Wednesday, May 31, 2023

Don't Know What You've Got, Until It's Gone.

I could drone on and on about things from my past I've lost that now I wish I had back, but one is a bit more widespread than just to myself. Today's topic kind of pertains to video rental stores. Now, I was never a Blockbuster member, nor did I rent there. The closest I came to Blockbuster was the bassist in my high school band worked there, and I went in a handful of times looking for used video games. My childhood bestfriend's Mom may have rented us video games from there on one of my many, many stayovers at his house on the weekends, but beyond that I didn't grow up a Blockbuster patron.

We had a local chain that had four stores throughout their existence that offered it all, and was pretty quick to add new stock. The chain in question was Movieland USA, with their main building in Mooresville Indiana. I remember going in and seeing video games before I saw any of the movies, the latter of which I really had no interest in anyway. My childhood bestfriend's Mom would rent us a video game or two for the weekend and we would make the best out of it from two days. Later on I purchased a few video games they were selling to clear out old stock. I believe they were $5 each or three for $12, and of course I always bought three NES games at a time. All of them came with their original boxes and manuals, albeit with a sticker stain, but they were all in good shape. I purchased a CIB Wayne's World for the NES from there, which I stupidly sold to FukyuLand, I mean FuncoLand, many years later, but I've already told that story.

 Years passed by, I moved out of state and on the odd visit I would sometimes drive by the old Movieland building, which was a historical building on main street in downtown. Movieland must have started feeling the strain of more competition such as RedBox and Netflix, before they were a streaming service, and moved into the basement of the same building, and closed all their other store. They stuck it out and stayed in business until about 2018 and finally called it quits, closing up shop for good. If you want to read the full story you can find it here.

Since then the owner has passed on and the former building in Mooresville was destroyed by a tornado on the 8th of April 2020. There's absolutely no going back, even for a quick nostalgic trip, the heart and soul of the owner, and the physical location of the building are both gone. I'm not sure if I still have any of the NES games I purchased from them all those years ago, but I believe I might have some of their SNES games. It's hard to tell as they're in the boxes, with Xeroxed manuals and only the circular sticker stains would be a sign of having been their rentals, but I can only guess whether or not that's from Movieland USA. So amongst the many things I've lost in my life, it may seem silly, but one little rental store in a small rural Indiana town still calls to me from the past, yet there is no way for me to answer that call.

Friday, May 26, 2023

Never Meet Your Heroes

There is an old saying that goes something like never meet your heroes, you'll only be disappointed. I've found this also applies to components for building guitar pedals. Way back in 2021 I built my ROG Umble that required J201 JFETs, which was a whole new experience for me in more ways than one. Up to that point I had shied away from building circuits that required JFETs solely because they were all obsolete and out of stock, rendering many circuits I wanted to build useless endeavors. Eventually I did manage to find SMD J201s and horribly soldered up my own conversation boards to make them through-hole. Luckily the Umble circuit has biasing trimpots, because my homemade versions needed quite a bit of dialing in. Even though I made my own and the pedal sounds pretty good, after getting it dialed in, I still wondered what it would sound like with real, true J201s.

I dreamed of the day when I might actually be able to touch a real J201, like so many other pedal builders before me. I wouldn't take for granted that I was part of the club, while so many may never have that chance to do so. Well, that was how I felt before I managed to purchase some real J201s. I mean, they're J201s. When I received them I put them into the Umble circuit and it sounded the same, meaning my slapdash homemade J201s were no worse than the authentic through-hole component I had always dreamt about.

I'm not saying they're bad, it's just that I built up this wonderful world in my head based on a component being unobtainium and not on the reality that it's just a J201. With my delusional head I thought that the authentic through-holes would have yielded far better results than the little things I soldered up so poorly. The differences, if any, were imperceptible. I guess it's my fault for thinking they would be any different, or perhaps having little faith in my homemade J201s being any good. What I've learned is components are components and they'll generally do what you expect from them, no matter if they're hard to find or the easier to find equivalent.

Wednesday, May 24, 2023

DIY Sound Probe

When it comes to pedal circuits that give me grief, usually standard troubleshooting, solder bridge finding and "oops, I'm an idiot" admitting will suffice. In the case of my Blues Driver and Kay tremolo clone circuits these methods just aren't good enough, apparently. Both circuits function well, but have their own mysterious issues that ruin the functionality of the circuit entirely. This is why I decided I needed to construct myself a sound probe and see if that could get me anywhere closer to figuring out the problems.

A sound probe is simple enough to build. You take a length of double wire and on one end you solder on a clip for your ground, and a probe attached to the other wire. On the other end you connect the ground and probe wires to a male audio jack of your choosing, using a 100nf capacitor on the probe end, and you're good to go. I chose 6.35mm so I could plug into my mini amp. I've had some really thick car audio wire hanging around for a while, so I took about a foot of that and slapped my audio probe together. Quick and dirty, but it works.

To use the audio probe I plug it into my mini amplifier, attach the clamp to the (hopefully) grounded enclosure of the circuit and start probing along the circuit board to find the signal path. Don't forget to inject a signal into the pedal circuit first, or there will be nothing to hear. To do this I chose to play music from my cellphone and used a 3.5mm to 6.35mm adapter to connect a guitar cable between the input jack of the circuit and output of my cellphone. This all worked pretty well.

Despite having one more tool in my arsenal to help troubleshoot circuit gremlins, I'm still no closer to figuring out what's wrong with the Blues Driver clone. However, it took me posting on reddit to find out the tremolo circuit was bad from the beginning. I could go on and on about how so called verified layouts can still yield absolute worthless results, but the counter to that usually ends up being that I need to be Nicola Tesla to even touch a circuit. Reddit is a weird fucking place. Once the volume issue on the tremolo was corrected I still need to figure out why it distorts. But that's for another day.

Ugly but functional!

Saturday, May 20, 2023

Knocking Down the To Do List

I have a Windows notepad documenting all my effects pedal builds since day one. I put the name of the circuit built, the date it was built and alongside them I put anything that needs to be checked, fixed, adjusted, troubleshot, etc. I also add an asterisk for pedals that still need an enclosure. Some of my pedals, including the LPB-1 clone, which was my second build, still to this day needed attention. So, I decided I would start knocking things off the list of repairs and adjustments that I needed to do.

The LPB-1 clone used an old A100k pot that worked just fine, until I closed up the build and let it sit. Afterwards, every time an ant would sneeze within a 500km radius when the pedal was on it would crackle. So I decided it was time to clean the pot, so I cleaned it with rubbing alcohol. That fixed it, for now. Should any issues arise from having used rubbing alcohol I'll just replace the pot entirely.

Next I moved on to my 8-bit fuzz, which is a Shoe Pixel clone, and that too needed a pot cleaning. After doing so I managed to clean up 90% of the pot, but it's still crusty and crackly on the furthest top range. That's good enough for now, but this pot will most likely be replaced in the future, should it become more problematic.

Finally I decided to tackle the oscillation of my Rat clone, Rata Blanca. This pedal had been nothing but trouble since I built it, back in March of 2022. I socketed the op-amp to test what sounded best, and also to put in a LM308, should I find one for less than the price of a whole Klon Centaur. For now it's an OP07 Rat and it sounds ok, it just loves to squeal though. I thought I had tested everything and it really was driving me insane, so I packed it away for a few months. When I hooked it up to fix it the oscillation was gone, except in bypass, furthering my feeling of being driven insane. So I decided to tackle the issue of the bypass oscillation by grounding the footswitch and attaching the input to ground when bypassed. This cured the bypass oscillation, but then I noticed the oscillation was back when the pedal was turned on, but it went away if I touched the enclosure. Simple! Ground the enclosure! So I did, and that cured that. Rata Blanca is now finished.

I still have some things to figure out and fix, but I've knocked all the 2020 and 2021 builds off my repairs list. I only have four repairs necessary for my 2022 list, considering I built fourteen pedals last year that's not too bad. The biggest problematic circuits still remain the Blues Driver clone and this year's tremolo circuit, the latter of which actually scored bit of success. It was confirmed the layout I used had an error in it that I'm just not smart enough to have recognized, and the volume issue was rectified in full. However, now the tremolo is an overdrive tremolo. I'll keep plowing through and see what fixes it though.

Thursday, May 11, 2023

The Yay and Meh of the Last Few Weeks.

Since we last spoke I've added three more builds to the completed bag. The completed bag is one of Tayda's huge pink bags where I keep the completed circuits in smaller bags, while they await their enclosures. Each circuit's enclosure has already been picked out, it's just a matter of purchasing them. I also have a bag of knobs already setup for each circuit once those enclosures are purchased, drilled and ready to become each circuit's new permanent home. The three newest completed builds are a Box of Hall reverb, an EQD Chrysalis clone, and a JHS/EHX Lizard Queen clone. I'm still working on the tremolo circuit though. It will be done sometime -- sometime!

The Box of Hall comes with the Heaven Within mods and feedback control. This one still requires the Belton reverb module, but I figured I should build at least one of the two reverbs on this years list so that I have something to test the Belton modules in when I get them. I really enjoyed the Rub-a-dub reverb I built last year, even though the reverb is a bit ripply. I feel the more reverb choices I have the better, and they can always be stacked to see how good that might sound. The other reverb I plan to build this year is an EQD Ghost Echo clone.

Speaking of EQD, the EQD Chrysalis is fucking awesome! I listened to a few demos and it sounds like a 1970s op-amp overdrive, but it's actually a very low component count overdrive that uses two transistors to get its sound. Even through my Kalamazoo Model 1 this pedal sounds amazing. It sounded good through my little test amp rig, but once it hit real tubes the pedal, and the amp, came alive. I'm shocked at how good it sounds, when I'm so much more familiar with transistors producing a fuzz style effect.

Speaking of fuzz style effects, I built a Lizard Queen clone. I hate it. I've made it well known that I'm not a big fan of fuzz, and I'm actually even less a fan of octave. Every octave I've heard sounds like someone farting in an attempt to cover up the original note, and failing completely as you can still hear the original note. Now, 50% of the two people who read this blog will be asking their screens "Why did you build it if you weren't going to like it?", and to that I would say, because I could. Every component for each of this year's builds were bagged up, tagged and sorted many, many months before I even got started. With the Lizard Queen being such a new product, I decided to use one of the many layouts and try my hand at it. Is it a waste of components? Nope. I'll just put it in an enclosure and let it be what it is.

As of the posting of this entry I have completed eleven builds this year, putting me three builds away from each of the past two years total builds, and bringing my overall total up to forty-five. There have been a lot of things learned, and I still have a lot of things to learn, but it's still a fun process. Aside from the tremolo, a Boss Blues Driver clone from last year is also being a bit of a pain in the ass. I've gone over it a few times and it's still not putting out much volume, kind of like the tremolo circuit. I felt this might be the right time for me to build a sound probe, so I can probe around and see where the volume drops. That should help diagnose and troubleshoot some issues. No, I don't consider the sound probe one of this years builds, as it's a utility TO build pedals. Hopefully this helps get a few things sorted and I can put these two pains into the completed builds bag!

Thursday, April 20, 2023

All The - Small Things. Odds And Ends.

Pedal building 2023 is off to a roaring start, but I must pace myself. If I fail to pace myself I will be out of things to build within a few months. Oh, that doesn't mean this year will have fewer builds than the previous years, no, in fact this year actually has more. And what's more is even though I've already prepackaged my builds for this years, I still plan to put more together. When you take into account that I built fourteen circuits each in the previous two years, and this year I've already built eight, with nine prepackaged circuits still to be built, this year is going to be the biggest yet. Hopefully that means I'll have more time to write songs with these pedals once they're all built.

This year does feel like it's going smoother than the previous years. I don't know whether that's due to experience, or just because I'm really excited to get this year's circuits built. That being said, I've also started experimenting with smaller things, such as BMP tone stacks and blend circuits. The blend circuit I put into my tremolo build turned out really nice, so now I'm wondering what else needs one. In anticipation I went ahead and built a few more that are awaiting their destiny.

As far as full circuit builds since my tremolo I've built a Colorsound inductorless wah pedal clone, which is actually really awesome. I'm having a hard time figuring out a way to make it into a function wah pedal but the circuit sounds really good. Then I built an orange squeezer clone that I'm really liking. Since I realized my Really Cheap Compressor pedal actually does work, I've been going compressor build crazy, which I'll talk more about in a second. Then I built a crunch box deluxe, which is a clone of the MI Audio Crunch Box, but with an external presence control. That is an amazing distortion. But if you really want to hear amazing, you need to try the Kingtone Blues Power. After just one demo video I decided I had to build a clone of it, so I did and I'm so impressed with how great this pedal sounds. The Blues Power clone may just become my main drive on my pedalboard.

Now, with both the Really Cheap compressor and the Orange Squeezer clone you would think I should have enough compression, but you would be wrong. Ideally I want to own a Keeley Compressor Plus, but since I can't, I would love to build one, but since I can't, I'll build something and make it as close as I possibly can. I already have optical, I already have FET and now I need to build The Engineer's Thumb! That's right, the closest circuit, in terms of control layout, I could find is The Engineer's Thumb. It may not be anywhere near the Keeley Compressor Plus, but it's something I want to build anyway. Most of this build is prepackaged and ready to go, but it still needs a few parts.

On the docket is also a pair of reverbs, which I'm excited to get done, but they also still need parts. Aside from one more wah pedal circuit everything else is a distortion pedal build this year. I wouldn't mind building a long delay, if there is a circuit layout out there for one. I built a Disaster Transport Jr. clone last year and it's great, but I'm looking for something with a longer delay so I can stack them. The unspoken builds consist of something I'm pretty sure won't even work, but since I have a SIP IC that I believe might work, it's worth a try. I'm trying to build more modulation circuits, but the reality is drive pedals are just so fun to build! We'll just have to see what sounds cross my path and make me want to build them.

Monday, April 10, 2023

Journey or Destination? ¿Por QuĂ© No Los Dos?

In my last entry I talked about my preference to keep a pedal circuit build as close to the circuit it's cloning as possible, and avoiding customizing it to my person desires. But what happens when a circuit doesn't turn out anywhere near what I expected, or even what anyone expected? Well, that's where the journey begins! As 2023 is a year of experimentation and broadening my horizons by building things that I wouldn't normally build, I attempted a simple tremolo circuit. Much like fuzz I'm not a big fan of tremolo, but I do love the sound of it on those 60s classics, so I figured if I could add one to my arsenal I may as well. The build layout was chosen, and after it was completed the resulting circuit turned out far too quiet to be of any use.

This seems to be a common issue for this circuit. I won't blame the layout because it could be based on a bad schematic, I don't know. The standard troubleshooting was done, going over it part by part to make sure everything was the right value in the right place and after a dozen or more times it seems I built it correctly, the layout just doesn't work as presented. I removed both power and all of the transistors and found the circuit still passes signal, that doesn't seem right, so I know something is wrong. Regardless, I went a step further and experimented with a few things to see if I could get it to at least become something I could actually use.

I started by fiddling around with the four required transistors, and swapped a few out for lower gain transistors. In doing so I found which one of them was pulling the signal down, and removing it helped with the overall volume loss, still not anywhere near unity, but it was better. With three of the four required transistors the circuit was working pretty well, but it still wasn't as loud as it needed to be. Then a plan formulated in my head. Many years ago I built an LPB-1 clone with a BMP tone stack, completely misunderstanding that the boost circuit was the recover stage for the tone stack, and not in fact an LPB-1 with a tone stack, so I always considered it a failure. I removed the BMP tone stack, tacked the LPB-1 circuitry onto the backend of the tremolo circuit and voila!

Now that the circuit was not only amplified, but actually can act as a bit of a boost, the effect felt a bit too thick. What I mean is even though the depth control works, there is no clean signal getting through. I know that's just how it's supposed to work, but when everything goes through the tremolo effect, it just feels too thick to me. I did a little bit of research and found a tiny blend circuit, which worked out perfectly. I can blend in clean with the tremolo effect and it actually sounds a bit like a poor man's uni-vibe with a 50/50 mix. It's still not perfect, but I feel this is now much more useable than it originally was.

Again, I won't blame the layouts or their creators, because maybe it was a bad schematic. I've noticed this issue being reported for a few different layouts of this tremolo circuit. I've gone over mine time and time again and it's exactly the same as the layout, so who knows. I'm leaving everything as is with the circuit, plus the tacked on boost and mix boards, just in case a solution comes up in the future and it's an easy fix. But really, with all the fun I've had building and experimenting to make this tremolo circuit work, I'm not sure I would want to take it apart to fix it later down the road. I'm not even sure how often, or if I'll ever use it much at all, but the destination of where this circuit is now, after all the experimentation, I can say I'm pretty happy with what it became.

Wednesday, March 29, 2023

Tonal Accuracy or Personal Preference?

An important part of picking another effects pedal to build, for me, is always audio research. I listen to online reviews of the genuine pedal over and over again so I get an idea of what it should sound like once I've built my clone circuit. Most of the time a pedal circuit sounds close enough to what I was aiming for, but when a pedal build doesn't turn out the way I was hoping, I turn to the layout website comments to seek answers. My latest build came about because I was super impressed with the Silver Jubilee model in my brother's Marshall Code 100H he let me borrow. I know the Marshall Code has to go back eventually, but I still want to have that sound available, so I figured building a pedal to sound like that would be my best option.

The circuit in question is the Lovepedal Jubilee, and the issue I'm having seems to be a widespread complaint for not just this particular layout, but every layout cloning the circuit. On initial testing the pedal sounded fantastic, but I fear that was clouded in the fog of excitement that rolls through my mind when a pedal build works the first time. After playing it more, and testing it with various guitars, I noticed there is a muddiness that detracts from the overall clarity that I don't remember being a part of the reviews I had seen. This issue was exacerbated by the fact I built the pedal many months after having listened to any of the reviews, therefore voiding any recollection of whether this was how the genuine pedal sounded or not. 

At this point I had a few choices, I could either leave the pedal as it was and just deal with it, or I could do a handful of suggested mods to make the pedal, what most people who have built it seem to find, more useful. I actually chose to revisit the audio research stage and see if the genuine pedal sounded like my clone. Now these are youtube videos of people playing the real pedal, and what I do is wear one headphone while playing my homemade circuit set to the same settings they use, to see what differences I can detect. Clearly this isn't 100% scientific, but it's a good enough reference. It only took one video for me to detect the muddiness that I was hearing in my circuit from the genuine pedal. So what do I do now?

Well I continued my audio research, and I also continued to hear the same muddiness from the genuine pedal in other reviews. Then I found a video from Anderton's, from six years ago, where they were testing the Lovepedal JTM and Jubilee. During the review Mick is playing a Fender Custom Shop Michael Landau 1963 Signature Strat and declares: "A bit more bottom coming out of this strat, which is weird, or maybe it's just a bit muddier.", while playing the Lovepedal JTM, but when the pedal is turned off the strat rings like a bell. Again, this isn't exact science, but my assumption is that, to an extent, Lovepedal's amps in a box style pedals may just be inherently muddy. The same occurred in another review comparing the Purple Plexi to the Jubilee, where both of these pedals lacked a bit of clarity, but oddly enough the Jubilee was less muddy than the Purple Plexi.

For a final time I will admit this is not scientific, this is not written in stone, this is just the way I personally test my pedals alongside pedals I do not own and can not A/B in person. So let's assume I am right, and the genuine Lovepedal Jubilee is muddy and lacking a bit of clarity, what should I do with the clone circuit I've built? Should I leave it as is to emulate the genuine pedal, or should I modify it to what everyone else who has built one seems to think is a more pleasing version? Well, when I decided to build a pedal after the audio research stage, that's really the ideal result I'm after, for better or worse. The genuine pedal isn't bad, nor is the clone circuit I've built, it simply is what it is.

Now, let's assume I'm wrong and the compression of youtube, my headphones, and the lack of frequency reception within my old ears are giving me a false sense of what this pedal should be. If I were able to A/B them in person and find out my clone sounds like a speaker in a lake compared to the genuine pedal, what would I do then? Well, maybe then I would consider the mods to see where that might get me, but I doubt it would be that drastic of a change from review video to A/B them in person. I could be wrong about that as well.

For myself, the point is I want what I hear from the review videos, and that's what I tend to aim for. Naturally a lot of my pedals are built using liberal substitutions as I don't always have the exact part on hand to build that circuit, so sometimes good enough is truly good enough. I know I've tested many different transistors in different pedals, to give myself a more pleasing sound, but generally they still tend to sound around about the same as what I heard in the audio reviews I used to inspire the build in the first place. I think the final test will be testing this pedal against the Marshall Code's Silver Jubilee to see if I can match the pedal with the amp. I might be completely wrong on all fronts, but if I can still dial in something that sounds good using the pedal as it is, that's all that matters. I aimed to build what I heard from the handful of reviews I watched, and having rewatched them now and compared it to what I built I believe I've achieved what I set out to do.


After testing the pedal against the Marshall Code, granted this is modelling vs pedal, they don't sound exactly the same, but I could dial in them close enough and find what I like. The pedal has more treble and is still a bit more muddy than the amp, but I still believe that may just be the nature of the pedal circuit. If I could stage one complaint it would be the noise that comes from the pedal circuit, but it is high gain. Overall I think if I were to fix or modify this pedal in any way it would be to reduce the background noise. Final verdict: The circuit, although a bit muddy, is still very useable and sounds fantastic, to me at least. My build will stay the way it was built.

Monday, March 20, 2023

Rockin the Mockman!

Way back in September 2021 I built a RunOffGroove Mockman. The circuit is based off the distortion found in the Rockman amp line, from Tom Scholz Research and Development Inc. Once it was built it sounded amazing, but I still wasn't completely happy with the circuit. Firstly it was way too loud, of course it could be turned down but it felt odd that the volume was on five or six and my ears were bleeding. Second, there was no way to control the gain, which really isn't the point of the circuit, but still it would be a nice option to have. Finally there was no tonal adjustment at all, leaving you to shape everything else around the Mockman. Again, the circuit sounded great as it was, but I felt it could be improved.

The circuit sat in my done, but no clue what to do with box, and from time to time I would ponder what its future would be. Eventually I decided to search around to see if anyone had added a gain control to the circuit, and to my surprise there seemed to be only one person, besides myself, who wanted to do so. A person by the name of Elijah-Baley said they adjusted a resistor value, replaced a jumper with a potentiometer and created a working gain control. I decided to try this out and see how well it would go. As suggested by Elijah I replaced the furthest left 1Meg resistor with a 22k-47k resistor, I split the middle (kind of) and used a 33k. Then I replaced the jumper that is left of that with a 1Meg potentiometer, I used a C1Meg on mine. And to their credit, it works! It doesn't clean up the pedal entirely, but if you need a clean signal just turn the pedal off. As far as the gain control goes, it does work.

Stock layout from

Gain Mod
(Wired for Reverse Log)

Now that I had a working gain control I felt I'll take it one step further (in for a penny!) and try to get more tonal adjustment out of it. A well known mystery of the ROG Mockman is the bass switch, which really only seems to serve as some kind of placebo. I mean, if you flip a switch you expect something to happen, so your brain does something to fulfil that request, but in reality whether the switch does anything is still up for debate. Still being new to building and modifying circuits I looked for the easiest way to add tonal control, and the answer came in the form of a BMP tone stack. The BMP tone stack is named such as it is the tone stack used within Big Muff Pi pedals to give some form of tonal control.

Stock BMP Tone Stack

The BMP tone stack is super simple, highly customizable and doesn't require very many components at all. I downloaded the tone stack calculator and fiddled with the components until I found what I thought would be a pleasing tone stack. In the end I used two 22k resistors for R1 and R2, and upped C1 to a 47nf capacitor. This gave me a bit of a scoop but also allowed me to adjust for single coils and humbuckers. Now I'm not saying this is the definitive tone stack for this circuit, it's just what worked best to my ears. Many of you familiar with BMP tone stacks will also notice I used the passive version, which helped cut down the extra volume from the Mockman perfectly! You know what else adding the tone stack did? It made the bass switch noticeable. Before there honestly seemed to be no change whatsoever, but now there is a noticeable bit of bass/bass cut, depending on which position the switch is in.

Now that my Mockman has a working gain control, tone control and, while still loud, it no longer makes my ears bleed, this circuit is much more versatile and useable than I felt it was stock. These mods might not be for everyone, and that's fine, but I figured that someone else out there might enjoy tinkering with theirs, and these are a few good mods to start with. Both mods are highly adjustable, super simple and add much more versatility to the circuit. I give all the credit of the gain control to Elijah-Baley (if there was a way to link to them I would) and the BMP tone stack to Beavis Audio. I hope this starts someone on their own journey in modifying not just the Mockman, but other circuits as well, the same way it did for me. Please use these mods, adjust these mods and perfect these mods, and report back as to how you adjusted them to your own tastes. My chosen values might not be the best, and I'm very interested in seeing what someone else comes up with.

Wednesday, March 15, 2023

Felt Like Makin' Fuzz!

In part one I talked entirely about the Fuzz Face clone that I built, but I feel the other fuzz pedals I've built also deserve a little more attention. So in this edition I'll be displaying each fuzz pedal I've made and talking a little bit more in-depth about it. What better place to start than with the very first pedal I ever built? Aside from starting with the first fuzz pedal I ever built, they will be presented in no particular order.

This is my bazz fuss circuit that I've dubbed the Fuzz Throat. I had wanted to build DIY pedals ever since the mid 2000s but I lacked confidence and never took charge in making that dream come true until January 2020. When I started looking around for easy pedal circuits the Bazz Fuss was the most talked about beginner build, so I dug through my parts bin and found what I thought I needed. Eventually I found out I was using a transistor with the wrong pinout, and kept testing transistors until one finally worked. The first transistor that worked is the transistor I stuck with.

The Fuzzy Throat is a gritty (and dare I even say shitty) gating fuzz because the transistor was just what I happened to have on hand, and I never dared to bias it correctly. This build stays as it was built the first time it functioned (at least semi-) properly. With this it's either all go or no go, as it will gate out completely with even the slightest adjustment of either the guitar or the pedal's volume control. Packed inside a Sucrets tin with a yellow knob to simulate the lozenge this little bastard started it all. 

This is my EHX op-amp Muff Fuzz clone. I did discuss this one previously, so there isn't much left to say, but I felt it still needed to be included in this entry for the sake of completion. I built this circuit in September of 2021 and never really knew what to do with it, until I started formulating a plan to keep it much like the original EHX Muff Fuzz and cram it into a 1590a enclosure. Once it was all done I can say it's a pretty neat little fuzz that sounds pretty good.

Besides guitars and guitar equipment, something else near to my heart are vintage video games. When I first heard the Shoe Pixel fuzz I knew I had to find a way to build one for myself, and here it is. The 8-bit tones that can be dialed in with this thing are incredible, not to forget there is an absolutely useable musical fuzz that is hidden in there as well. The overall design choice is meant to evoke the totally radical and carefree childhood of the late 80s and early 90s with its vibrant colors. The input is designated green and the output designated red, with a red power input on the very front of the pedal. The LED came from a McDonald's Super Mario toy (a little too on the nose?) that constantly shifts through a handful of colors when engaged. This is the one pedal that makes me feel happy just by looking at it.

Finally we have reached Purple Face, my Fuzz Face clone. I was originally going to add some form of triangular material by the footswitch, in the same vein as a real Fuzz Face, but I decided against that. After I finally heard this circuit's true voice, through Kali, I put it into a purple enclosure and finished it off with antique bakelite knobs that I plan on putting googly eyes onto. Again, I'm not sure whether this sounds anything like a true germanium Fuzz Face, but I get a really nice range of tones from it. Fiddling with the knobs I can get anything from an almost overdrive crunch to a thick fuzz. Don't tell the others but I think this one is the most versatile, and probably my favorite, fuzz pedal. But only by an exceedingly slim margin over my Shoe Pixel clone.

So will my building schedule and a fuzz circuit ever crossed paths again? YES! Pardon me for shouting at you there, but yes, yes it will. You see, a few years ago That Pedal Show and Josh Scott went to Macari's in London to test Tonebenders, and in doing so they made me fall in love with the MKI. One of them can even be quoted as saying "That sounds like modern gain to me.", which I completely agree with. To my ears, so take that for what it's worth, the MKI has a very fuzzy edge, but sounds very much like a modern high gain circuit. That type of characteristic, whether right or wrong, is almost like what I can dial in with Purple Face, albeit Purple Face is more like an overdrive pushing a 5 watt tube amp to her screaming point, but DAMN does it sound amazing. So, a MKI tonebender circuit will be on my build schedule at some point, even though I've read many times it's a real beast to dial in correctly. I feel I'm up for the challenge! At least I feel that way currently, we'll see how I feel as I continuously fail to get it right when that time comes.