Home GuitarPCB Forum Forum Rules MUFF'n Build Essay w/ photos using most mods

  • This topic has 8 replies, 5 voices, and was last updated 3 years ago by Anonymous.
Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #548
    Barry
    Keymaster

    The journal of the Klingon Muff . . .

    or . . . How I Built & Modded a Big Muff Pi Based on Barry’sMuff’N

    Reposted by Barry courtesy of Cybercow

    ***WARNING*** (long read – but there is circuit porn towards the bottom)

     

    After a long bit of planning, analysing and researching what a “Big Muff” is, I finally got aroun to completing my rendition of a BMP based on Barry’s “Muff’N” PCB. Barry’s build docs for Muff’N intriqued me and prompted me to dig into an analysis of each section, including Barry’s Muff’N Advanced Builder’s Guide.

     

    This is mostly a long read for beginners. You building\modding veterans can skip this if you like. I think you’ve all visited the following processes at one point or another and either embraced or rejected them in your own fashion.

     

    If you do elect to read this, please take a moment to refresh your favorite beverage, relax and read on. I suggest you do all the reading before going through the photos. The photos are in order of the discussion but not referenced.

     

    It started with a Guitar PCB “Muff’N” PCB, cuz I wanted to see what a Muff build was like. I wanted a non-IC Muff build for the specific nature of transistors over opamps. Then I read about all the mods available for this particular PCB. So I dug around and found not only several mods directly associated with Barry’s PCB, but all the variants of Muffs and a couple different “analysis” papers about the fundamental Muff circuitry and a plethora of other mods, studies, alterations, quirks and brag-brags.

     

    I saw that the default Muff’N build provisions two different diode clipping options with a simple SPDT (On-Off-On) switch. “Neat!” I thought to myself, as I remembered Guitar PCB’s 4-way rotary switch PCB called the “Roto-Tone”. Then I noticed the ‘Mid-scoop” pot mod; and the Lo and Hi pass switch mods; the Bender mode switch mods; the Pi filter selection switch mod – and thought once again to myself: “It would be pretty neat to have a Muff that can reach into and touch bits of all 8 of the factory variants out there; with a couple of extra clipping diode options.” That’s when it occurred to me that an enclosure larger than a 1590B or 125B box would be required. Something deep inside was whispering for this build to have a “rough” appearance. So I found a bare aluminum 1590BB on Hammond for $5.40 and some nice grey anodized aluminum knobs for $2.20 each.

     

    The PCB population process went smoothly, ‘cept I had to “tent” a few resistors to come up with required values. I din’t select any specific model build from the BOM that came with the Muff’N PCB, rather, I selected averages, I wanted in my build to swing more towards the middle of the ranges for each particular resistor and cap – constantly referring back-and-forth between the original schematic, the BOM and all the other schematics I could find and print for the various versions of the Muff that are out there. I also elected to hold off on just dropping the default clipping diodes onto the PCB because I knew I wanted to add the option of the ‘Roto-Tone’ PCB to add more clipping diode options.

     

    For clipping diode switching, I went withe the stock SPDT (On-Off-On) to switch between the on-board set of my choosing and the ‘Roto-Tone’ with four sets of clipping diodes. Before just grabbing diodes and populating that board, I did some homework and came up with combinations not normally used in pedals. The diode clipping switch in the left position would engage the on-board clipping diodes. The center position would disengage ANY clipping diodes. And the right position of the clipping diode switch would engage the ‘Roto-Tone’ PCB packed with my special diode combo-sets. Those sets are:

     

    1) Bat46 (x2) & OA91 (x3) – Ge diodes internally throwing the symmetry off with matched Vf on each side)

    2) 1N914 & 1N5817 – Si (the Schottky will ideally provide an asymmetrical Vf slope with the 1N914)

    3) Blue & Red LEDs – Si (these are water clear colored LEDS and a slightly higher Vf – again, asymmetrical due to different Vf drops of each color)

    4) 2N5088\2N5087 & Clear White LED (a PNP/NPN transistor pair with their B’s & C’s tied together with a single water clear white LED)

     

    In summary of the clipping diodes selection(s), with both the Clip and Muff Bender mod switches, all clipping diodes can be bypassed completely. Perhaps on a future build, I’ll use another ‘Roto-Tone’ circuit on the second clipping diode stage as well.

     

    The basic Muff’N build also has options for Lo-Pass and Hi-Pass capacitor selector switches and I opted to go for both mods, but chose to use the DPDT (On-Off-On) switches so each could utilize a less/default/more option.

     

    While a deep “mid-scoop” has its place, I also wanted to be able to fill that trench and added the ‘Mid Scoop Control’ mod by adding a linear 25K pot to the user interface.

     

    Finally, the Pi filter had so many options with regard to establishing the cross-over points at which the lo and hi cuts, I went with the first Big Muff Pi fundamental circuitry and added the DPDT (On-Off-On) switch between bypassing the Pi filter altogether, engaging the Pi filter, or just running it through a 150K resistor.

     

    OK, so now all the electronic prerequisites are established, planned and in the works. Now, onto laying out the hardware within the enclosure. With 4 pots, 5 toggle switches, a rotary switch, the stomp and LED indicator, the 1590BB has enough room, even with all the wires. The biggest concern was getting the drilling done right, even if only somewhat aligned, neat and pleasing to the eye. That was just meticulous planning with pencil and paper using actual-size cut-outs of the PCBs, pots and switches, (They were just footprint outlines for spacing and layout inside the enclosure.) Once I had a workable layout of to-scale cut-outs laid down inside of the enclosure, I then went to Photoshop and re-created everything there in a to-scale drawing of the enclose and all the components that I wanted to fit inside. Doing it again Photoshop allowed me to create accurate drill points for the enclosure and get a better idea what the resultant guts layout would like. Sometimes, I will take photos of the populated PCB’s, pull those photos into Photoshop, scale them to life-size, then place those within the enclosure diagram.

     

    Meanwhile, there was also planning for the graphics on the front of the enclosure. Trying to decide if wanted to do a full-face decal, or just cut-out labels for each piece of control hardware. Considering the pedal was no longer just a Muff’N, or populated as a Ram’s Head, Triangle, Civil War, 3rd Edition, Foxy Lady, or Green/Black Russian, I thought to make it rough, dark, dank and dystopian in a Klingon sort of way. (I don’t actually read Klingon, so I used a font that only somewhat emulates Klingon glyphs.) I went with the full-sized waterslide instead of cut-out labels.

     

    I used several iterations of Photoshop edits to get it right, and once I started getting close to the final waterslide design, I started printing it out on plain paper, cutting it out an placing it on the enclosure’s face and hold it up to the light to see thru the drilled holes and determine how far off the labeling was. It’s a process I kept repeating until it got very close. The I would tape the plain paper print-out, cut it out, tape it to the enclosure’s face, then mount the hardware with the knobs and switches so I could dial-in the labeling line up. Another rinse and repeat process.

     

    With the holes drilled, the main and daughter PCB’s populated, I then set about populating the enclosure with all the components – PCB’s, pots, switches, and jacks. This let me then plan out my wiring scheme. I knew there would be lots of wires to run, so I wanted all the components in place within the enclose so I could plan, layout, measure and cut all the wiring runs. This is quite tedious as it took several removals and replacements of the various components in the enclosure to get the wire lengths laid and cut correctly. So as I got wires cut, I’d then rehearse the soldering order and process. The rehearsal was important as I found out had I not done so, I’d have had to unsolder some things in order to solder something else first. I like to use “outtie” DC jacks so I can take the entire guts in and out without the need to unsolder anything.

     

    So once all the components are soldered together, I thought: “What the heck. It’s all ready to be fired up. Let’s give it a go.” Usually, in pedal building, it is best to “rock it before you box it”. But I was already in the box and knew the guts had to come out again anyway to finish the waterslide portion of the enclosure. So I fired it up. I found three problems. The Mids control was wired backwards. The Hi-Pass switching option did not work. And the Pi switch only worked in one of it’s three positions. Even with all of that planning and research, there were problems. The meds pot was an easy fix. and it turned out that the Hi-Pass switch was crap from eBay – so I swapped with a better quality switch and it worked. But the Pi 3-way switch really befuddled me. It seems to working just fine when checking with an ohmmeter. But in-circuit, it just wouldn’t engage the other two positions for the Pi\Mids\Direct options. I decided to just swap out the switch and voila! It worked as planned. So, now, out of the box, the guts worked and I needed then to get on with finishing the enclosure.

     

    Now, just because the holes were drilled in the right locations and waterslide artwork properly reflected that point, I knew there might be issues with label placement once all the hardware’s washers were in place and knobs applied. So to better approximate the labeling placement in the artwork, I reprinted the artwork onto plain paper; cut-out and applied the mock waterslide onto the top of the enclosure and cut out the drill holes; then re-stuffed the guts into the enclosure and applied the knobs to see how well things lined up or not. Well, they did not the first time. So I took a pencil and ruler and made notes of what was off by how much, where and in what direction any corrections needed to be made. I removed the guts one last time and removed the mock-waterslide from the enclosure. (I just used small bits of scotch tape to hold it in place.) Then back to Photoshop with the marked-up mock-waterslide that has the corrections for the graphics layout. This would be the last adjustment to the waterslide artwork.

     

    Once I was as sure about the waterslide graphics as I could be, and measured and tested all that I could, I printed out the waterslide using clear-background waterslide inkjet paper. After the printer passed the printer paper, I let it dry for two hours before applying two, light coats of clear Krylon spray with two hours of dry time between those two light coats of clear.So, while that second coat dried, I went on to the final enclosure prep.

     

    Being bare, unpolished aluminum, the box was a bit rough, so I took 400 grit sandpaper to it first. After that, I used 1200 grit. Then finally, after a good rinse & dry, one of those large, brick-red erasers. Finally, before applying the waterslide, I wiped it clean with an alcohol pad. Just rubbing alcohol with paper towel will work. After applying, drying, clear-coating the waterslide decal and allowing drying time again, I put the final touch on the enclosure by applying several thin coats of MinWax furniture polish on it. This greatly enhanced the protection of the waterslide and greatly reduced the ‘roughness’ of the unfinished aluminums surface without looking too ‘pretty’.

     

    OK now. I had a finished enclosure, a working set of Klingon Muff guts and commenced to stuff the finished enclosure and apply the knobs. There was no more required drying time or further testing. This thing was boxed, already tested and ready to rock. So I took to my amp room, plugged it in, saw the red light of “it has power”, tested that the amp was hot with a bypassed signal from the Klingon Muff, stepped on it’s stomp switch and played the first chord . . . . nothing. No sound in activation mode. I hit the switch again – bypass: OK, engaged it again: dead. Harrumph! This is how building\modding can go. You spend all kinds of time planning, researching, prepping, testing, building and the result is zilch. So I opened it back up and broke out the audio probe to see if I could quickly suss out the problem.

     

    I was able to audio trace the signal up the base of the last transistor, after that the signal was like -96db. I swapped out the transistor – no change. I kept a signal going through it with a small MP3 player and and began to pick up the nest of guts when suddenly it kicked in – and loud too!. I set it back down and it went quiet again. Hmmm . . . . I took a small plastic straw, so the capacitance of my skin would not influence anything. I just wanted a physical tool to poke and prod the solder joints around that last transistor and it needed o be something non-conductive. I touched and tried to wiggle every single solder connection around Q4 and as I was about to give up, I repeated the poke-n-prod test on every component near Q4. THERE! There it was. The collector output resistor moved a bit when I pushed straight down on it and it got load again for a split second. It would not wiggle at all from side to side as I had been prodding everything else, but when I pushed straight down, it moved and engaged ever so briefly. I unplugged the power and ins & outs and took my 10x magnifying loop to examine that solder joint more carefully. There was solder on BOTH sides of the PCB – but on one side, there a gap between the solder and the PCB pad. I quickly took some fresh flux to solder joint and reflowed the joint with just a hint of fresh solder. Visual re-examination looked good. Plugged it all back in and hit the stomp switch. BAM! That was the problem. Unit now worked. And wow did it work. The diode selections rotary switch for the first clipping section did its magic, the “Bender” mode switch removed the clipping diodes from th second clipping section; the Pi Filter On\Bypass worked, and with the Pi off, the Mids control would flatten out that BMP scoop; the Hi-P filter switch worked in all three positions, all the pots worked as expected, but when I got the Lo-P filter switch – only one position worked – center. I checked the switch. I checked the caps. I checked the wiring. Everything tested for continuity and the caps tested within the ratings of their respective values as they should.

     

    Confounded by the issue, I referred back to the schematic and build notes and BMP analysis dissertations. I was ready to consult a crystal ball. But then I saw it in one of the analysis docs – “caps with too high of a value in the Lo-Pass filter will cause a sever volume drop”. I double-checked the value of the caps and realized that I had applied resistor math to adding two caps together. In resistance math, when two resistors are places in parallel, the total resistance is always reduced to ‘less than the value of the lowest value component in the parallel pair\set. (You can find TOTAL RESISTANCE in a parallel circuit with the following formula: 1/Rt = 1/R1 + 1/R2 + 1/R3 + … ) But that’s not true for capacitors. For caps in parallel, it’s 1/Ct = 1/C1 + 1/C2 + 1/C3 + … so I was drastically increasing the cap values of the Lo-Pass filter switch circuit instead of decreasing them with Lo-P switch mod. I re-examined the circuit and elected to go with much smaller values for the Lo-P switch. So now, instead of adding 220nF or 100nF to the 470nF Lo-P cap, I changed the switch cap values to 10nF and 22nF values for a more conservative approach, making only small tweaks in the Lo-P filter switch mod. And that did the trick.

     

    OK. it was all complete and so I put on my rubber gloves for a final test before re-stuffing the guts into the enclosure. With everything plugged in and turned on and running audio through the activated Klingon Muff, I carefully picked it up and delicately passed it back-and-forth while rolling th guts nest over in my hands. There were no glitches, errors, popping or problems that I could hear. So I stuffed the guts into the finished enclosure; tightened everything down and ensured the switch were lined up and orient properly oriented. Applied and tightened the knobs; closed the back; connected everything – but this time with my guitar; then fired it up. It music to my ears, (pun absolutely intended).

     

    With all the mods on the BMP, it is important to note the the ‘Mids’ control does not do anything when the Pi switch is on. When the Pi switch is off, the Mids control is re-connected to the circuit. It was also interesting to learn how some of the controls seemed to have little to no effect when my guitar’s bridge pup was in humbucker mode as opposed to single-coil mode. Most interesting indeed.

     

    I especially like the “Muff Bender” mode option and I am going to repeat this build, but the next one will have more diodes in series on each side of a diode set so the forward voltage drop is not so small with each selection. I also intent to use a footstomp switch for the “Muff Bender” mod option as I don’t want to bend over to switch that option in\out of use.

     

    So now, here is some eye-candy and circuit-porn. The pics should be self explanatory and may or may not be in order of the discussion. (Except the first pic – it’s the finished product.)

     

    Thank you for taking the time to read this far.

     

     

     

    #1582
    Anonymous
    Inactive

    Wow, great work! I have the same board and almost the same ideas after looking at the filter section of a Radial Tonebone pedal. But I never started it as I was looking to complete my Super Sonic build. This article, rather a short book, has inspired me to break out the board and dig deep into my component stock.

    It’s members like you that keep this place rocking. Thanks so much for all your work. I was wondering though how you apply your graphics sheets so thanks for the idea. Everyone has their way and I have yet to find the perfect way. In fact, it has been the toughest challenge in all my builds. How about the diodes? Did you ever try Mosfets?

    Anyway, thanks again. Keep up the good work and let’s wear those soldering irons right out! -Dave

    #1583
    Anonymous
    Inactive

    ” It was also interesting to learn how some of the controls seemed to have little to no effect when my guitar’s bridge pup was in humbucker mode as opposed to single-coil mode”.

    Possibly an impedance change between the single coil and humbucker? You can get unexpected results in picking your components without running the circuit though a Spice program. Just my .02 ‘sense’.  -Dave

    #15494
    Anonymous
    Inactive

    Hello to all,

    I am new to the past time of building my own effect pedals, and have now purchased a few kits from a supplier here in Australia. I have requested some advice from them regarding the building of the standard Muff’n kit, but as yet haven’t heard back from them about what to do, so am hoping for some guidance from someone here in the forum so I can complete the project.

    This will be a 2 part question, so please be tolerant of this.

    1. I have noted from the build guide pages that R25 (33K resistor) and R27 (3k3 resistor) are omitted from being installed on the main PCB when using the separate 3pdt stomp switch pcb, so do I need to use a wire jumper where the resistors should be or not? These resistors were not included with the kit.

    2) The second question is that when using that separate switch pcb and the Bi-color LED, what value does the resistor need to be that’s shown on the build guide (picture shows it occupy’s the (CLR) set of pads, as that resistor didn’t come with the Switch pcb or was included with the kit either. As it appears from the build guide, would I need to use the 3k3 (R27) resistor on the switch PCB to assist the LED to work?

    I hope these questions are too newbyish for this forum.

    Regards

    Flarts

    #15499
    Cybercow
    Moderator

    Flarts . . . .

    1. When using the separate 3pdt stomp switch PCB, simply OMIT (no jumper) R27. R25 is a separate issue involving the transistors you select to complete the build. For R25 (as stated in the build doc:) – “as noted at the bottom of the parts list, the 2N5133 was lower gain, in fact, the specs state the minimum gain as 60hFE, whereas the 2N5088 has a minimum gain of 300hFE. If you use a 2N5133, R25 should be a jumper. If you use a 2N5088 or other higher gain transistor, you will need to insert a resistor in R25 to keep the pedal from squealing when all knobs are all turned completely up. 33K is a good value to start with. A stock Mayo calls for the 2N5133 and a 2k7 value in R25, but this same guidance applies if you change transistors.”
    2. For the bi-color LED, the CLR value simply determines how brite the LED will shine when on. The lower the CLR value the brighter the LED will be. And yes, the CLR needs to be mounted on the switch PCB to get the LED to work. Personally, I use a 4K7Ω or sometimes a 6K8Ω CLR value to keep the LED brightness down. Those indicator LEDs can be obnoxiously bright.

    Good luck with your build! Cheers!

    #17097
    Anonymous
    Inactive

    Hey Barry,
    my question comes about three years late but perhaps it can still be answered 😉

    I am currently thinking of adding a rotary switch do my DIY DOD 250 preamp. And while doing my homework I came across your post and now I am particular curious about your 4. clipping option, the “2N5088\2N5087 & Clear White LED (a PNP/NPN transistor pair with their B’s & C’s tied together with a single water clear white LED)”.

    This sounds interesting, but I do not completely get the idea. Is it a Sziklai pair with the LED between the PNP B towards the NPN C? Any further explanation on how you connected the two transistors with each other, the LED and the poles of the switch would be great!

    And finally, what kinda clipping sound does this fourth option give? Rather a clean volume boost, more fuzz or something different?

    #17100
    Cybercow
    Moderator

    BullFrog – the 5088\5089 trannies are NOT set up as a Sziklai pair. When I get back to my PC, I’ll post a diagram of how they are wired to utilize just the PN\NP junctions of them for the clipping.

    #17102
    Cybercow
    Moderator

    BullFrog – I found it. The typical Fvd of silicon BJTs is about 0.7v – so together, these two BJTs (PNP & NPN) act like two 0.7v Fvd diodes in series, with a total Fvd of ~1.4v

    The added LED (with a Fvd of ~2.0v) is oriented in the opposite direction to complete an asymmetrical clipping device configuration.

    IMO, using a pair of BJTs in this manner is pointless because there are so many other means of using different combinations of diodes and\or LEDs to achieve the same thing. I just did this one clipping option to see how it works and sounds. I found nothing special about it other than it uses BJTs instead of diodes.

    I just find the whole electronics thing, and the experimenting therewith, quite stimulating to see what works, how and where such implementations are practical.

    #17105
    Anonymous
    Inactive

    Ah, great, thanks for the diagram!

    Yes, I can imagine that the sound is not too different from other clipping diodes while BJTs are more tricky to solder. Anyway, I like the idea to try a combination of an NPN and an PNP transistor, to get a little different taste.

    I’ll see whether it get it to work, just for fun 😉

Viewing 9 posts - 1 through 9 (of 9 total)
  • You must be logged in to reply to this topic.