4 scratch build

[Rabbit]

The saga continues...after some research I found out that even though they are related species, the bubinga Ric was using at the time for fretboards is almost twice the hardness of the ovangkol they used for the middle strip of the neck. I went on a quest to Trend out near Windsor (a long treck for me ) to see if they had any bits left (bubinga being listed it's really rare). They did, but the piece was about enough for an entire instrument and I don't have the many hundreds of $ (or thousands knowing this joint) so I didn't enquire further...it was a wasted trip. Did find they had gidgee for a fellow shedder. After some trawling it looked as though the best substitute was south indian rosewood. I orderd a piece , but I have no idea which of the three pieces photographed it will be or even if it will be one of those three..could turn out to be a different timber alltogether,depending on how dodgy this joint turns out to be.

I'm trying to ease my way into being able to use the cnc router at the shed to mark out fretboards...cutting the slots would take way too long... each pass is about 0.3mm deep.. no idea what the feed rate will be but slow would be my guess...the bit is so thin it would snap if you breathed on it .

I made another jig for routing the twin truss rod trenches of the Rics... just hope it works.. the arc is very shallow compared to the centre rod trench in the Buzzard.

[Rabbit]

Further progress....kinda sorta.

I managed to get my hands on a piece of Spotted Gum that will suit the Cliffenebacker. As an added bonus it's quarter sawn.

I managed to get a passing assessment on the Men's Shed's CNC router. After doing a miniature test piece of Spotted Gum, I was able to find a nice safe set of plunge rate/cut depth/pass speed that wouldn't destroy the delicate 0.5mm bit (took me 3 broken bits to get that right.. good thing they were 10 for $30).

Then came the "accuracy" test. Running the full fretboard cycle on test blank of Spotted Gum. 20 fret slots to a depth of 1mm, a centre line to a depth of 1mm (to indicate the final thickness of fretboard... the blank is 7mm, final fretboard should be 6mm), and a second run with a 1mm bit round the fretboard perimeter to mark the edges of the fretboard and the binding (which is 1mm thick). All up there's close to 2 hours run time...It would be a *lot* longer if I actually cut the slots to full depth. Here's what I ended up with:



And a peek under it's skirt to see the colour once radiused:



After doing the measurements I discovered a few things. One is that the second run (the outline with the 1mm bit) I must have re-homed slightly off by 1mm on the X axis as the outside edge rather than the inside edge of the line was the correct location of the nut. Luckily the rout is only 1mm deep so it will disapear once I radius the board. The second thing I discovered was that due to an error on my part with the file, the board at the nut was 42 not 43mm. The measurements of the fret location had them all placed within 0.2mm of where they should be. I feel that given variation in people's playing techniques, and the larger scale of a bass, this is good enough. I wish I could say the same about the fretboard I did for the buzzard.

I was so painstaking in my measuring and marking that I was sure there was no creeping compound errors in measurement. I only actually checked the first 7 fret slots. Turns out things started to go pear shaped after the 10th fret. The 20th fret was out by about 4mm. I snapped it over my knee and tossed it in the scrap bin. Luckily I hadn't put any of my stainless frets in it yet. There is an up side. The test fretbaord happens to be right for the buzzard!

Progress is slow, but progress none the less.

[Trevor Davies]

Great persevering Rabbit. I'm glad progress has been made.

[Simon Barden]

Ahhhgh!

Have you any idea where the 4mm error crept in? Is the file measuring from the previous slot position, and so compounding any positioning errors inherent in the cutter? If so, can each slot be referenced back to the nut instead?

And is it worth in the future marking the slot positions in pencil first and doing a very shallow run with the cutting bit so you have a quick visual guide as to whether the slots are going in the right position? It could save a lot of wasted time (and wood) if it's obvious there is still a problem next time.

[Rabbit]

Ahhhgh!

Have you any idea where the 4mm error crept in? Is the file measuring from the previous slot position, and so compounding any positioning errors inherent in the cutter? If so, can each slot be referenced back to the nut instead?

And is it worth in the future marking the slot positions in pencil first and doing a very shallow run with the cutting bit so you have a quick visual guide as to whether the slots are going in the right position? It could save a lot of wasted time (and wood) if it's obvious there is still a problem next time.

The board with the error was marked and cut by hand. The one done by CNC was within 0.2mm all the way along the line. To prevent the bit from breaking, each pass is done at a depth of 0.008 of an inch, at a feed rate of 5in/min (the software is set for imperial.. it kept rounding up in metric)

The machine refernces off the home position, so it effectively does each slot from nut to fret. The dud board was marked by hand with a pair of digital calipers and a marking knife. As someone with a scientific background I was all too aware of the dangers of compounding error of fret to fret measurement, but that seemed like the only option since I don't have a 600+ mm set of calipers. I did see a method that reduced the number of fret to fret steps, but there was still the inherent risk of compound error.

[Simon Barden]

Ah, I see.

0.2mm should be fine and I’m sure that’s better than the accuracy of all those classic Fenders and Gibsons of the 50s and 60s.

[Dikkybee007]

I used to operate a Roland engraver as well as Okuma NC Mill and Lathe at my last place of employment and have been using these types of machines since 1980. Not knowing what your machines specs and programming software are I can only give you information about the machines I have used.

The Roland I used to make small fuel cells that were 25mm square and had 0.3mm wide tracks that were 0.3mm (roughly 0.012") deep and were machined out of brass and graphite. I ran the cutter at 30,000 RPM at a feed of around 100 Ft/min in the brass due to machine limitations but it should have been a lot higher, nearer 60,000 RPM and 570 ft/min. I was only taking a depth of cut of around 0.01mm or 0.0004". Brass shares similar cutting characteristics as wood so if you were to decrease your depth of cut you could increase the feed considerably. Not knowing if the spindle speed is variable I cannot say what your feed should be but if you reduce the cutting depth you could increase the feed. 0.008" depth on a 0.020" cutter is pretty harsh on the cutter especially if the RPM are too slow.

I used to use a soft jet of air to help remove the swarf as if I used the full force of air, 100 PSI, it would break the cutter, even when the machine was not running. Not sure if they will use compressed air on the machine you are using as some places are a bit funny about spreading dust around the shop but if the machine is fitted with a good extraction system then the compressed air is not needed.

As for the 4mm error in fret measurement I take it your using a 150mm caliper so if you combine that with a 600mm rule you can achieve the same accuracy no matter how many frets are on the board. Once you get to the limits of the caliper in absolute measurement just place the ruler at the last fret measurement and then start again with the calipers.

You need a chart with all the measurements in absolute and if the last measurement that can be measured with the caliper was say 130 then place the steel rule and line up the 130 mark with the nut and I would probably clamp it in place. Then if the next measurement is another 30mm from the last then just set the calipers for 30mm and work from there. Then each time you reach the capacity of the caliper just move the rule.

You could also just cut a series of wood pieces and measure with the calipers and mark the measurements on them and then do the same as the rule. Just add the sizes together, take it away from the size you need and make up the rest with the calipers to achieve the length required to place the fret.

[Rabbit]

As usual progress is slow. I have finally started to get a handle on the shape of the inlays for the Lado Unicorn copy. It looks like they are abalone, so I have some to experiment with, and a jewlers saw on it's way.

As part of the experiment, as well as seeing how well I got the depth of the fret slots right, I marked up a test board with fret slots for the 2nd & 3rd fret, and 14th &15th frets (where the widest inlay lives). Between the 14th and 15th frets I also routed a rectangle 3mm deep (it will be 2mm deep at the centre once the fretboard is radiused) and poured some epoxy with red mica powder mixed in to give a red pearloid effect. Will be interesting to see how it looks after radiusing and a little wipe with oil or epoxy. :



The test piece is Jarrah in case any one was wondering.

[Rabbit]

Progress as usual is slow. This time too many other non related projects. After the failure of using my neck shaping jig to radius fretboards, I made something, it starts with a J........

I did some modification but I stole the basic idea from a pdf I downloaded from a place called nextproject; This one is for making a radiused sanding block:




And this is how I set the depth of the final cut. The final cut being 25.4mm from the surface of the jig curve:



The high tech instrament used to determine where to set the depth is called..... a toothpick! The bolt is adjusted to just contact the toothpick and this gives a reference to set the depth of the router bit.

The finished sanding blocks:



I made the totes from American Red Oak. The knobs at the front was turned from European Beech. The two knobs are different because I was able to refine my technique on the second one. I'm not allowed to use the wood turning lathes at the men's shed, to become assessed is akin to a 6 month apprenticeship. I was however able to get assessed more quickly on the smaller of the two metalwork lathes, which can still be used to turn timber.

You may have noticed that one of the sanding blocks has been laminated up. 5 pieces in all. The short reason for this, as anyone who has been following this thread has probably already surmised, is that I am batsh*t crazy! The longer answer, is that the neckthrough section of a Rickenbacker 4001 bass is a three piece construction of a strip of Ovankol ( often called Shedua in luthiery circles) sandwitched between Rock Maple. The fretboad radius on a Rickenbacker 4001 bass is 10 inches, so the lammed block is a 10 inch sanding block. The 12 inch block is Bluegum.

And this one is for radiusing a fretboard.:



Here's the finished surface:





The lines would suggest that the bit is not square with the surface to be cut. When it comes time to actually do the fretboard I will probably shim the router base to correct.

I had intended in making the sides of the radius trolleys by hand, but after a little while it dawned on me that I have access to a CNC router. So I used that to more accurately cut out the sides.

I hadn't read all the printed material on the pdf, and at first couldn't wrap my head around why the fretboard trollys needed to be 1 inch larger radius than the intended radius. For some reason my brain was stuck on the notion that it should just copy whatever the radius of the jig was. I had to picture in my mind a bunch of circles within the circumferance of the circle of the jig radius, then a light went on. this was also why using the neck shaping jig didn't work.

Conversely the sanding block trollys are 1 inch smaller than the final radius.

Well that's all for now, but I think the ducks are all slowly lining up.

[fender3x]

Rabbit you are truly the king of the jigs. I had seen something similar online, but your is much more compact.

I wonder if it might not be quicker to clean up the fretboard using the sanding block rather than shimming the router? In the pic it doesn't look for off.

The other thing that occurred to me when you mentioned the 1" difference between the trolley radius and the fretboard radius is that if you were to add a mechanism for raising and lowering the router, you could use the same jig to cut fretboards to almost any radius, and still retain the compactness of your current jig.

Amazing work!