This started out as someone on TDPRI asking for some enlightenment on what the various components do in a 5E3… before I realized it I had described nearly every component! Here it has been consolidated and reordered, along with a copy of the Weber 5E3 Deluxe layout that was used as reference.
I find it easier to use the schematic to get a sense of what the components are doing. The layout describes the physical arrangement of the pieces and wires, but the schematic is what really tells you what they’re doing. But let’s give this a shot!
The Inputs
First comes the input jacks and the 1M resistor referencing to ground. The purpose of the 1M resistor is to set a ceiling on the input impedance that the amp will see. Realistically the highest impedance you’re going to see from a guitar is 15K ohms, so the 1M value has practically zero effect on the sound. Next are the 68K input grid resistors. The purpose of these is to prevent radio frequencies from using the input as an antenna and turning your amp into a radio. 
It works off of the principle of Miller Capacitance, but to sum it up, the 1M resistor to ground and the 68K input grid resistor combined with the tube in V1 create a small amount of capacitance that works as a low-pass filter, effectively blocking out extremely high (beyond the human ear’s range) frequencies.
The First Stage
So back to V1′s jumpered cathodes and the right-most side of the fiberboard… Yes, the 820R resistor is providing cathode bias (DC voltage) for the two triodes. It’s roughly equivalent to cathode biasing each triode separately using 1.5K resistors, but lazier and cheaper. The 25uf electrolytic cap is stabilizing the AC voltage of the cathode. This prevents the AC voltage of the cathode from wiggling along with the input grid, and as a result the triode amplifies the signal more efficiently.
The value of the electrolytic cap controls how low of a frequency will be stabilized (and more efficiently amplified). 25uf works well down below the human ear’s listening range. If, for example, you were to put a 1uf cap there instead you’d definitely hear less bass.
Next up – the 100K plate resistors. They connect to the high voltage supply one one end and the triodes’ plates on the other. Thanks to the principle of ohm’s law, as current is drawn by the triode through the 100K resistor, the voltage on the triode side of the resistor decreases. Exactly how much it decreases is based on the biasing of the valve, and I’ll defer to Merlin’s explanations on that. One way to think of it is that the voltage drop across that plate resistor is what gives a valve the potential to amplify signals. As the input signal comes in, those tiny voltage changes affect the biasing of the valve, which changes how much current it draws. That change in current results in a large voltage change at the plate, and voila! we have amplification.
AC Coupling 101
So now we have an amplified signal coming out of V1. The trouble is that the DC voltage at the plate is somewhere around 160v, and we have to prevent that voltage from going on to the next stage of the amp. The .1uf capacitor is used to block the DC voltage (and LOW frequency AC) but allow the (higher frequency) AC voltage to go through. Again, figuring out what frequency ranges it allows and blocks is based around the Miller Capacitance principle. Basically the goal here is to allow all audible frequencies through and block the uselessly low ones out. Here is a very convenient tool for figuring out how to use cap/resistor combos to set rolloff frequency.
The Tone Stack
Next in the path is the tone stack. Tone stacks are complicated beasts… thank goodness this is a 1-knob affair! For the sake of simplicity, imagine that the first .1uf cap and volume pot don’t exist – let’s just focus on one preamp channel. Now that the DC voltage is blocked, the AC signal comes out of the .1uf cap and into the wiper of the volume pot. One side of the volume pot is connected to ground and the other is connected to the output that goes to the next stage. This is a very crude(*) voltage divider.
(*)A side note on basic voltage dividers… When a pot is used as a voltage divider you end up with one side having a fixed load and the other side having a variable load. Preamp valve outputs are high-impedance and don’t like pushing a variable load. Valve inputs are more tolerant of lower loads. For that reason the “normal” volume arrangement is to run the input into the “top” lug and the output from the wiper. The 5E3 does this backwards, and as a result there is a significant muddying of tone as the volume decreases.
Now that we understand what the volume knob is doing let’s look at the tone knob. The full-strength input signal runs through a 500pf cap to the “top” of the tone pot. On the “bottom” side of the tone knob is a .0047uf cap running to ground. The output goes out the wiper and connects to the volume’s output. At full-on the tone knob is adding high frequencies to the volume’s output and at full-off the tone knob is removing mids and highs from the volume’s output. This, combined with the backwards volume arrangement, is why the 5E3′s controls are so interactive.
Intermediate Gain stage
The signal leaves the tone stack and enters the first triode of V2. It has lost a bit of its strength from all that tone-stacking. This is just a basic amplification stage that boosts the signal up nice and strong. On the cathode we have a 1.5K cathode resistor (cathode bias) and 25uf cap (AC stabilization, maximizing gain). On the plate there’s the 100K resistor providing the amplification potential, just like V1. Then it goes out the .022uf capacitor to block the DC voltage so the next stage can do its thang.
The Phase Inverter
Why do we need a phase inverter? In your basic “single-ended” amp you’d just push the signal to the power valve and that’d be the end of it. But somewhere along the way some genius figured out that if you take two power valves and you arrange them face-to-face like a pair of boxers and feed them perfectly opposite signals, they put out a LOT more power. One valve “pushes” while the other valve “pulls”, hence the name push-pull. The increased effeciency is why one 6V6 in a Champ puts out 5W and two 6V6s in a Deluxe put out 18-22W.
SOOO… phase inverter. In this case we have the cathodyne phase inverter, which is one of the simplest. Our goal is to end up with two equal and opposite signals. Let’s pause and reflect on what we know about valve amplification… When a signal is pushed into the grid, the amplified signal coming out at the plate is inverted. Also, if we push a signal into the grid and we don’t have a cathode bypass cap, the cathode tends to wiggle in phase with the grid. So there’s the possibility, somewhere in there, of having one signal in phase and one signal out of phase. The challenge is to arrange the valve so both signals are about the same strength.
If you think of the cathodyne PI as being a standard gain stage, but the plate resistor is chopped in two and the second piece is stuck underneath the cathode (and the bias resistor), well… then you pretty much have it right. There still needs to be a stable relationship between the input grid and what was the “ground” of the gain stage, so a 1M “reference” resistor is added at the input.
With 56K resistors on either end of the stage, the whole valve is floating in the middle of our voltage range. When signal goes into the grid, the voltage at the cathode wiggles in phase with the input at nearly the exact same strength as the input. The voltage at the plate wiggles in opposite phase, but because the valve is “floating” in our voltage soup(*) it doesn’t amplify the signal. There we go, two (basically) equal and opposite signals that we can send to the power tubes!
(*)”voltage soup” is a highly technical term, though some heathens and crazies prefer “local feedback”.
Approaching the Power Stage
The coupling caps after the PI are there for the same reason as in other stages – they prevent DC voltage from going on to the next stage. .1uf is a very large value and allows all audible frequencies to make their way to the power tubes. Most later Fender amps use smaller caps like .022uf. This reduces some of the lower frequencies, which is most noticeable in the tighter overdriven sound compared to tweed amps.
You’ll see a trend present in practically every valve… There is always a reference resistor between the input grid and ground, or in the case of the cathodyne PI, the relative equivalent of ground. This is due to how cathode biasing works – the cathode’s voltage is raised by a specific amount above the grid’s voltage. The reference resistor does what its name implies and provides a reference voltage to tell the grid where it should be in relation to the cathode. That’s what those 220k resistors are doing between the power tubes input grids and ground. Every valve has a maximum grid reference resistance it can tolerate… Values of 330k or less are common for power tubes, while preamp stages often have up to 1M reference resistance. The smaller the resistance, the higher the load on the previous stage, and the “rounder” the sound. Using high resistance at every stage provides an edgier, harder overdrive.
At last, the power tubes! So far we’ve taken this tiny AC signal from somewhere between 50mv and 400mv and amplified the voltage up around 30v or so. We’ve got plenty of voltage but now we need something that will turn that voltage into amperage – it’s current that makes speakers move!
Grids and Grid Stoppers
6V6s are beam-tetrodes. Compared to triodes there’s an extra grid in that bottle. Pentodes have yet another grid, but we can pretty much ignore it for the sake of this explanation. The “screen grid” as it is often called makes the valve more efficient than a triode, especially in power amp duties.
You’ll notice that the input signal goes to pin 6 (which is unused on the 6V6) and then hops a 1.5K “grid stopper” resistor before it goes into the power valve. The purpose of the grid stopper here is a little different than the 68Ks found at the amp’s input. It does help stop supersonic frequencies, but here the goal is to prevent the power tubes from creating their own oscillating feedback. It also serves as a means of protecting the input grids from sucking current from the phase inverter during overdrive. This helps prevent all sorts of bad stuff, ranging from ugly blocking distortion to arc-shorts and valve destruction.
Grid stoppers are also added to the screen grids on many amps. The 5E3 “cheats” and relies on the 4.7K/2W resistor in the power supply network to do the job. The main purpose of these is to prevent the screen grids from conducting too much current and shorting out. Many (dare I say “most”) guitar amps run the voltage on the screen grids much higher than the valves were designed to handle. Adding screen grid resistors (470R-1K is typical) doesn’t affect the sound and is very cheap protection to prevent burning up power valves.
Biasing and Final Output
The 5E3 uses cathode bias on the power tubes, just like the preamp. The 250R/5W cement resistor on the far left side of the board is providing cathode bias for the 6V6 power tubes. Similarly to V1, the 25uf electrolytic cap is stabilizing the AC voltage. You can think of it in pretty much the same way as the preamp – the valve is a gain stage, the cathode resistor biases the valve so it works, and the cathode bypass cap helps maximize the efficiency. What’s different is that instead of a high-value resistor helping us create voltage, we use a lower-impedance inductor (our transformer) to safely transfer the power tube current to the speakers. The transformer is actually doing two jobs. Its first job is to act as an inductor, taking the place of the plate resistor, but capable of handling a lot more power. Its second job is to convert the high-impedance (6600 ohm) primary impedance on the power tubes to a low-impedance (8 ohm) output that’s suitable for speakers. There are a number of other ways to perform these two actions, but the output transformer is the most practical.
The Power Supply
Here’s the last thing on our fiberboard. Between the phase inverter and the power tube cathodes are the power supply capacitors. They get their power from the rectifier, which in the case of the 5E3 is a GZ34/5AR4 rectifier tube.
When it comes to running vacuum tubes we have a slight challenge on our hands: they need very high voltages to run efficiently, and pretty much no one in their right mind works with batteries that put out 300-500V DC. That’s because DC kills, Thomas Edison was a stubborn sonofabitch, and his protege Nikola Tesla had the right idea when he proposed using high AC voltages to send electricity over long distances.
We’re getting something between 110v and 240v AC from the outlet, depending on where we are in the world. The power transformer kicks this up to something like 600-700v AC inside our amps. We then use a rectifier to take the bottom-side of our AC waveform and flip it over so it’s facing upright. Great, now we have a power feed coming in that looks like a series of mountains. What we need is a way to chop the peaks off the mountains and fill in the valleys to get a stable DC voltage.
The big bad filter capacitors are our mountain-choppers. The first capacitor (16uf, nearest the power tubes bias resistor and cap) after the rectifier does the heavy lifting and gets the voltage fluctuation down to somewhere around 20-40V. The average (DC) voltage is dropped to around 70% of the original AC swing. This first filtering stage is where the power tubes get their power. Because the power tubes are in opposite phase of each other, the 50-60hz AC hum from those 20-40v is cancelled out.
There’s then a (4.7K/2W) resistor or a choke (big spender!) followed by another capacitor (16uf). Because there’s current being drawn through this resistor there’s a voltage drop, which gives the capacitor more “room to work”, and the AC fluctuation at this point is usually down around 2-5V. This is where the power tube screens get their voltage. Again, since the power tubes are in opposite phase of each other, the AC noise is cancelled out.
For the third stage there’s a bigger resistor followed by another capacitor. The resistor is bigger because less current is getting drawn through this section, so a larger resistor is needed to drop the voltage to where we want it. By the end of this filtration stage the AC fluctuation on the voltage is really low, down in the millivolts. This is good, because this is the part of the power supply that feeds the preamp. Any AC hum that’s left in the power feed gets amplified by the preamp, so the less AC, the better.
