PDA

View Full Version : Guitar Effects Explained



jugglindan
22-07-2020, 02:38 PM
I picked the Jack Orman ebook Guitar Effects Explained from Amazon today. So far I am finding it very readable. It avoids too much math, and reads like a plain English description of how effects work at the circuit level.

I highly recommend this to anyone interested in building pedals, and also to anyone who just wants to know more about the magic.

There is some assumed knowledge required to understand parts of it, but nothing so far that a quick Google search can't fix. For example, the early section on boosters says "our signal is AC coupled into the booster so the clipping takes place symmetrically". Jack doesn't explain what coupling is though, or how it leads to symmetric clipping in the gain stage. Hopefully he explains the related role of the coupling capacitor after the gain stage.

Sent from my moto g(7) using Tapatalk

Bakersdozen
22-07-2020, 02:45 PM
Sounds great, exactly some of the things I am currently learning with building a few fuzz circuits and playing with assym and symmetrical clipping. I was actually troubleshooting that decoupling cap at the end for the last few days, but seemingly the problem was elsewhere. Would be great knowledge have .. thanks for the tip

Simon Barden
22-07-2020, 04:46 PM
Small Signal Audio design by Douglas Self is a great book on audio electronics. Now in an updated (2020) third edition. https://www.amazon.co.uk/Small-Signal-Audio-Design-Douglas/dp/0367468956/ref=sr_1_1?dchild=1&keywords=small+signal+audio+design&qid=1595407224&s=books&sr=1-1

But like most technical books, it's not cheap. But it's one of those books that will help you expand your understanding of audio circuits in general. I doubt it's light reading (I haven't got it), but if you want to do more designing rather than just copying circuits or making basic modifications, it could be a very useful book.

jugglindan
22-07-2020, 05:02 PM
Small Signal Audio design by Douglas Self is a great book on audio electronics. Now in an updated (2020) third edition. https://www.amazon.co.uk/Small-Signal-Audio-Design-Douglas/dp/0367468956/ref=sr_1_1?dchild=1&keywords=small+signal+audio+design&qid=1595407224&s=books&sr=1-1

But like most technical books, it's not cheap. But it's one of those books that will help you expand your understanding of audio circuits in general. I doubt it's light reading (I haven't got it), but if you want to do more designing rather than just copying circuits or making basic modifications, it could be a very useful book.

Yes, it does look good, and definitely won't gloss over details like Jack's book does. But at nearly 120AUD I might give it a miss for now. The Orman book does a good job of explaining how things work, and what the design options are in general terms, so it makes a good starting point.

JohnH
22-07-2020, 05:32 PM
Might have to get a copy

Simon Barden
22-07-2020, 05:44 PM
AC coupling itself doesn't determine the type of booster clipping, though you'd normally only use a coupling capacitor before and after the boost circuit in an asymmetric clipping circuit that added a small a DC offset to achieve the asymmetric clipping. No point in adding components in the audio path that aren't required.

The capacitor in the audio signal path before the boost stops any DC voltage getting back upstream of the boost/clipping section, and the capacitor in the audio signal path after clipping section removes the applied DC offset so that you just have the AC component of the (now asymmetrically clipped) audio signal as the output. The 'booster' part of the boost/clipping section is the pedal's gain control, and will set the level of the signal passed to the clipping section. A higher signal level before the clipping section will give more signal clipping, so more overdrive/distortion.

Symmetrically clipped signals use the same type and number of diodes to clip the positive and negative halves of the audio signal. So the positive and negative halves of the waveform have the same basic shape. (In reality the initial audio waveforms themselves are rarely truly symmetrical, but assume you are passing a sine wave through the circuit).

Adding a small DC offset is one way of achieving an asymmetrically clipped signal. The DC offset shifts the whole audio signal up (or down), so by adding a +0.1Vdc offset, instead of both the positive and negative peaks reading say +2v and -2v, the positive peak now peaks at +2.1v and the negative peak is now at -1.9v. At this point, the audio signal waveform is the same basic shape, but its centre line now rests at +1v, not 0v. If the signal then passes to the diode clipping circuit, then the signal peaking at +2.1v will clip its half of the signal a lot more than the other half peaking at -1.9v. So the resulting waveform is no longer symmetrical. Diodes will typically start to clip the signal once it exceeds 0.7v, so the resulting signal is now clipped and peaking at ±0.7v. Passing the signal through another capacitor removes the DC offset, so the signal now peaking at +0.6v and -0.8v, with the shape of the top and bottom halves of the waveform looking very different.

As the maximum output level of the audio waveform is fixed by the diode forward voltage value, then it gets passed to a volume boost circuit, which can raise or lower the output level to your requirements (but without affecting the wave shape).

Another way of asymmetrically clipping a signal is to use different types of diode for the positive and negative halves of the signal, that have different forward voltage values, and so will clip at different starting signal levels. So one may clip at 0.65v and the other at 0.75v. There is no DC voltage introduced in this case, so there is no need to isolate the clipping section with capacitors either side of it.

Or for even more flavours, you could combine a DC offset with different diode types.

I believe you can also use different numbers of diodes in series for the positive and negative halves. You'll need a bigger signal level to do this, so the gain section will need to be beefier, but you can probably get more asymmetry this way (if you want to).

Note that I may have got some details of that wrong, as it's a very long time since I did any of this stuff in anger!

DrNomis_44
22-07-2020, 07:44 PM
Maybe, once I get back to Darwin, I could breadboard a diode clipper circuit and do a screenshot of the display of my oscilloscope so we can see what diode-clipping actually does to a guitar signal, has anyone checked out my Oscilloscope tutorial thread?

jugglindan
22-07-2020, 08:21 PM
I mostly know what the coupling is doing in the circuits. Just pointing it out as an example of things that the book glosses over. The passage "our signal is AC coupled into the booster so the clipping takes place symmetrically" is talking specifically about an op amp booster with no clipping diodes anywhere. The DC offset is applied to get the input signal centered in the middle of the voltage range handled by the op amp so that when it does clip, it will tend to clip symmetrically. Pretty sure this is one reason for the standard voltage divider in op amp drive pedals to get Vr of around 1/2 of the supply voltage. The largish capacitor after the booster gain stage then removes the DC offset, getting the signal back to centered around 0 volts (more or less).

In my pedals I have done asymmetric clipping by various methods including uneven numbers of diodes in each direction (multiple diodes on one side in series to raise the voltage at which they clip), by having no diodes on one side, or by using different types of diodes with different forward voltages. They all sound different, and there is an interaction with the amount of gain. On my TS clone, some diode combinations sound the same as the diode lift position because the voltages aren't getting high enough to cause audible clipping. On this circuit I found single diodes of different types in each direction gave the best asymmetric clipping. Although it can be worked out in theory what they are doing to the signal, it would be interesting to visualise that.

Dr Nomis: yes, I have been reading your oscilliscope tutorial. It's also bookmarked for when I finally get one of my own.

jugglindan
22-07-2020, 08:29 PM
I believe you can also use different numbers of diodes in series for the positive and negative halves. You'll need a bigger signal level to do this, so the gain section will need to be beefier, but you can probably get more asymmetry this way (if you want to).
That's exactly what I noticed when experimenting with my TS clone. Multiple diodes in series didn't clip much. Blue LEDs (any colour except red actually) didn't clip noticeably unless I maxed the gain, the guitar volume, and stacked a booster to really push the pedal gain stage. I think I ended up using a 1N4001 (measured Vf 0.6v). It sounded a bit more open, less compressed and had higher volume than the stock 1N4148/1N914. All of which mean it clipped a bit less. I just left 1 diode in so it's asymmetric - 1 side clipped and the other isn't.

But of course, as you point out if I had higher voltages coming from the gain stage the result would be different.

Simon Barden
22-07-2020, 09:24 PM
If you don't have an oscilloscope, but you have a DAW, you can simply use that to show you the waveform. Not quite as immediate as an oscilloscope, and you may miss some of the finer detail due to the limited bandwidth, but good enough for rock 'n' roll. All the important stuff for your ears is in the audio frequency range anyway, especially as your guitar/bass amp speaker will filter out most of the frequencies above 6kHz or (depending on speaker size). So you'll actually hear a slightly different waveform to that your oscilloscope or DAW shows you.

DrNomis_44
23-07-2020, 01:54 AM
There's a free vst plug-in Oscilloscope, called Smexoscope, that you can download and install and run in any Daw that supports vst plugins, I think there's only a pc version:

http://bram.smartelectronix.com/plugins.php?id=4

jugglindan
23-07-2020, 06:06 AM
There's a free vst plug-in Oscilloscope, called Smexoscope, that you can download and install and run in any Daw that supports vst plugins, I think there's only a pc version:

http://bram.smartelectronix.com/plugins.php?id=4

Doh! That's a good reminder. I already have Friture (http://friture.org/) which can capture audio from my THR amp via USB. It has very little latency. I always forget I have it though, and only used it once to see what the tone control on my Direct Drive clone is doing (not much as it turns out, just cuts a bit of high freq noise - could try changing the cap but I like the sound and don't want to mess with it). As well as a scope it also has various frequency domain displays.

It's open source and supports Linux, Mac, and PC. Worth a look.

36607

Simon Barden
23-07-2020, 03:11 PM
It's often just as easy to record the audio and then look at the captured waveforms. As you aren't generating regular sound pulses that you could set your timebase to in order to display in a stationary manner, you'd certainly need a freeze/capture function on any scope/plug-in to see a still picture of the waveform.

Once the sounds are captured, you can look at them at leisure to your heart's delight.

jugglindan
23-07-2020, 03:18 PM
I frequently test pedals by putting my looper in front. Lets me just loop the test music while I adjust trimpots or whatever. That would also work for this since I don't think I can get Friture to consume a recorded track directly (might be able to fake it with a loopback or something).

Simon Barden
23-07-2020, 03:45 PM
You wouldn't need it to, as DAWs will display the waveform directly.

But a looper is a good test tool (unless you are using something like a Fuzz Face or some wahs, that really need to be directly connected to the output of guitar).

Bakersdozen
25-07-2020, 06:06 AM
Any ideas on this circuit? I am currently playing around with it and sounds really good, however I'm getting the fuzz dying out and a kind of starved fuzzing out once the bulk of the sustain from the signal has gone. The fuzz lasts for around 4 or 5 seconds, when playing a chord with a decent attack on the strings then dies away with a BZZT breaking up kind of decay. The chord is still ringing out of the guitar but the tail of the Singal goes towards clean then.

I have swapped opamps and the diodes with different ones and although the fuzz tone does get altered, the starved decay still occurs in the same way. I've also tried with different power supplies to ensure that the voltage isn't starved and contributing to this.

Do you rekon it's just the circuit, or would it be likely the electrolytic cap maybe leaky/possibly faulty. It would probably be the next thing I try, once I get a new cap in 10uf. Or is the issue somewhere else? I do have spare 22uf or 47uf elec caps here, should I give them a go?https://uploads.tapatalk-cdn.com/20200724/d532393270602567979aa2667a3249d8.jpg

jugglindan
25-07-2020, 06:29 AM
I am not an expert (or a fan) of fuzzes, but it sounds a bit like gating which is just part of the design of some fuzz circuits. On some transistor fuzzes, as the signal level starts to drop, the transistors go into cutoff. I have no idea how op amp fuzz behaves.
I would try the 22u cap, as it looks like it's just used to filter out the DC offset before the signal hits the output. 22u should work for this. But there is a chance it's just part of the design.

Are you building on breadboard or soldered? I always like to breadboard unknown circuits like this to see if I like how they sound.

Bakersdozen
25-07-2020, 06:35 AM
Thanks DC, this is straight on Vero, as I dont have breadboard yet. I see how breadboard will be a good investment though. Does it act like Vero in long rows? Or do you have to create the circuit as such?

jugglindan
25-07-2020, 07:01 AM
I am not sure that the 10u cap would cause that sort of problem. The impedance of a cap varies with frequency not voltage level, although I am not sure how a fault would affect that.

Breadboards are not connected the same as veroboard. They are like this:
36662
36663
36664

All images taken from this intro to breadboarding (https://computers.tutsplus.com/tutorials/how-to-use-a-breadboard-and-build-a-led-circuit--mac-54746). It's a short read and a very well written intro to using a breadboard. For myself, I find it easier to breadboard a circuit direct from a schematic, but it can also be done by systematically working through a vero layout. I just find it a little harder since the layout is arranged for compactness not readability.

A lot of the Tagboard Effects layouts are modified from available schematics though, so it can be worth working directly from the layout you plan to build.

You can pickup a decent breadboard from Jaycar for not too much money. Here's how I have mine setup for pedals:
36665
Other holes on the angle bracket are drilled to accept switches as required. No power switch, I just add/remove the ground lead from the battery. That circuit is 3/4 of a Rat (no tone control or output buffer at that stage).

Bakersdozen
25-07-2020, 07:07 AM
Awesome, thanks DC. I will have a read and will force me to polish up on schematics.

Simon Barden
25-07-2020, 02:55 PM
A circuit diagram would help for that fuzz, rather than just the component layouts.

I'd guess that the signal boost could be increased, so that as the input signal level drops, more of it is pushed up to clipping level. So you could try adjusting the booster gain resistor values.

Bakersdozen
25-07-2020, 04:38 PM
A circuit diagram would help for that fuzz, rather than just the component layouts.

I'd guess that the signal boost could be increased, so that as the input signal level drops, more of it is pushed up to clipping level. So you could try adjusting the booster gain resistor values.Thanks Simon. I actually stumbled upon the answer when searching for the schematic. Someone else was having same issue. I must have been using the wrong terminology when searching previously (so much to learn).

Turns out the schematic had missed a resistor in parallel with the diodes, which is obviously what this layout was copied off.

I'll give it a burl tonight.

Joe Garfield
22-09-2020, 04:43 AM
I picked the Jack Orman ebook Guitar Effects Explained from Amazon today. So far I am finding it very readable. It avoids too much math, and reads like a plain English description of how effects work at the circuit level.

I highly recommend this to anyone interested in building pedals, and also to anyone who just wants to know more about the magic.

There is some assumed knowledge required to understand parts of it, but nothing so far that a quick Google search can't fix. For example, the early section on boosters says "our signal is AC coupled into the booster so the clipping takes place symmetrically". Jack doesn't explain what coupling is though, or how it leads to symmetric clipping in the gain stage. Hopefully he explains the related role of the coupling capacitor after the gain stage.

Sent from my moto g(7) using Tapatalk

Thanks, for $10 I think it's a good recommendation! I looked at the preview and enjoyed it. It does leave me with questions, but if it's giving me enough information to ask legitimate questions then it must be doing something right! I didn't really understand the part about 'direct AC coupled' but realized the point was that it's a symmetric system with no other components in the way to influence that. I think it does require at least familiarity with electronics terms, but definitely not a degree in electronics.