set the max current. Capacitor C_E is a standard bypass cap to boost gain. Capacitors C₁ and C₂ are standard coupling caps.

Here's the transistor emitter follower. It bears an uncanny resemblance to a tube cathode follower:

Triode tube cathode follower on the left, transistor emitter follower on the right. Remove transistor resistor R₁ and this circuit looks just like a tube cathode follower. The two circuits serve the exact same purpose.

Again, the big difference between the tube and transistor amp circuit is the addition of resistor R₁.  Resistors R₁ and R₂ form a voltage divider which sets the transistor's bias voltage on the base (input). A transistor emitter follower can be used as an easy to implement tone stack buffer or effects loop buffer because they do not need a cathode heater.

How Voltage Dividers and Pots (potentiometers) Work

When I first got into electronics I had a hard time understanding voltage dividers but they are really pretty simple. Understanding them is important because a tube amp is full of them. Even tone controls and tone stacks can be thought of as voltage dividers. The Soldano SLO-100 high gain amp uses five voltage dividers to bleed off excess guitar signal to prevent harsh over-overdrive.

Classic Voltage Divider

Note how there is 100 volts at the top of the resistors and 0 volts at the bottom (ground). The resistance above the output is R1 and the lower is R2.

A voltage divider literally divides voltage. Take a look at the diagram above. We have two 500k resistors with 100 volts at the top and  0 volts (ground) at the bottom. If we tap the resistors at the top we get 100 volts, tap the resistors at the bottom and we get 0 volts. Now if we tap the resistors at the middle, half way down at 500k, we will get half the voltage, 50 volts. Tap higher and the voltage moves toward 100v, tap lower and the voltage goes down toward 0 volts.

Note the equation at the bottom of the diagram above:  The formula to calculate the voltage out of a voltage divider is:

Voltage Out = Voltage In * resistance below the output / total resistance

We can calculate the voltage out of the voltage divider by multiplying the Voltage in * R2 / (R1 + R2).

For our voltage above we get: 100v * 500,000 / (500,000 + 500,000) = 100v * .5 = 50 volts out.

The R2 / (R1 + R2) determines the ratio of the upper to lower resistance. If you make the lower resistor smaller, more voltage will be bled to ground so you will have less voltage out.