loading

Add Diode-Clipping Distortion to your Guitar Amp

Featured
Picture of Add Diode-Clipping Distortion to your Guitar Amp
Here's a relatively simple way to add some "bite" to your old guitar amplifier.

Amplifier overdrive and distortion is usually achieved by signal clipping--pushing the gain until the signal peaks are chopped off. "Real" tube overdrive isn't possible without major changes (adding additional preamp stages, etc.), but here's another option:

Preamp clipping can be simulated by installing a diode-clipping circuit. In fact, some big-name amp manufacturers (Marshall and Fender, for instance) have used diode clipping for a more ballsy sound (especially at low volumes.) In fact, it's made a bit of a resurgence in boutique amps...

The danger of converting a tube sound into a harsh "solid state" sound exists. Tubes clip with a softer "shoulder" and have a more pleasant, warm sound....of course, it's a matter of taste, after all.

But a carefully designed diode-clipper circuit can mimic those tube characteristics. "Peak clipping" more closely approximates tubes than "crossover clipping," which is common to transistor design. Plus we can utilize asymmetrical clipping to emphasize the "correct" harmonics. And since our new sound is contained within a tube amp, it will help to soften the effect.

Of course, diode-clipping can be used (and is) for solid state amps, too.

Video clip shows the three switch-selected settings: 1) clean; 2) medium clipping; 3) max clipping:


Warning: This is a "mod," and does require a bit of rewiring. All the usually cautions about discharging power supply filter capacitors apply here. To restate them:

Don't touch the internal components of a guitar amp without first draining the filter capacitors. Please. You could really harm yourself. Here's a link to my older amp project, and the section Discharge those Caps!
 
Remove these adsRemove these ads by Signing Up

Step 1: Let's examine diode clipping...

Picture of Let's examine diode clipping...
unclipped.jpg
Wow--at first glance, a diode connecting the the signal path to the ground looks like a short circuit! How can this work?

In theory, diodes conduct current in only one direction. But they don't operate "perfectly." All diodes have a "forward voltage" -- they won't conduct until that voltage is reached.

If a diode is connected between the signal path and the ground, the signal won't be channeled (shunted) to ground until it exceeds the forward voltage. And even then, the whole signal isn't dissipated, only the portion of the signal above that forward voltage. So it's peaks are simply "chopped off."

But it also results in more attenuation, as some of the signal is lost. That isn't all bad! Clipping the loudest parts is also a form of "limiting" -- you loose some dynamics, but accentuate the lower-volume sounds.

Plus, many amps don't begin to distort until turned up too loud for many venues. Like a stomp-box, diode-clipping lends a killer sound at a lower volume. Your family, roommates and neighbors will thank you.

Step 2: Types of diode clipping

Picture of Types of diode clipping
asymmettrical.jpg
The clipping effect works in both directions relative to ground--so a circuit that clips both the positive and negative signal peaks is two diodes connected in opposing directions. The diodes can be chosen to pass as much of the signal as necessary. More clipping equals more distortion.

The first illustration shows:

1) An unclipped signal
2) Asymmetrical clipping, one side of the signal (not limited to a single diode.)
3) Symmetrical clipping, where both the + and - signal is clipped equally.
4) Asymmetrical clipping, both sides.

Variation #4, asymmetrical clipping, offers the most natural and "tube-like" sound. Plus, it's the most flexible--any number of diodes in series can be used to shape the effect.

Picture #2 shows just a few variations of asymmetrical clipping. Notice that LEDs are used as diodes in examples B and C! LEDs have a much higher forward voltage than typical diodes, so the clipping can be softer and more subtle.

Step 3: The design

Picture of The design
diodeclip_circuitdrawing copy.jpg
image0001.jpg
As mentioned, using a combination of diodes, including LEDs, is a great way to begin. I've included a schematic, a wiring diagram and a photo (showing just how simple this circuit is.)

After much swapping in-and-out, I choose a combination of one germanium diode, one silicon diode and one LED.

-- a 1N4148 (D1) and a 1N60 (D2) for a combined forward voltage of ~1.05V
-- a red LED (D3) with a forward voltage of ~1.7V

The diode pairings are connected to SW1, an on-off-on double-pole switch. The center setting is "off," or no diode clipping at all. The other two settings are:

-- the diodes connected directly to the signal path.
-- the diodes connected through a pair of resistors (R1: 47K, R2: 100K)

The resistance softens the clipping, which shapes the effect. When the resistors are connected, more of the natural tube signal leaks through. There are many ways to accomplish this, and this is merely a simple example (See the "Other options" step for more info.)

To help you choose your components, here are the forward voltages for some common diodes:

~790mV -- 1N4148 (Silicon)
~265mV -- 1N60 (Germanium)
~1700mV -- LED (red)
~205mV -- Schottky 1N5819
~740mV -- 1n4001 (Silicon)

Germanium diodes tend to have softer transitions which give a non-linear, more "tubey" sound. But silicon diodes can give a sharper "square wave" metallic distortion, if that's your thing.

Before you build it:

It's important to understand that the peak signal voltages will differ greatly, amp-to-amp. No one combination of diodes will give the same effect in different amps.

And no one combination of diodes sounds good to everyone, either. Experiment! Try 2, 3, 4 or more diodes in series. Keep the sides unbalanced, or use switches to tap the signal in different ways.

(Note: on some amps, the LED(s) will actually light up--they don't in my amp, the peak voltages aren't high enough.)

Step 4: Inserting the circuit

Picture of Inserting the circuit
single channel Kelly.jpg
jcm900.jpg
Exactly where should the clipping circuit be added? Well... first, where NOT to place it...

--Don't insert at the input circuit (before the preamp.) Guitar signals are quite weak, and even if the circuit functions partially with a really hot pickup, it likely won't work at all with other guitars.

--Don't insert in the driver portion of the power amp (output transformer primary.)

--Don't insert in the output transformer secondary / speaker loop.

The last two examples will greatly attenuate (lessen) the volume of the amp, probably "fry" the diodes, and possibly ruin the power tubes or the output transformer.

So, where should it go?

--Within, or just after the preamp circuit. The red line illustrates the signal path between the preamp and the output stages. This amp has a single preamp stage, so this is the only acceptable place (in this case.)

The points B and C are all possible places to connect the clipping circuit. I chose C, since this amp is never too loud, and turning down the volume will change the character slightly. Connection point B are probably better if you wish to retain the full clipping effect when the volume control is turned down. Point A is unsuitable, due to the DC plate voltage.

Amps with a dual-tube preamp are ideal--any signal attenuation lost in the clipping can be recovered by increasing the gain of the second preamp stage. (Of course, that pre-supposes that you have the expertise to do so...)

There will be some attenuation with diode clipping--some signal will be lost. This isn't really tragic, however. Most players would prefer to get "that sound" at lower volumes.

Step 5: Other options...

Picture of Other options...
Although I chose an on / off / on circuit, the amount of distortion can also be controlled by simply adding a POT. Now the effect can be adjusted, and the desired asymmetry "dialed in." It should be noted that I've add a fixed "mix control" to mine--the two resistors. But a variable control might be of value....

And there are many diode types we haven't covered: zener, silicon, other LED colors, etc. Even MOSFET transistors have a voltage drop.

Flipping the asymmetry--switching the directions of all the diodes, will also have an effect. This can emphasize odd or even harmonics. Choose whichever sounds best to you.

There are a great many variations of diode clipping circuits. Here are a few links to get you started... :
-- DIY Fever: Diode Clipping
-- MOSFETs as diodes
-- AMZ - Diode clipping staturation controls
-- AMZ - "Warp" controls--balancing diode clipping
-- AMZ - Mosfets and Zeners as clipping diodes

Step 6: But how does it sound?

Well, IMO, pretty good...

It definitely gives a harder-edge to the natural distortion of my amp. But it's a small amp, and the attenuation is fairly significant in the "full-on" setting.

At least one thing must be noted:

A simple mod like this isn't going to give you an over-the-top metal-chorus-delay-flange-mega-distortion effect. You're better off using a stompbox or fancy rack-mount fx unit. But if you like the sound of your amp clean, you probably will like it even better with a little "dirt."

To be honest, a good tube amp with some flexibility really brings out the character of different guitars better than the external FX. And you can always use the stompboxs with the mod anyway...
1-40 of 67Next »
valve.king.510 months ago

I have been wondering if I connected a rectifier tube in the same manner, would it have the same effect in a tube preamp circuit? Isn't a rectifier just a big ole' gassy pair of diodes in tube form?

gmoon (author)  valve.king.510 months ago

Yeah, you can use the diodes in a tube rectifier for a clipping effect. I've never tried it, but you can search the web and find examples.

freeza362 years ago
I love the sound of overdrive. I bought a tube overdrive pedal for $20 at Guitar Center because they came out with a new version. Score!
Calligraphy6 years ago
So I have a cheap-o First Act bass amp in combination with a Peavey 126 and I want to have my first act act as a dirty channel.(since the volume is lower) Would I have to make any changes in these plans to accommodate bass freq.? This is probably a stupid question but oh well.
Btw, first act guitars are not all that too baad, i refinished one and it came out pretty nice. and im pretty sure that the bass would not have a difference with that schematic.
gmoon (author)  Calligraphy6 years ago
The most important consideration is the peak voltage of the signal--where ever you decide to install the clipping diodes. The diodes shouldn't change the frequency.

Find a point along the signal path after a coupling capacitor and before an amplifying component (transistor, opamp, tube, etc.)

Look at step five for ideas to control the amount of clipping.
eyeball2264 years ago
I just had an idea while reading this... what if you were to have 2 pots (only using 2 lugs), one for each diode. Then you'd have constantly variable asymmetrical clipping. What do you reckon?

gmoon (author)  eyeball2264 years ago
Yeah, I'm sure that's one way to do it...That AMZ article notes quite a few.

It should be reiterated that the total voltage drop for the diodes should be calculated for the peak voltage of the signal path, and not just taken verbatim from my project...
oweng4000 gmoon4 years ago
You have the basis for a circuit by tim escobedo called the "Simple Square Wave Shaper made simpler."
http://folkurban.com/Site/SimpleSquareWaveShaperMadeSimpler-712.html.
Check It out
gmoon (author)  oweng40004 years ago
That particular example (note the capacitor, in particular, and read the description) doesn't really work like this one...

Incidentally, diode-clipping distortion is MUCH older than any of Tim Escobedo's circuits. TONS of guitar effects use it.
oweng4000 gmoon4 years ago
I meant what eyeball226 was talking about. I use diode clipping for all sorts of circuits. I didn't mean that you where copying tim escobedo.
gmoon (author)  oweng40004 years ago
Ah, got it.

Tim E. made some cool (and some oddball) FX...
GatedSound5 years ago
I'm confused at the wiring to the lugs of a pot. Which are the diodes attached to, and which goes to the signal?
gmoon (author)  GatedSound5 years ago
The "wiper" (the arrow) is always the center tab of any POT. Here you see that goes to the signal path.

The other two tabs are identical for linear POTs, other than effecting the direction (whether resistance changes up or down, right or left.)

Log or audio POTs also have a taper, so greater "change" is bunched up on either side. They'll have to be tested in the circuit.

Check out those AMZ links (Step 5) on saturation and warp controls for more options...
fastcar1235 years ago
so im thinking of triing this and i always love to make things at easy access so wat i want to know is can this be embeded into the guitar itself?

also would it matter what color the leds are and can i use leds for all of the designs or would it sound beter with real diodes ?

this does work with transistor amps right?
gmoon (author)  fastcar1235 years ago
This can work in the guitar itself, but remember that it clips peaks, so the output of the pickups will be reduced.

Each diode type--LED, silicon or germanium--has a "forward voltage drop," and that determines when clipping starts.

General purpose silicon diodes have the smallest voltage drop, So they will clip at a lower voltage (MORE clipping.) LEDs have a larger voltage drop (less clipping), and different color LEDs have different voltage drop.

Transistors amps will work too. Just place the diodes after a coupling cap.
gmoon (author)  gmoon5 years ago
Sheez, I mean general purpose germanium diodes have the smallest voltage drop...
so this means i will just need to turn the amps at a higher volume right? or does this work like the volume control and grounds the signal and some of the gain

silicon diodes are the plastic ones that are usually black or grey in color right?
do 3mm leds have lower voltage drop than 5mm

what is a coupling cap
gmoon (author)  fastcar1235 years ago
Yes, there will be some signal attenuation. That's a "feature" of diode clipping--more distortion at lower volume. You can always switch it out for more (clean) volume.

The color of the LED is more important than it's size. Check the diode specs., it will indicate Si or Ger.

Capacitive coupling -- amplifier stages are separated by a coupling cap. Stages amplify DC voltages as well as the AC signal voltage. The cap blocks that additional DC.

A clipping diode must be placed after that cap, because the clipping won't work with the DC offset voltage present. The diode would remove the AC signal entirely.

but it wil still sound as good right?
 

so is this stuff sold in radio shack or do i need to buy online

ok none of this matters if it is inside the guitar right?

 

gmoon (author)  fastcar1235 years ago
If you build it before the amp (in the guitar--or in a stompbox), you'll probably have to raise the signal a bit. You could do that with a simple boost circuit, then follow with the clipping diodes.

But at that point, you might just start thinking about building a "real" stompbox from one of the many plans online. That would get you more predicable results, since it will take some experimentation with the diodes, etc., to come up with a workable sound.

The diode setup I used was "custom fitted" to the amp. I hope that's clear from the 'ible--that every amp is a little different.
aa boost circuit just makes up for the lost volume right? so is it easy to build and is it cheap ?

i ve thought of bulding one but i dont have any way of powering them (no 9V or AC adapters) also i like to make things as easy as possible

so if yours was custom how do you think i should do mine (just useing signal then wtih the booster)
gmoon (author)  fastcar1235 years ago
Yep. And it increases the signal strength (voltage) to the point where the diodes have a more pronounced effect.

Surely you can get your hands on a 9V battery or two?

Customizing the effect--

-- Find a likely spot in the signal path; EARLY in the signal path. Not at the very input, but definitely not in the power amp section.
-- Test it for DC voltage with a voltmeter. Should be zero (it's after a coupling cap.)
-- Ideally, you would then use an oscilloscope to find the AC peak voltage (help to chose the diodes.)

You probably don't have a 'scope. So try this instead--

-- Find a likely spot in the signal path; EARLY in the signal path. Not at the very input, but definitely not in the power amp section.
-- Test it for DC voltage with a voltmeter. Should be zero (it's after a coupling cap.)
-- Breadboard the diode clipper. Connect one side to the GND. Then very carefully jumper the other end to the signal path. You might have to solder a test lead in place, to prevent shorting the amp with your jumper connection.

DISCLAIMER: Hey--even a running solid state amp COULD have enough voltage to kill you. YOU PROBABLY SHOULDN'T BE DOING THIS WITH THE AMP ON--OR OFF, FOR THAT MATTER. These instructions are for educational use only. THIS IS POTENTIALLY DANGEROUS.
yes i can can get some 9V so is that all a boost is? just a 9V before the diode circuit?

i dont have a voltmeter either so would i just use some diodes htat you have used here?

I didnt know that solid state amps are potentially dangerous because ive disassembled my amps 4 times i think so that concerns me

but doesnt the guitar have a very low (if any) stored electricity when unplugged?

so could i damage my pickups by putting a 9v in its path because they are passive pickups

i was also wondering what would happin if this low volume circuit was turned to a high volume?

the overall question is it this possible to put this in the actual guitar with a booster curcuit with no damage to my equipment?
gmoon (author)  fastcar1235 years ago
I gotta give a disclaimer--especially if the amps' internal wiring is hooked up temporary via breadboard. Even SS amps could have 24-48 volts in the PS, which is enough...

There's more to it than just the battery...but a boost circuit doesn't need to be very complicated. Here's the Fetzer valve circuit (which itself gives some overdrive, too.) It's just a single transistor and some discrete components. You could add diode clipping to that very easily.

You could try the circuit as-is, but like I said: silicon diodes will clip the most, LEDs the least... Play with different combinations. There will be no danger if working with a low-voltage boost.

Using a correctly-wired boost will not harm your guitar. But you shouldn't try wiring a battery in there by itself--it won't do anything, anyway. I was merely asking if you had access to batteries, since originally it seemed to be a problem...

And the guitar itself poses no danger if unplugged.
ok 48V isnt enough to kill just to know you back right?

i can try that

and the best way is to have asymetrical on both sides

i can get 1 or 2 batteries yes

\i didnt think it did i just needed to double check from someone more experianced
what about if it is plugged in and the amp is running
gmoon (author)  fastcar1235 years ago
From Wiki on Electric shock: however, death has occurred below this range, with supplies as low as 32 volts.

And just to drive the point home, here's Tim's 'ible on DC welding using batteries. It's a 36V setup. You wanna take a chance with that? Even if it doesn't kill you, it could do major damage...
thats pretty bad  i need to be more carful with what i do ive been known to fix things and do lots of tests with low voltage things but now it doesnt seem so voltage now
ok so ive bulit mine useing 1n914 type diodes but it doesnt seem to do anything i get the same sound no matter where the switch is and ive built it the same way you did only in guitar itself
Any tips on finding the part where pre amp and power amp separate?
gmoon (author)  LeviMan_20015 years ago
Anywhere following a coupling cap (C2, for example) should be good.

The coupling cap will remove any DC bias.
indrekhaav5 years ago
You mention that you chose, quote, "a combination of two germanium diodes and one LED", yet immediately afterwards you specify a 1N4148 which is a silicon diode. Was it a typo?
gmoon (author)  indrekhaav5 years ago
Nope, just me assuming that all glass-envelope diodes are germanium...which were the most common type when I was growing up...

Obviously I was confused about this myself when I wrote it--I list 1N4148's as either Germanium and Silicon.

Definitely has a forward voltage appropriate for SI, though.

I'll need to fix this soon, thanks.
No problem, and thanks for the clarification.
Lenny245 years ago
I've Used A Zener Diode in my Diode-Clipping Circuit. It doesnt' Work really good, but that could beause I only used a LM741 as "Preamp Stage", and then a 1-Watt-Amplifier-IC out of Old PC-Speakers. I Attached the Diodes after the 1Watt Amp and Put an 10K Resistor after that and Connected it to an Old Guitar amp (That hasnt' OD). I also tried Retifier-Diodes. That Worked a lot better.


gmoon (author)  Lenny245 years ago
That sounds about right. Thanks for the info!
Lenny24 gmoon5 years ago
No Promblem! ;)
Mandela6 years ago
I'm still confused at here (im beginner) If guitar voltage output is beetween .1V and 1V, arent the voltage never been cut because the forward voltage is more than 1V ?
gmoon (author)  Mandela6 years ago
Good question. The diodes follow a preamp stage, which boosts the signal peak voltage. But yes, if the voltages don't exceed the forward voltage drop there won't be any clipping.
Mandela gmoon6 years ago
thanks oh yes! i almost forget that you made this effect to the amp ! :D But if i want to make guitar effect (not at the amp) i just have too buy lower(under 1V) forward voltage ? thanks again !
1-40 of 67Next »