I have kind of a lot of questions related to the same general theme, so bare with me. The questions are the product of many hours spent pairing schematics with tone stack, coupling cap, bypass cap etc. calculators. I do have some educated guesses to some of these questions though.
-If I design a tone stack with corner frequencies of (lets say) 80Hz and 5KHz, why would I need to decide coupling cap values? 
-Why couldn't I just use a 20nF in every spot to easily allow all frequencies to pass?
-What's the difference between cutting bass with a coupling cap vs a bypass cap? 
-Why would Dumble use 4.7uF and Fischer use 22uF? 
-Why would a Marshall use a .68uF if the corner frequency reduces the gain of the low E by roughly 7db?
-If a speaker has a range between 80Hz and 5KHz, why would you need to design the amp to amplify/pass frequencies below or above those corner frequencies?
-I've seen (specifically LTP) phase inverter input coupling cap values range from 0.001 to 0.1.
-Wouldn't the smaller values cut a lot of low-end that was designed to pass in the preamp?
Thanks for any input!
			
			
									
									
						The Logic of Big-Picture Tone Shaping
Moderators: pompeiisneaks, Colossal
Re: The Logic of Big-Picture Tone Shaping
The simple answer that covers almost all of your questions is: DISTORTION!
With the exception of the first stage, which almost never distorts, each additional stage in our guitar amps can distort the signal, adding harmonics. If you do NOT trim bass frequencies out early in the circuit, these frequencies get "wooly, woofy, flubby, mushy" and swamp out the frequencies you WANT to hear.
As you point out, there are many opportunities along the signal chain to shape the tone. The answer to, "why here, not there?" is many years of obsessive experimentation by a lot of solder monkeys. You can leverage this experience and "stand on the shoulders of giants," or you can do your own experimentation. All good!
			
			
									
									With the exception of the first stage, which almost never distorts, each additional stage in our guitar amps can distort the signal, adding harmonics. If you do NOT trim bass frequencies out early in the circuit, these frequencies get "wooly, woofy, flubby, mushy" and swamp out the frequencies you WANT to hear.
As you point out, there are many opportunities along the signal chain to shape the tone. The answer to, "why here, not there?" is many years of obsessive experimentation by a lot of solder monkeys. You can leverage this experience and "stand on the shoulders of giants," or you can do your own experimentation. All good!
I build and repair tube amps. http://amps.monkeymatic.com
						Re: The Logic of Big-Picture Tone Shaping
xtian is right - tone shaping in successive stages is cumulative, and works sequentially on whatever the previous stage did.  As the math and physics guys might say, it's a path-sensitive function.
Even discounting distortion - which you can't really, but bear with me here - the effect of two 80hz high pass filters sequentially is not the same as one 80Hz filter. It's not even the same as one two-pole 80Hz filter all by itself. Designing filters is its own art. There are no real-world brick-wall filters; they all start sloping off before the -6db point that is the nominal cutoff frequency, and slant downwards at fixed slopes from that at rates that depend on the number of time constants ("poles") that the filter contains. The filter itself may be maximally flat in the pass band (... "Butterworth") or with some degree of ripple allowed to get things like fastest slope into the stop band, minimum ripple in the stop band, minimal phase shift, other things.
Generally the traditional guitar amp used only one time-constant/pole filters at a time with a gain stage between possibly overlapping filters. This cuts down on some of the passband/stopband issues above, but especially when your power output stage uses feedback, the multi-pole issues can pop up. The experienced solder jockeys have learned by either math or tinkering what works in a world of cascaded filters.
Hmm. Maybe the better answer is that like many things, what appears to be simple at first glance turns out to be complex.
			
			
									
									Even discounting distortion - which you can't really, but bear with me here - the effect of two 80hz high pass filters sequentially is not the same as one 80Hz filter. It's not even the same as one two-pole 80Hz filter all by itself. Designing filters is its own art. There are no real-world brick-wall filters; they all start sloping off before the -6db point that is the nominal cutoff frequency, and slant downwards at fixed slopes from that at rates that depend on the number of time constants ("poles") that the filter contains. The filter itself may be maximally flat in the pass band (... "Butterworth") or with some degree of ripple allowed to get things like fastest slope into the stop band, minimum ripple in the stop band, minimal phase shift, other things.
Generally the traditional guitar amp used only one time-constant/pole filters at a time with a gain stage between possibly overlapping filters. This cuts down on some of the passband/stopband issues above, but especially when your power output stage uses feedback, the multi-pole issues can pop up. The experienced solder jockeys have learned by either math or tinkering what works in a world of cascaded filters.
Hmm. Maybe the better answer is that like many things, what appears to be simple at first glance turns out to be complex.
"It's not what we don't know that gets us in trouble. It's what we know for sure that just ain't so"
Mark Twain
						Mark Twain
- 
				thisismyname
 - Posts: 41
 - Joined: Fri Feb 19, 2021 7:05 pm
 
Re: The Logic of Big-Picture Tone Shaping
Thanks, guys! This is all really interesting! 
I just bought Richard Kuehnel's Guitar Amplifier Electronics Circuit Simulation book on using LTspice. I'm really curious about the finer details of amp design.
I'm looking forward to learning more
			
			
									
									
						I just bought Richard Kuehnel's Guitar Amplifier Electronics Circuit Simulation book on using LTspice. I'm really curious about the finer details of amp design.
I'm looking forward to learning more
Re: The Logic of Big-Picture Tone Shaping
Good on you. Keep digging for understanding! You might enjoy Lancasters's "Active Filter Cookbook" if you can find a copy. It's a great single-capsule introduction to filters. It has a fair amount of the math, but the math is not central to what it says because there are good text and graphic explanations of what the filters do. 
I was extruded through undergraduate courses on circuits and filters, and I still reach for this book instead of my old textbooks when I have a quick filter question.
			
			
									
									I was extruded through undergraduate courses on circuits and filters, and I still reach for this book instead of my old textbooks when I have a quick filter question.
"It's not what we don't know that gets us in trouble. It's what we know for sure that just ain't so"
Mark Twain
						Mark Twain