Another week in the shop

The most important lesson that I was taught by guitars #3 and #4 was that you should do fit before finish: get everything together in rough form first, check it sits together, tweak it until you're happy, and then do the "finishing" work. And where finishing is still half the job: sanding things smooth, dying, oiling, final fret finish, and so forth. These two guitars were made of the course of 2017 as I found time away from my day job, and it was just easier for me to take each part through to completion before starting the next. But this meant I made life harder for myself later on when inevitably I found small issues as I assembled each guitar and they were now much harder to deal with, even down to trivial things like fear of scratching the finish as I tweak the electronics. It's not that you can't pull it off this way, as I'm really happy with these guitars for the most part, but it just makes your life a lot harder than it need be. This theme dominated my week, and I pass it on as a lesson to anyone else starting out making physical objects for the first time.

With this lesson mind, this week I continued to build the prototype amp with a focus on getting the physical aspects through to rough completion to demonstrate the design before I move on to the electronics side of things. The initial render I posted a while ago of the amp looked like this:

Screen Shot 2017-12-12 at 13.51.40.png

So I was quite pleased when I started putting it together and got to a stage where it looked like this:


Not bad for my first full design to CAD to CAM to assembly! You'll see I'm just using MDF here for the body, where as traditionally amps and cabs are made from birch plywood, but that again is just part of prototyping. Birch plywood is quite expensive, so you don't want to go making your mistakes on that if you can avoid it. For a prototype MDF has enough strength for what I want, comes in the right dimensions: 18mm for the body, 12mm for the speaker baffle.

Part of the reason for the modelling in Fusion 360 was not just to generate the panels, but also let me check all the parts fit. I modelled the chassis I bought, the speaker, and transformers, etc. and put them in place in my model so I could check everything fitted. The reason the speaker baffle is 12mm not 18mm is because otherwise I'd be too close to the power transformer for comfort. If you look at the side profile below, with the front of the amp on the left, you can see all the parts in position. The chassis has the controls on it, so has to be flush with the front, and as does the speaker baffle for obvious reasons. The power transformer is mounted on holes pre-drilled into the chassis I bought, so is forced onto the left, and as such you can see the speaker baffle has to be 12mm to make sure there's good clearance from the baffle.


This is why modelling everything, including the bits I'm not going to personally manufacture, in Fusion is great, all the variables like that I can tweak and make sure everything will fit. However, I only modelled the final finished product, I didn't model the assembly stages, and herein was the lesson of the prototype as early as possible: although things may fit together in final position, you can't guarantee you can get to that final position!

If you go back to the diagram, you'll note the chassis is fit quite snuggly between the front and rear panels, which are glued into place. However, with the speaker baffle fitted I can no longer easily slide the chassis up into place, I now have to put it in at an angle so the front goes into the gap between the baffle and the top, and then I rotate the rest of the chassis in. So, dodged a bullet there? No, if you have the power transformer on the chassis (as it will need to be fitted to the chassis before the chassis is fitted to the amp), I can no longer get the front of the chassis into a position where I can rotate it up! So, much like the sofa in Dirk Gently's Holistic Detective Agency, the chassis does fit, it's just impossible to get it into that position :)

Thus the lesson here is not just to think about final position at design stage, but also about the steps of assembly too. For this prototype I shall just move the power transformer by drilling new mount points in the chassis to let it sit further back. In production I'd just need to either deepen the who amp by another 10mm or so to let the chassis be put in cleanly, or have the rear panel screwed in rather than glued in, which is a design you see on many amps.

After all that, I also laser cut a prototype front panel for it (more below) and finished fitting the feet and handle, so it's all looking quite good. Time allowing, I hope to get onto the electronics next week.

Guitar #3 also came back into the shop briefly. This guitar was finished before christmas and has had a bunch of play testing since then, and it came back in for some final fettling.  Firstly, I wasn't happy with the fret ends. Whilst they didn't catch your hand too much, they still felt a little more present that I wanted, so I took another pass at the fret ends and made them lovely and smooth. I also took this opportunity to adjust the truss rod again now that it's been together for a while under tension.


Again, to my opening point, this was made more fiddly by having oiled the neck already and not wanting to damage that finish. But it's now all done, and I can declare success on my first full neck build being finally finished.

I also decided I wasn't happy with the quality of my soldering on this one compared to guitar #4, so I went back and strengthened up some of the joints. Soldering is something I'm continually trying to improve on, and practice really does help improve your skills here.

Finally I did a little work in Adobe Illustrator (not a tool I have had much cause to use before). I generated both the design for the control panel and a small flyer (more on which next week I suspect). 

Ideally I'd have done the panel design in Fusion 360 and used that to generate the laser cutter instructions, but I didn't for two reasons. Firstly, at Makespace the laser cutters are plugged into dedicated PCs, and driving them directly isn't encouraged, thus I'd need to generate vector files rather than tool paths to load into the laser cutter software anyway, so the CAM section of Fusion is no use for me in this case. Secondly, the logo for Electric Flapjack is already in Illustrator, and my brief attempt at importing it into Fusion was frustrating, so I wimped out on grounds of time (something I'll revisit at a later date though).

Screen Shot 2018-01-13 at 10.49.53.png

For those interested, I didn't do the tick marks on the volume and tone controls by hand: I found a website called scale-o-matic designed for making control faces for potentiometers. You give it a bunch of parameters and it'll generate an EPS file which you can easily import into tools like Illustrator. There truly is a website for everything out there.

Having made the design, I then got some laminated perspex and use the laser cutter to both cut the panel out and etch in the labels. The final bit needs a bit of tweaking, but overall as a rough draft fits in place, and I can do a nicer one in the finishing phase!


First week in the shop of 2018

Happy new year! Looking forward to doing lots more in the shop this year, so let's get cracking.

I spent a little time earlier in the week getting guitar #4 from hanging on the wall to being an instrument that can be played. The first new thing for me was mounting the soap bar P90 pickups. For those less familiar with pickup designs, P90s are the design Gibson used in the 1940s until the late 1950s when Gibson employee Seth Lover invented the humbucker pickup, the Gibson mainstay ever since. However, for blues and rock the P90 has remained a popular pickup and is still used today.

The P90 typically comes in one of two of packages (referred to as soap bar and dog eared),  both of which seemed designed to be clamped straight to the guitar body, rather than a more traditional floating mount that lets you adjust the height of the pickup as you adjust things like strings and bridge heights through the guitar's life. The soap bar P90 design, which you can see below, is mounted via screws that go through the middle of the pickup and into the guitar body. The problem I hit was how do you mount the pickups so they are adjustable? Typically you want to mount pickups on springs of some sort, as you adjust the screw holding them a sprint compresses/expands to keep the pickup stable at the new position. But for a P90 you'd have to put the springs over some pilot holes, put the pickup on top without the springs falling over, and then put the screws through the pickups and springs into the pilot holes, at the same time as trying to tuck all the wiring out the way neatly. Theoretically possible, but a right royal pain to actually pull off.


The solution I used in the end was to use several strips of neoprene to create a cushion for the long edges of the bottom of the pickup, leaving a clear space for wiring and the screws. The neoprene will compress nicely as you try and adjust the height by adjusting the screws, and using several thin strips means you can vary the amount of padding readily to suit the maximum height you want. The pickups are now snuggly in place.

The neck of guitar #4, as beautiful as it is with that spalted tamarind fretboard, has been something of a problem. When the fretboard was glued on the maple neck, as the glue set for some reason the entire neck warped into a slight banana shape, leaving the fretboard side concaved. At this point, I could have just scrapped it and started again, but as a learning exercise I went through seeing how much I could fix it, and it never hurts to get more practice at fretting and all the other bits of neck finishing.

The first step I took was when I sanded the radius into the fretboard I sanded it more at either end to level the freboard part somewhat. Between that and truss rod adjustments I could get it mostly level. Having the fretboard roughly level, I then still had the problem that the overall shape is still curved, so the tip of the headstock is higher than it otherwise would be. On it's own this might not have been a problem, but it was compounded by my trying out locking tuners on this neck, where the string isn't wound wound multiple times around the peg, it just is clamped at the top of the peg, so sits quite high; which meant the strings didn't sit firmly in the nut when I strung the guitar up. The solution here was to add a second string tree for the middle strings to pull them down (and having multiple string trees isn't unheard of anyway). The final recovery was that if I just screwed the neck into the body as is, although I'd straighted the middle, the bit where it joined was still slightly curved, so I had to put a small shim of wood at the back of the neck pocket to get the neck angled level.

All in, it's recovered to a point where it's quite playable now, but clearly not an ideal neck, and not something I'd be happy providing to a customer (unless perhaps you're Jack White, who said he likes guitars for their quirks rather than the absense of quirks :). But, I gained a lot of experience going through the process, and I'd encourage others trying to build guitars (or indeed other things) not to look at something going wrong as the end, but rather turn it into an opportunity to experiment and learn. Even if you can't save the thing you're working on, you can still get value out of working with it.

The other main progress this week has been starting to take my amp design out of the virtual world of Fusion 360, and bring it into the physical domain.

Screen Shot 2017-12-14 at 09.45.01.png

I sound like a broken record here, but one of the things I really love about Fusion 360 is the fact it does an entire end to end workflow in a single tool. Having completed the overall design, I then took the components that make the cabinate and created a new layout in the same design where I laid them out flat within an area defined by a bit of material I expect to cut them from (in this case a 1220x610x18 bit of MDF for my first trial run) I can do this using a copy of each panel that is linked to the original, so that if I tweak my amp design later, my production layout will automatically reflect those changes.

Screen Shot 2017-12-18 at 15.06.26.png

Having done the layout (including iteratting it a few times based on trying to minimise material usage and physical contraints of the machines I have access to), I can then create CAM tool paths for a CNC Router, such as the one we have at Makespace, and work out all the cutting actions necessary to actually cut the wood. Again, this is all linked, so if say I go back and change the radius of the rounding of the edges, the tool paths just need a one click to regenerate with the new values. Having done that I could then export the design ready for cutting, which I spent Thursday afternoon doing.


A lot of people assume using a CNC router is a lazy way of doing things, but don't think that because the machine is doing the cutting I can just forget about it. Firstly as this CNC router isn't the most advanced, I need to keep cleaning up dust from the grooves it makes etc., so I spend the better part of the three hour cut run chasing the head with a hoover. I also need to watch for where I've made mistakes in the toolpath generation: at one point in this run I spotted it starting to do the wrong thing and only because I was watching it was I able to abort that stage, generate some new tool paths and set it going again. Had I not been watching it, I'd have had to start again with new material. It's also loud and dusty, so by the end of it all you're quite tired. I don't expect sympathy for this, as it is still easier than sawing by hand, but for those looking at this and thinking it's a trivial task, it is still quite involved.

Having cut the panels and checked they all fit as expected, both with themselves and with the metal chassis, they're now gluing in the workshop, so hopefully a box will be ready next week and I can move on to the electronics.

A New Workbench Amp for the New Year

Not having the luxury of my own workshop where I can leave a regular sized amp, I need a small workbench sized amp I can use to test guitars as I solder them together. Space is at a premium in Makespace, so I can't even just leave a small practice amp there, I have to head down to really tiny desktop amps. Also, given it's a shared workshop where things might get knocked or broken accidentally, I don't really want to go all the way to the small but very good sound quality (and not cheap) amps like a Yamaha THR10 or Vox Adio Air GT.

For the last year I've been using the Orange Micro Crush, which is tiny, fairly cheap at forty quid, and has a built in tuner. However, whilst it's a fun little product, it's clean tones are not great (or, indeed, existant), and as I move into doing more complicated wiring circuits with coil taps and serial/parallel pickup options, I've been struggling to work out if things are working as they should with the sound quality of the Micro Crush.

Thankfully, if you need a diminutive amp that actually sounds nice clean, there are a couple of good seeming options out there now, so I decided to go play at the weekend in my local guitar shop, PMT Cambridge, and see what they were like.

First up is the Boss Katana Mini, which is a 7 watt solid state amp with a 10 cm speaker. It has the option to switch betwee clean, crunch and brown channels (brown for some reason is the metal channel, but I've no idea why it's called that), a three band EQ, and a built in delay effect. It comes in at 23x18x11 centimetres, so a bit bigger than my old Micro Crush, but still reasonably small. The second one I tried was the Black Star Fly, which is a 3 watt amp with a 7.5 cm speaker. It has a clean and overdrive channel, a weird knob labeled IDF that lets you move between a fenderish to marshallish sound, and again a delay effect circuit. Though it is quieter, it is also notably more compact at 17x13x10 centimetres, more comparable to the Orange Micro Crush.


Having watched reviews of both amps online I think I went into the testing with unreasonable expectations, particularly of the Boss Katana Mini. If you watch reviews on YouTube like this or this, then you'd expect it to sound like a much larger practice amp, but it reality it does not in the flesh, having more of the slightly tinny sound you'd expect of a smaller speaker. However, it is streets ahead of the Orange Micro Crush in terms of delivering a clean sound that you can hear the subtleties of your guitar with. At 7 watts it's also reasonably loud, so will make a nice party in the park amp I suspect. But it is still a little tinny, though with the three band EQ you can dial some of that out. It is really good for the money and size, but I think if you've just watched demos on youtube you probably want to try it out in person first.

The Black Star Fly was quite similar, probably a little more tinny than the Boss Katana Mini and definitely not as loud, but still has quite a reasonably clean tone, and if you mess around with the weird IDF knob you can get some reasonable body to the sound as with the Katana.

For a workbench amp though, both are more than capable of letting me know if I've wired things up properly and give me a reasonable idea of what I've got. But neither are going to give me an great tone compared to plugging it into an amp with a 12" speaker, so final testign will need to still be done otherwhere for me. These are nice portable practice amps that you can stash in your rucksack and let you play where you otherwise wouldn't, but not a replacement for something as basic as say a Fender Mustang or Champion practice amp with a much larger speaker.

Downsides with both: neither has a built in tuner, which makes me a little sad. As I mentioned in a recent post, I use a clip on tuner most of the time, given I occasionally want to string up the guitars before I've soldered the electrics in them; but I don't always remember to have it on me, so a built in tuner would definitely be handy. They also have delay but no reverb, which seems odd to me, but I guess perhaps is something other people ask for despite most bigger amps just having reverb? Not a biggie given my use case, just seems odd to me.


In the end it was a tough call between the two: both have the sound quality I want and whilst not cheap, are in a price I feel comfortable stashing in Makespace. In the end I opted for the Black Star Fly, given I am space constrained, and for my use case the extra volume doesn't matter so much. But both really are a step up from your usual tiny amp, so if you need something in this space I really do recommend either of them over the usual novelty tiny amp from the likes of Orange, Fender, and Marshall.

Pre-xmas week in the shop

The week started early on Sunday, with a visit from Tristan Dales of IKARI brining his Chuncaster in for a little tune up. I wrote more about this already on the blog, so I won't go into it much again, other than to observe again that is was nice to see the Chuncaster again, and Tris as ever amazes me with his playing.

Next up I found I had a bad solder join on the tone pot in Guitar #3, so I fixed that. I failed to spot this first time around in part due to the test amp I have in the workshop not being the best and the Makespace workshop not being the quietest at times. Although the Orange Micro Crush amp for 40 quid is not bad, particularly has it has a tuner in it, the speaker really isn't that great, and it seems there's much better things out there since I got this. For example, the Boss Katana Mini has been getting great reviews, so I might look to upgrade my workshop amp in the new year.

Speaking of tuners, as a random aside, I got the TC Electrics Polytone clip-on tuner at the start of this year, and whilst it's very good as a player, a clip on tuner is an essential as a luthier. Clip on tuners work by detecting vibration in the guitar body, so when you have a solid body guitar without electronics that is too quiet for microphone based tuners and can't yet be plugged in to an electronic tuner, the clip-on style comes into its own and is in my mind essential workshop kit. There's others out there, but the TC one is quick, accurate, and convenient, so I can't recommend it enough.


I spent yet more time finishing off the amp design in Fusion, and now I'm in a position where I can start cutting the wood after a final review. Traditionally amp cabs are made from birch plywood for it's strength and resonance properties, but it's really quite expensive, so not something you want to waste. So I'll start with a practice run on MDF to get the gremlins out before doing anything in birch ply. That said, my daily player guitar is made of basewood as I got the cheapest guitar body I could for my first experiment in building a guitar, on the assumption that as someone who hadn't done woodwork in over a quarter century I was bound to destroy it. But because I was careful (and perhaps lucky) that cheap body made it all the way to the final build and I still play that guitar almost daily.

I started on the electronics on Guitar #4, which will be a Tele Deluxe style with P90s rather than humbuckers. If Guitar #3 represents a refinement of all I learned from the preceeding guitars, Guitar #4 represents trying to start branching out into new areas. Small steps, still being a t-style of sorts, but the manufacture is more based on my own methods, it has distinctive looks, and I'm really excited to try these House of Tone P90s that I got voiced for this guitars intended use as a blues player. Overall I'm hoping this guitar will have a clear Electric Flapjack stamp on it.


The soldering is all done, but I hit a snag printing the nut for this guitar which is now a blocker on completion. I've had real bad luck using the Form One to print nuts, with a less than 50% success rate. My theory is that because of the nut's narrow profile, the print base doesn't adhere properly at the start of the printing progess, so I've tried printing them out in pairs to give the overall model a wider profile. Based on my most recent run this appears to have solved the base problem, but the supports failed to come out this time, so I'm back to debugging this. I may have to give in and actually file a regular bone nut if I can't solve this, but I really like the clear nut I printed for guitar #3 and I'm taken with the idea of a black nut on #4, so I'll persist a little more before giving up.

With xmas coming up I'm not sure how much I'll get done in the coming week, but I hope you all have a great festive break, and week notes will come back hopefully in the new year!


The Chuncaster comes home for a quick tune up

It was a nice surprise visit to see The Chuncaster, our first guitar that went on stage, back for a visit this week for a little tune up. This guitar went to my brother Tristan who is in the band IKARI, and as such this guitar has seen action both in the recording studio and on stage, so whilst Tristan was dropping past to look at our latest builds we took time to see how things were bedding in.


Although when you ship a guitar you think you've done everything right, and when you receive the guitar you might think all is great, it's really only over time that a guitar settles in and any small niggles may appear. One thing Tristan noted was that the action by the nut was a little tight on on some more recent Ikari tunes this was notable, so we took the guitar into the shop and took the nut height down to make it more comfortable.


We also went over other bits and checked everything was still nice and tight and gig worthy. In the end Tristan left pleased with the changes, and I look forward to getting an update as to how it plays after the next few outings. Guitars, just like anything, take time to bed in and get used to, so occasional servicing is recommended, and it's always nice to hear the tales of how an instrument has been used since it left the stable :)


A week mostly not in the shop

Due to being under the weather for most of the week I only had one day in the workshop, but it was an important day, getting guitar #3 to live! This involved final assembly after the previous week's screw failure and soldering the electronics, and indeed fitting the electronics into the tiny cavity inside t-style guitars. For this guitar not only does it have a humbucker in the neck, but I've given it a coil tap to add an extra voice.


For those of you who have never needed to know about the insides of pickups, let me explain: a single coil pickup is just that: a long coil of wire wrapped around a set of magnetic poles, one under each string. As the strings vibrate this induces a current in the coil of wire (a very tiny current, which is why you need an amplifier). The problem with this is that a long coil of wire acts as an antenna to pickup other radio noise, which makes the pickup hum slightly if in a noisy environment. A way of countering that noise (indeed, of bucking the hum) is to have a second coil in the opposite direction that will also pickup the hum inversely and cancel it out. This is what a humbucker is, and why it's normally twice the size of a regular pickup, as it is literally two pickups. So a humbucker gives a cleaner sound than a single coil pickup, yet you normally see humbuckers used for heavier rock sounds and single coils for cleaner tones, why is that? Well, because you've got two coils you now get twice the power output from the humbucker, which will drive the amp heavier, and cause it to distort sooner.

But, one thing clever people spotted after some time was that if you have one pickup that is made of two, can't you just get it to switch between being treated as both a single coil and a double coiled humbucker, giving your guitar an extra voice? Indeed you can, and it's now quite a common feature seen on guitars with humbucking pickups. There is a switch that will simply ground the point between the two inner coils cutting one of them out of the circuit on demand. As such, this tele now has 5 sounds rather than 3 just by adding a simple switch! This appeals to me greatly as versatility of an instrument is something I feel very strongly about. Most people have fewer guitars than they'd like, so I want them to be able to get the most out of the ones they do, and even if you have many guitars on stage you want to minimise the amount of changes you need to do in order to get the sound you want.


Anyway, guitar #3 has now gone through a week of play testing (the best bit of being a luthier perhaps :) and will have some final fettling this coming week after which I'll do a video so you can hear how it sounds too.


Coming down with a cold meant I put down my soldering irons etc. and turned instead to getting some quality time with Fusion 360 to learn how to make more complex designs, using my amp project as the catalyst for learning. Although I did some simple 3D printing with Fusion a couple of weeks ago, the amp design is a much more complex thing with many parts that will need manufactured, plus other parts that I want to check will fit correctly without interfering with each other.

Screen Shot 2017-12-12 at 13.53.12.png

As part of sharing back progress, here's some tips for newcomers to Fusion 360 I've picked up on building a model like this. Firstly, treat every discrete physical part in your design as a new component in Fusion, not just a series of bodies on a single component (which is the default behaviour). This is just because components are the things that you can then place and orientate relative to each other so that they will automatically update as you iterate your design, where as bodies are considered going together to make a single part and are much more work to adjust later. Getting into the habit of making everything a component as you go along will pay dividends later in the design process.

The next tip is make use of custom parameters to save sizes, such as material thickness, overall sizes and so forth. How this works is you define a paremeter, e.g., one called Material and we set it to 18mm as that's how thich the plywood we'll use for the amp will be. Then when building a panel, rather than specifying the wood thickness as 18mm in each place, I can just specify "Material" as the dimension, and it'll set it to the correct thickness automatically. Even more importantly, if I then decide I want to use wood of a different thickness later, I just change it in the parameters editor and all of my design will update to use the new thickness automatically without me having to change everything manually.

Screen Shot 2017-12-14 at 10.08.41.png

Even better, when specifying dimensions in Fusion you can set it not just to a parameter, but you can use formula based on the parameter, so if you have a side panel that is the overall height of your design but minus the top and bottom panels' material thickness, you can just write that into the dimension field as a formula and it'll automatically work out the size for you.

Again, getting into the habit of specifying every number you can as a parameter will pay dividends later, particularly when you come to the final tweaking stage where you want to fiddle with values to check how it impacts your design.

The final tip is to learn to use joints, a feature of Fusion 360 that allows you to define how your components are assembled relative to each other. In the picture of the amp you see here I never specified the positions of the parts using numbers, I just told it that the side panels attach to the base, the lid to the sides, the metal chassis to the lid, and the vacuum tubes to the chassis sockets. It's really easy and quick to put together that model this way.

Screen Shot 2017-12-14 at 09.45.01.png

The double whammy of using joints and parameters is that now my design is effectively configurable: I want a bigger speaker, or I want the cabinate a different size, I just change the parameters and the entire design just changes to reflect that. This makes that final design stage where you want to adjust things lots of fun, as you can readily do lots of "what if?" tweaking just by changing the parameters and immediately seeing the results.

To Mr Hillhead, my high school technical drawing teacher, I hope you're proud that some of what you taught me is still there!

Next up now that I have an amp design I'm happy with is using Fusion to generate the CNC Router controls so that I can cut the panels from a sheet of ply. I really do like the fact that Fusion both for this and 3D printing is an end-to-end tool.

A(lmost) a week in the shop

This week was meant to be about the assembly of the guitar #3 and #4, but I hit a bunch of small setbacks that have pushed things out another week. Whilst I don't wish to whinge, I think part of trying to document progress is explaining what to do when things inevitably don't go to plan.

First up, is the lack of standardisation on parts for guitars in terms of sizes. This is something that is in part down to my naivety: you'd assume that if you got part A for a particular guitar and part B for that same type of guitar that they'd just fit together; that after all is part of the benefit of starting out recreating a well known design, to simplify the process in part. But no, you often find this happens:


Over time a lot of guitar parts have changed shape subtly, switching from imperial to metric, or just model evolution over time. Unfortunately, unless you are aware of this, you end up with surprises like the above. In fact, that was one of two such surprises I had this week, as on the orange deluxe it turns our the bridge I bought was a few mm longer than the standard too. In both cases a little modification to parts is possible, but for the want of a couple of days it's much nicer to get other parts that fit and then use the now spares on other guitars. Anyway, as a lesson for any would be luthiers reading this: don't just measure twice before you cut, measure twice before you order; the only problem being that a lot of parts don't specify these things, so you do in the end need to learn which suppliers use which sizes and find ones you like.

The other thing I'll take from this is that I should do a test assembly as early as I can so I don't find all this out at the last minute. However, the order in which I put things together as I moved from planks to guitars didn't lend itself to this, so I'll rework that flow next time.

The other setback I hit was one I documented here a couple of weeks ago: yet another screw split as I was screwing it in:


As before this was another 2mm sized screw that came with a part, and I think the lesson here is that the screws that come with parts aren't necessarily the best, and if they're very thin and your wood is tough, then consider using screws you know to be good rather than the ones that come with the part (a lesson imparted to me by others now that I discussed this with them). Of course, sometimes things just go wrong, but two in a month smarts a little. I had drilled a pilot hole, I had made it deep enough, and so forth, but it still just snapped. Thankfully, having gone through this once before, I was able to recover it quickly using the same method, but it's still annoying and still costs me a day as I needed to let the wood glue I use as part of the repair set.

So a week of small setbacks, and a few instances of counting to ten slowly. I try in general to share the positives of progress on the guitar, and I suspect to some degree I make it look easy on my instagram stream by posting all the stages of completion as I go along, but it's not all plain sailing, and you have to expect the occasional setback. Part of what I like about luthierie is that it's a very different discipline compared to writing software or designing electronics: you can't hit undo if you make a mistake, you have to find a way to move on regardless. At times it is very frustrating, but then that's also in part what makes it satisfying as your design comes to life as a musical instrument.

Onto happier notes, guitars #3 and #4 are close to completion now, and starting to look wonderful:


Hopefully next week will mostly be spent assembling the guitars assuming a smoother week than this one.

Aside from working towards the end of the current guitars, I've done some more work towards the first amp build. Mostly that has involved a combination of notebook sketching and continuing to learn how to use Autodesk Fusion 360. The build will be a 5W 1x10 combo amp for practice/rehearsal. For those not familiar with all that, it basically means it's a single box that'll contain both the electronics and the speaker (a combo), be reasonably quiet (5W is not enough to gig with, but plenty for home and practice use), and it'll contain a single 10" speaker.


Having got the parts from various sources, the next thing I need to do is design the cabinate to house it in. The plan is to produce the design in Fusion 360 and then produce the panels on the CNC router I have access to at Makespace. To do that I've started first modelling both the metal chassis for the electronics (which also has the controls on it) and the speaker. This has been a nice introduction to Fusion 360 before I try building my own bits from scratch.



Finally, as guitars #3 and #4 do draw to a close, note that I'm still have the slots for guitars #5 and #6 open, so if you fancy a guitar that is built for you, get in touch. Pricing and timing will depend on what you want: I'm pretty adept at producing t-style guitars now as you can see, but I'm itching to have a go at other solid body or semi-hollow body designs: perhaps an S-Type takes you fancy, or a Les Paul Jr for something a little more raw? Or even something that it's just hard to get like a baritone or a custom scale length variation.

Using a laser cutter to make fretboards

There's many ways to approach most problems when building a guitar, and the method you pick will be in part dictated by which tools you have access to at the time. As such, when it came to making my own fretboard for the first time, rather than buy new tools for the job, I first had a look to see what things I had access to at the maker space I work from that might help. This has resulted in my using a combination of low tech and high tech in a way I've not seen used anywhere else, so I thought I'd document the approach.

To start with, I'm making necks with a maple back and a glued on fretboard, so I start with a piece of wood thickened to around 6mm that I'll use for the fretboard. I leave this wider than the neck itself, and I'll use a hand router to make it flush with the neck once glued in place.


My aim here is to make very narrow slows so that I don't have to glue the frets in. Fretwire has a toothed piece called the tang which will stick into a slot in the fretboard. The tang is typically 0.5mm in width, so ideally you want your slot to be slightly smaller.

The main tool I have access to that my technique is based around is a laser cutter. A laser cutter is basically a machine into which you place a bit of wood and it has a laser head that can be moved over the piece in 2D to either cut or etch the surface, depending on how much power you provide to the laser. It's a really good tool for cutting sheets of ply wood or etching nice designs into thicker bits of wood. But the question is, can we use it to cut fret slots? To work this out, I spent an evening playing with various powers and slot designs until I ended up with something that worked:


In the end I realised I couldn't reliably cut the slots using the laser. To cut the wood you need quite a high power beam, which will cause some charring around the slot, and it's hard to control how deep the slot is; normally you're just cutting through the piece, so it's not an issue, but there's no reliable way to specify you want to cut to a given depth, and even if you can estimate the right power, the depth would vary with the wood density and the mood of the laser cutter that day.

So instead I use a hybrid approach: I use the laser cutter to etch a 0.5mm grove that is tenths of a mm deep for each fret slot, and I use that as a guide for a simple back saw to cut the slots by hand. With a suitably narrow saw (I have a Japanese Dozuki back saw that has a 0.3mm wide blade) I can then hammer in my frets without needing to use glue to hold them in.


The nice thing about the laser cutter for this is that etching the slots takes a couple of minutes at most, whereas if I'd tried to use the CNC router to do the same task it'd have easily taken an order of magnitude longer. You can get a sense of the speed from this video:

The final result works quite well, and I think is a nice combination of new tech to ensure the slots are accurately positioned, and old tech to cut the slots for that hand built quality that people want in a custom guitar.

So, if you have access to a laser cutter (a common tool at a community maker space, which is how I access mine), then all you need to do this technique a design file that'll generate the slot patterns to feed to the laser cutter. Which is why I wrote the fret design generator tool - simply tell that the scale length and fret count, select the laser etching option, and you have a design ready to feed to the laser cutter software. If you give this technique a go, do let me know, I'd be keen to see if it works for others as it has for me!

Another week in the shop

Once again I've had the luxury of being mostly in the shop this week, so again I'll take this opportunity to write up some of the goings ons.

This week guitar wise has been about getting those two necks finished and oiled. Monday was a repeat of the previous monday where I sanded the rosewood neck and finished off the frets so they're both nicely polished and feel smooth as you run your hand up and down the fretboard. Having then finally got both necks frets happy, I ran them both through setup: stringing both guitars up, checking the action, doing truss rod adjustments and so forth, confusing anyone who came into the workshop room at Makespace to see a bearded hippy playing guitar in there.


This means both necks are now into the stage where I'm oiling them twice a day. For this I'm using the Crimson Guitars finishing oil, which I quite like as a finisher on maple necks. Although people expect maple necks to be lacquered and glossy, I find that half a dozen coats of oil (either the stuff I use or Tru-oil is the other popular choice) leaves the neck with a satin like smooth finish that feels to me much nicer as you slide up and down the neck. Ultimately it's down to personal preference, and the fact I don't have a spray booth.


As part of setup, on both guitars I tried out my new nuts that I've 3D printed. The nut on a guitar is the bit of bone or bone substitute that sit at the top of the neck near the tuning pegs that guide the strings down the neck. Whilst not the hardest of things to make, I do find shaping and filing the them a bit tedious, and as is probably becoming a running theme in this blog, I like to see if I can remove tedium where possible.

So for fun, and as a way of giving me a project to learn more about designing in 3D and using the different types of 3D printer available to me in Makespace, I designed, prototyped, and produced my own guitar nuts. To do the initial design I used Autodesk Fusion 360, I did a rapid print using an Ultimaker 3D printer to get something I could test fit, iterated through that again, got something I was happy with, and then did a final print using the Form One to get something that is of more production quality than I can get from the Ultimaker. I use the Ultimaker for the test prints as they tak 20 minutes at high resolution vs 90 minutes on the Form One, and have a similar proportional materials cost, with a test print costing 5p on the Ultimaker vs 75p on the Form One. You can see all the stages here:


On the left is a bone nut that I filed, then the first prototype which was too flat, the second prototype which had the right radius, and then a final print in resin to get a nicer finish for actually stringing up. Note the string channels are in the design to give me the correct spacing, though I do need to fine tune them with a file, it's way less work than for the bone nut.

I'll publish the design files once I've got some confidence this actually is sensible. The first resin version I printed was a bit too tall in the end, and the string channels a bit too wide, so I got some buzzing on the lower strings. My main concern with this approach is that although the resin can withstand the stress (at least so far) of being strung up, I'm worried about the strings cutting into the resin over time as the guitar is tuned up and down. So I'm not yet confident to encourage people to take this route, but it's still an interesting learning experience for me. Here's a picture of the nut in place: I quite like the clear look as it's both unusual but not distracting.


The other reason I've not published the design files yet is just where to publish them. I could make a public link through Fusion 360, but I can't see a way to add LICENSE and README type files. I could use Thiniverse, but it's list of supported files doesn't include Fusion 360 archives, so I'd have to put up just the STL file, and I'd rather publish the original design as people will need to tweak it to fit their particular neck measurements. I could just use github as I've done for software (see below), but I'm not sure that's a good place for 3D models. If anyone can guide me here, it'd be much appreciated.

Moving on, I have a tutorial post I want to write about how I use the laser cutters at Makespace to make fretboards, but before I do that I needed a way to ensure other people could readily follow in my footsteps, which will require having a suitably made up fretboard design file. I did this by hand the first time, and I was going to just publish that design file, but I decided it wasn't that hard to make a tool to generate those files, so I spent a couple of days on that instead.

I made a simple webpage based tool that will generate fretboard design files so you can just enter the details of the neck you want to make, and the page will spit out either an SVG or DXF design file that you can then load into your design tool or CAM software. This was also an excuse to brush off my dusty HTML5 and Javascript which I've not had to use in anger for a while. I used MakerJS, which is an open source project from Microsoft that is aimed at making it easy to programatically generate parameterised designs that you can then send to 3D printers, laser cutters and so forth. It's really quite nice, though lacking in a couple of areas, but if those become an issue then it's open source so I can just go fix them and contribute it back. My fretboard design generator is also open sourced, and you can play with it here.

template generator 2.png

Finally, I ordered the parts for my next project piece, a simple 5W 1x10 combo tube amp based on the old Fender Tweed Champ/Princeton circuit. The circuits for these are very simple and well understood, so I thought a good place to learn. However, it's not just the electronics I want to get back into (I actually did two years of this at Uni, but haven't touched it in two decades!), I also want to use this to push more into 3D design, modelling the cabinate in Fusion 360 and hopefully generating the tool paths for the CNC router for the panels from that. I imagine that will take a while, so don't expect this to be done by next week's post!

Fretboard design generator

Fretboard layouts are one of the more fiddly bits of building a guitar. Even if you're using CNC machinery as part of your workflow, taking the output of a fret spacing calculator and entering it into your design tool is very tedious.

For my first fretboard, I found an existing design file at the right scale length, but now someone has asked me about a baritone neck, so I was back to square one. Being both lazy and a software engineer, I decided to automate the generation of design files into a simple webpage, which you can access here.

template generator.png

It's (hopefully) simple to use: you enter the details of the neck you want such as the scale length, the number of frets and so forth, you get a preview of you fretboard along with the positions in a table for you to confirm it is what you want, and then you can export the design as SVG or DXF. This means you can import it into most design and CAM software for final tweaking and then to production. Here you can see one imported into the tool I use for driving the laser cutters at Makespace:


At Electric Flapjack Guitars we're a big believer in contributing back to the luthier community that continues to help us out, so this tool is open source for others to play with and contribute to. The tool was mostly created using MakerJS, a nice Javascript library from Microsoft that is targetted at people trying to make it easy to generate designs for the kit you find in maker spaces programatically. The library isn't perfect, for example it won't generate polylines in DXF, so you still need to close them in whatever other design tool you use if you pick that option, but overall it made it nice as easy to create this tool.

A tip of the hat to Matt of the awesome Fidelity Guitars who gave me some feedback on this to help improve it!