Faultfinding DIY Pedal Builds

[DrNomis_44]

Rightio, what I'm going to be doing today is I'm going to post some screenshots of my DSE Q-1804 20Mhz Oscilloscope being used to look at the output signal of some of the effects pedals that I have in my collection, I'm going to focus mainly on pedals that produce a distortion effect since you can clearly see on the Oscilloscope's display screen what the pedal is doing to the signal from the guitar, in each case I'm going to be using my Fender Strat set to it's neck pickup, and with a capo placed just behind the 12th fret, as my signal source, the capo is so that I can free up one of my hands for holding and operating my digital camera.

The pedals I will be posting screenshots of are as follows:

Red Jim Dunlop JD F2 Germanium Transistor Fuzz Face.

DIY Tonebender MkI Germanium Transistor Fuzz Pedal.

DIY Tonebender MkII Germanium Transistor Fuzz Pedal.

DIY Rangemaster Germanium Transistor Vintage Treble Booster Pedal.

Ibanez TS-9 Tube Screamer.

Boss BD-2 Blues Driver.

DIY Baja Real Tube Overdrive Pedal.

DIY Doctor Overdrive Pedal.

DIY Big Muff Distortion Pedal.

Boss MT-2 Metalzone Distortion Pedal.

[DrNomis_44]

First up, it's my Red Jim Dunlop JD-F2 Germanium Transistor Fuzz Face.



Here's what the un-effected guitar signal looks like on the Oscilloscope's display screen:



Here's what the output signal of the JD-F2 Fuzz Face looks like with the pedal engaged and the Fuzz control set to minimum:



And here's what the signal looks like with the Fuzz control set to maximum:




At minimum Fuzz settings we notice that the bottom of the signal is gently rounded while the top is pointy, when the Fuzz control is turned up to maximum we notice two things, firstly the top and bottom are sharply squared-off, but they are also un-equal, this is called "Asymmetrical-Clipping" and is largely responsible for the reedy, nasally sound that the Fuzz Face produces.

[DrNomis_44]

Next up it's my DIY Tonebender Mk I Germanium Transistor Fuzz Pedal.



Again, here's what the un-effected guitar signal looks like:



Here's what the output signal looks like at the minimum Attack setting:



And here's what the output signal looks like at the maximum Attack setting:




We can see that at the minimum Attack setting that the output signal shows some sharp asymmetrical clipping, when we turn up the Attack control to maximum, we can see that it just gets more heavily clipped asymmetrically, notice how the's a bit of sloping of the top and bottom of the heavily-clipped signal, that's caused by a fair bit of high-frequencies in the resulting tone so therefore the Tonebender Mk I will sound brighter and buzzier than a Fuzz Face.

[Marcel]

Re your source signal photos, or 'un-affected' shots......Are you injecting real guitar or is it from a signal generator?

Real guitar then maybe okay, but if a sig gen then I'd be checking it as it doesn't look right. Even a single note from a guitar should look more sinusoidal...

[DrNomis_44]

Re your source signal photos, or 'un-affected' shots......Are you injecting real guitar or is it from a signal generator?

Real guitar then maybe okay, but if a sig gen then I'd be checking it as it doesn't look right. Even a single note from a guitar should look more sinusoidal...

Yep, it's an actual real guitar signal straight from the output jack of my Fender USA Strat guitar, going straight into the pedal with nothing in between except the instrument lead, I've connected the input of the Oscilloscope directly to the output of the pedal in each case via a short pedal patch cord and a lead I made up using a BNC connector, a piece of Response WB-1530 mic lead and two alligator-clips.

The pickups on the Fender USA Strat are a set of three Kent Armstrong Dual-Blade Humbucking pickups that are Single-Coil sized, and for each pedal I set the pickup selector to the neck pickup, all the controls on the Fender USA Strat were set to full-on.

The strings on the Fender USA Strat are getting a bit old, and I guess I should have put new ones on before doing the screenshots except that I'm out of both new strings and money at the moment.

There's a reason why I decided to use my Fender USA Strat as a signal source, most Signal-Generators usually have a low-impedance output and I've found that some vintage-style Fuzz pedals behave differently if a low-impedance signal source is connected up to their inputs, the Fuzz Face is a classic example, I wanted to show how the pedals behave when a real-world guitar signal is fed into them.


All my DIY pedals use what's called a "True Bypass" footswitching system, that is, the signal doesn't go through any buffers, when bypassed, the signal goes straight from the tip-connection of the input socket through to the tip-connection of the output socket, the effect circuit is taken completely out of the signal path.

[DrNomis_44]

Okay next up is my DIY Tonebender Mk II Germanium Transistor Fuzz Pedal.



As before, here's the signal from the guitar:



Here's the output with the Attack set to minimum:



And here's the output with the Attack set to maximum:



This time we notice a few interesting things, firstly at the minimum Attack setting the output signal is heavily clipped, it is also a bit more symmetrical (although there is still some asymmetry) compared to the Tonebender Mk I, and notice how the sloping of the top and bottom of the signal is more pronounced?, that means the sound is going to be much brighter and buzzier.

Turning up the Attack control to maximum just clips the signal harder.

[DrNomis_44]

So, what we learn from that is that most Fuzz pedals tend to behave in a similar way, taking the guitar signal and converting it into either a square-wave or pulse-wave, the sharp clipping is responsible for the buzzy-ness of the resulting tone, the symmetry affects the overall tonal-colour, or timbre, and the spikiness of the clipping determines how bright the tone sounds.


One thing to note is that what you see in the screenshots I posted is what I saw on the Cathode Ray display screen of my 20Mhz Analog Oscilloscope, if you were to use say a 100Mhz Digital Oscilloscope the clipping would look sharper because of the increased signal-bandwidth of the Scope.

For those of you who are interested in getting an Oscilloscope, you can sometimes find second-hand ones on sale on eBay for reasonable prices, for most signal-tracing work a 5Mhz Single-Trace Oscilloscope will do a good enough job if you're on a tight budget, you can sometimes pick up a good one for about $100.00 or so, make sure it's rated to run on the AC mains voltage of the country where you live though.


More screenshots to come....stay tuned.

[Simon Barden]

What you need to take into account is what you actually hear through a guitar amp compared to those oscilloscope waveforms will be something different again; the speaker acts as a low pass filter, so you won't get many of the upper harmonics produced by the distortion unit, resulting in a much 'softer' sound. A direct signal from the fuzz (or distortion or overdrive) unit straight into your audio interface will sound a lot harsher and not very pleasant. Even that will be filtered by the A/D sampling rate setting. Even at a sampling rate of 192kHz, the audio bandwidth is a bit over 80khz (depending on the filter used), which is a lot less than the 20MHz of the 'scope. And unless your monitors have ribbon tweeters, your playback system (be it a studio monitor or hi-fi speaker) will have around a 20kHz frequency response (maybe double that with ribbon tweeters).

Doc, any way you could show some of that effect? Maybe record the output into your system, then get screenshots of the raw, straight playback and playback through a cab sim waveforms?

[DrNomis_44]

What you need to take into account is what you actually hear through a guitar amp compared to those oscilloscope waveforms will be something different again; the speaker acts as a low pass filter, so you won't get many of the upper harmonics produced by the distortion unit, resulting in a much 'softer' sound. A direct signal from the fuzz (or distortion or overdrive) unit straight into your audio interface will sound a lot harsher and not very pleasant. Even that will be filtered by the A/D sampling rate setting. Even at a sampling rate of 192kHz, the audio bandwidth is a bit over 80khz (depending on the filter used), which is a lot less than the 20MHz of the 'scope. And unless your monitors have ribbon tweeters, your playback system (be it a studio monitor or hi-fi speaker) will have around a 20kHz frequency response (maybe double that with ribbon tweeters).

Doc, any way you could show some of that effect? Maybe record the output into your system, then get screenshots of the raw, straight playback and playback through a cab sim waveforms?

That's true and you do raise a good point, I guess what I could do is record the sound of each one of the pedals going into my Marshall amp that's miced up with my Scarlett CM25 condenser mic(it's the only good mic I have), I'll need to sort out my account with Soundcloud first though cause I only have enough play-time on it to upload a 4 minute audio clip, but it is entirely do-able.

[Marcel]

Something else to consider is the load impedance that the pedal/guitar is 'seeing'. Most amps are typically 1M ohm input impedance, so is your Cro which is replacing the amp also providing a 1M ohm load to the pedal/guitar? I'm suspecting your Cro's impedance may be a little higher leading to the 'discrepancy' that I think I'm seeing in your bypassed screen shots...

1M per volt is not a 1M ohm load.. and a x10 probe is definitely not 1M ohm i/p Z.