MIMF

by **Robert Smallwood** » Fri Oct 12, 2012 2:42 am

Hi.
I am road testing a fingerboard radiusing router jig with callipers spirit levels - you name it.. - but I keep getting different thicknesses at each end. I will reverse the ends in the jig tomorrow, but in principle, how much tolerance is 'liveable' with over the length of the board?

<0.5mm? 0-1mm? none at all?

Thanks,
Rob.

by **Trevor Gore** » Fri Oct 12, 2012 3:18 am

Typical relief on a fretboard is ~0.1mm to ~0.3mm over 12 frets. So that will give you an idea of how flat/straight your machining has to be. Any out-of-flatness is going to come off your frets when you level them. Typical fret height is ~1mm. Straight but wedged is not quite so much of a problem because the error comes out as action error or string height above the body error which can be fixed by other ways than messing with the fretting. A 0.5mm wedge error over 12 frets resolves as a 1mm string height above the body error, or a 0.5mm action error if you keep the string height above the body constant. How easy that is to fix depends on your neck joint engineering and your bridge height adjustment.

Bottom line, you need to get pretty close or you'll be spending forever dressing frets or fiddling with the neck angle to get the action/string height above the body in tolerance. For me, getting the fretboard exactly right is the most time effective way of working.

by **Mark Swanson** » Fri Oct 12, 2012 8:47 am

I just make sure my fingerboard is dead flat before fretting, which I do after the fingerboard is glued to the neck. My jig is as accurate as I could make it, but I always check and use a scraper and radiussed sanding block to get it dead level before I fret it.

by **Mario Proulx** » Fri Oct 12, 2012 9:01 am

I'm 'reading' Robert differently than y'all...

Methinks you're making nice straight(in other words, no dips or high spots) fretboards, but the thickness isn't identical at each end, correct?

Unless you've perfectly matched the compound radius(say, 10" at the nut to 16" at the end) to the taper of the fretboard, you'll always get a variation from one end to the other. With a straight radius(in other words, the entire board is machined to the same radius, like 16"), the variance from end to end is quite pronounced. Most of us will aim to keep the edge of the fretboard even, as that's the only part that shows; the guitar player/owner will never measure the thickness at the crown.. Simply put, nobody machines a fretboard that is the exact same thickness at each end, but all of us aim to keep the edges equal over the full length.

by **Robert Smallwood** » Sat Oct 13, 2012 1:22 am

Yes Mario you've stated what's puzzling me.

It's the thickness at the centreline of each end that for some reason/s varies by around .5mm despite all relevant measurable intervals being the same at the start levelness, height from board to bottom of sled at each end etc etc. and it is a 10 -16" compound radius I'm attempting.

I've put up with it previously & the necks seem to play OK, but I am refining the jig & wondered what I am missing.

by **Mario Proulx** » Sat Oct 13, 2012 10:17 am

You're not missing anything.. The guitar will be fine.

To achieve a "prefect" fretboard would require you to perfectly match the taper to the compound radii, for each and every instrument. And even if you did, there would be zero advantage, so be content that you have done an excellent job with your jig, and keep building!

by **Andy Birko** » Sun Oct 14, 2012 10:34 am

Mario Proulx wrote:Simply put, nobody machines a fretboard that is the exact same thickness at each end, but all of us aim to keep the edges equal over the full length.

Minor hijack - I actually have a CAD model in which you enter the nut radius and it automatically calculates the end radius to get a perfectly (well as perfect as I can machine it) even edge height on the fretboard. The CNC machine does a really nice job with that

.

by **Rodger Knox** » Mon Oct 15, 2012 12:42 pm

Andy Birko wrote:
Mario Proulx wrote:Simply put, nobody machines a fretboard that is the exact same thickness at each end, but all of us aim to keep the edges equal over the full length.

Minor hijack - I actually have a CAD model in which you enter the nut radius and it automatically calculates the end radius to get a perfectly (well as perfect as I can machine it) even edge height on the fretboard. The CNC machine does a really nice job with that .

Don't you mean the nut WIDTH?

by **Mario Proulx** » Mon Oct 15, 2012 4:30 pm

No, Andy's correct.

When building a compound radius fretboard, we first choose the radius at the nut that we will want, and the end radius will have to be what it will be.

Yeah, Andy, I'm sure that your CNC and CAD skills can do it, but you'd still have to calculate the exact taper for each and every fretboard, and once you begin doing custom instruments for clients who will all want a slightly different nut width and string spacing at the bridge, you would quickly tire of all the measuring and math for what amounts to such small differences that in the end, will never be noticeable, and will still be wayyyy better than the millions of guitars that have a straight radius.

by **Andy Birko** » Tue Oct 16, 2012 9:33 am

Mario Proulx wrote: but you'd still have to calculate the exact taper for each and every fretboard,

Which is why I chose SolidWorks for my CAD software. SW is a parametric modeler which although it has some downsides, allows me to have models driven from spreadsheet data. I collect a few key parameters: Scale Length, width at nut, join fret, width at the join fret and a couple others and enter these parameters into my spreadsheet. I then press a button in solidworks and the model is re-built to the new parameters. One more button and the tool paths for the new configuration are generated.

If there is a down side to this method, it's that the customer must select the exact taper of the fretboard in advance and the neck must be matched to the FB rather than the other way around but, that's a downside to the faux-bound (and probably any bound) fretboard.

Now that I've built this parametric fretboard model, the computer part takes less than 5 min of actual hands on time to create a custom model.

Now I'm totally hijacking but CAD and CNC offers some interesting possibilities. E.g. everyone set's their neck at around 1.5 degrees to get the proper saddle height. With a CNC fretboard, I could build the set into the FB by making it taper in height from nut to bridge end. Although this probably isn't too big an advantage for acoustic builders as there's always fitting to be done but I think this could open some design doors for through neck electric makers.

This probably belongs in the other thread but I think a lot of people have the impression that CNC is best for doing the same thing over and over again and while CNC is very good at that, where it really shines is in its ability to do one offs very quickly and very well if the proper upstream software is selected.

by **Mario Proulx** » Tue Oct 16, 2012 9:45 am

I collect a few key parameters: Scale Length, width at nut, join fret, width at the join fret and a couple others and enter these parameters into my spreadsheet. I then press a button in solidworks and the model is re-built to the new parameters. One more button and the tool paths for the new configuration are generated.

Makes sense.....

Now, stop being so darn practical!

by **Rodger Knox** » Tue Oct 16, 2012 11:39 am

OK, I see now. There were a few other parameters going in. I do CAD all day at work, so I can see the advantage of CNC for one-offs that can be relatively quickly modeled.

by **David King** » Tue Oct 16, 2012 1:55 pm

I would err to the nut end being thinner. Don't over-think to the point where you don't keep moving on. No one will be measuring your work after the fact.