How accurate should fret slots be?

Honestly, if you are concerned about accuracy why not just buy pre-slotted and radiused fingerboards.

Of the guitars done in my shop, the slotting and radiusing took hours to do correctly. Taking longer to make sure you have an accurate board at the end of all that work.

By the time you buy a board and do that work, you could buy one already slotted and radiused for only a little more than the cost of the board and certainly cost you less in terms of how much time it takes to do the work correctly.

Here are a couple of options for buying boards.

http://www.allenguitar.com/fretbd_pre-slot.htm

https://www.stewmac.com/tonewoods/wood-parts/fingerboards/

https://www.lmii.com/179-pre-slotted-pre-radiused-fingerboards

Personally all future build I attempt will be done in this manner. It's really not worth your time to do the work. Of course this is if you are in the US... in other countries it might be an issue or cost far more.

Of course, there are the folks that want to do every step of the way on their own. That's awesome! I say go for it!

But many folks are now going the CNC route, so it's no different, except you don't need to buy a CNC.

Just my unsolicited $0.02

 

Yup. Get yourself a pre-slotted fretboard and do this:

 

Did this at the beginning....

 

Clever...but is the razor blade the size of the kerf? Wood will shrink and lengthen with the change of humidity.... Both the stewmac and LMII systems are non wood materials in comparison. No doubt it's a good work around but....

My single edge razor blade in my scraper measures .018".

 

Agreed but it's a heck of a lot cheaper. The stewmac one is over $300 and the LMII one is about $225. If you make a lot of necks I get it but... A piece of tape on the razor blade would take care of the difference. Or the OP could make his own miter box and buy the slotting template from LMII for $23.. I'm cheap BTW..

 

I totally get it. The miterbox could be baltic birch plywood with a template and a pin. That's what the LMII one was originally.


By the way, he could have cut the top part instead of all the way though. Then he could have drilled larger holes on just the one piece to make it adjustable against the saw. the gap would be about 1/8 minus the saw thickness. That way the sides would be all in a line and more accurate against movement. It would be ideal for a one off scale but I still think having the most accurate system to cut kerfs is the way to go. As always, YMMV.

 

Assuming a well tempered scale I did the math on fret positions for a 25.5" fretboard up to the 24th fret.

I did a bit of a google to find out what was detectable by the average ear as 'out of tune', and all I could find was that the average ear could detect a 5 cents difference in tuning.
There are many factors that affect the tuning beside the position of the fret, e.g. the fret height / crown, the string action, the string type / gauge / age, how the note is fretted i.e. position, pressure etc, and all of these (and other) factors accumulate error in the tuning.

Just looking at the fret position alone I tried to do the sums on tuning, specifically what the allowable tolerance in fret position is to achieve a certain tolerance in tuning. Looking at just a value of tuning tolerance of 2 cents these are the results I got. Don't know if they are right or not, so don't rely on them for your measurements.

The results are in metric (units I'm used to) and Imperial units.
As would be expected the precision required in fret position is greater the further you go from the nut, and there is more tolerance on the flat side than the sharp side wrt the fret position.
I expect that given all the other possible factors that effect tuning the 2 cents allocated to fret position is a bit generous.

When I built my one and only guitar (so far) I marked my frets with a metal rule and a sharp pencil and used a fret saw and a hand-made block (indexed by eye). On the upper frets at least I don't think that I would have been able to achieve better precision in fret position than shown below, and my tuning is not noticeably off, although I don't know if my ear is considered average or worse.

Maybe a waste of time, but a fun exercise none the less.

 

Chickenbonejohn has a combined 25 1/2" & 24 3/4" scale 2mm Stainless Steel fret template for under £20.00, plus easy plans on how to make a mitrebox fret slotting system / or Beechwood mitrebox for sale if you really can't make one.

After I made up my mitrebox using some Oak scrap I slotted my first fretboard, which was no difference to a pre slotted board from forestguitarsuppies of the same scale. After 3 fretboards the money I spent on the template & fretsaw was already saved by buying unslotted hardwood for fretboards instead of expensive pre slotted boards.

https://www.chickenbonejohn.com/products/fret-scale-template

 

I've little experience but what I've read is that the frets should be +/-0.5mm accurate. I've also seen people scribe the lines with a knife and a ruler to get them in the right place before using the fret saw. One of these days I'm gonna do this...

 

Well, I guess it depends. How accurate do you want your tuning and intonation be?

 

l

Gang saws caused a lot of mass produced inaccuracy.

I measured a Gibson melody maker, a 63, because when compared to a modern jig cut
board I ordered...it was totally off. It is so inaccurate that I won’t set it up, because to my mind
there is no room for error here and I am seeing over 40 thou runout on the Gibson.
It causes wonder as to how rife this problem is.

A top luthier explained it as a set up and or blade dullness error. The guy at the assembly line
doing set up was having a bad day, and got sloppy with his shims, refused to measure results
or if he did, simply overcranked the arbor nut and called it a day.
The result is thousands of unplayable fretboards, but who is paying attention?
I mean...if you just banging strings hard enough, they’ll harmonize, right.

I have tried using set up blocks (gage blocks), calipers etc.,for a layout but it is time consuming and error prone.
I was advised to just get a good jig. There is no error and it can be done with eyes closed.
There isn’t any substitute for accuracy here.

 

Keep in mind that the string doesn't ring precisely at the contact point on the nut, fret, or the bridge either. There's no way all the strings intonate *perfectly* for all string gauges given a straight fret. Then the action height makes a difference. Then there's your sort of accidental bending as you'll never press the string exactly down. Then there's the precise location of the crown on that fret wire. Then there's the accuracy of your tuner... If you're using a jig there's the accuracy of your jig. Then after you play it a little the top is now flat and the intonation point moved. The tuners are also never perfectly stable as you're playing the intonation will move slightly. So... It's the sum of everything, as long as the contribution of the fret location error is small enough the other components will dominate. So maybe the correct answer is make them as accurate as you can, lots of guitar builders cut frets without any fancy equipment and people pay thousands of dollars for those instruments (though those guys have their technique pretty polished wrt/ to accurate woodworking).

A "real" CNC can be accurate to microns (so a few orders of magnitude more than needed).

Interesting read from a different forum (with some other references).
https://www.mylespaul.com/threads/fret-slot-tolerance.209589/

Don't let the error accumulate though, that seems to be the most common mistake (measuring fret to fret instead of nut to fret). And don't place the bridge at the wrong place either Also the further down the neck you are (towards the bridge) the bigger the impact on the pitch, so pay more attention to those.

 

It's not necessarily a problem; it's the reality of mass production. Fret slots should be as perfect as possible, but that's a goal--not a reality.

There is always variation from the ideal, but so long as it's within "go" specs, they go. The amount of acceptable deviation from perfection is defined by the maker. Also, strings are not fretted in a laboratory when music is being made. The notes produced by fretted-string instruments in the real world are imperfect due to things like temperature, humidity, how excited or relaxed a player is, etc.

Hand cutting in a jig is more accurate over time, but production numbers are a lot lower and production costs are a lot higher. Finding an acceptable balance is the trick. If you're making guitars one at a time by hand, it's possible to control the quality to a much higher standard.

 

Susan Gardner explains making a fret locating template from an engineering perspective, maths are not too complicated if you pay attention!



Whilst I like the idea of trying a DIY fret template the same way in theory, I'm even more impressed with the template she accurately makes in the video.

 

I slotted my fretboard by hand and did the measurements using a ruler. I wasn't sure about how accurate I was so I compared it to a pre-slotted fretboard I purchased afterwards. My fret slots were not dead on. Some were, and some were out by at least 1/64" and maybe 1/32". I really wanted a birdseye maple board (from the same wood I made the body from) so I used my fretboard anyway. I worried about the intonation being off, but I cannot hear it on any fret. Either my ears are very bad (a distinct possibility) or the error is minimal compared to the other things that also make guitar notes out of tune (e.g. too much pressure on the string, slight bending, etc.)