Official Luthiers Forum!

OK, time for some advice here. I have gone round and round on fret layouts, and am not satisfied with the accuracy. I have tried paper templates, layouts with a ruler, a local machine shop gave up on making me a template (not sure why,) and finally borrowed a Starret ruler in 1/64th gradations. I used a sharp marking knife and a magnifying glass, and made my own template for the Stew Mac fret slotting miter box. Good to go.

Today, I double checked the accuracy of my template one last time, and I am not satisfied that it is good enough. I put the ruler back on the template. Some seem dead on, others seem to be off by possibly as much as a 64th. I have come to the conclusion that for a number of reasons, this is as accurate as I can get. I think it may be a stackup of errors. I can likely mark the template well enough, but the act of cutting the notch is introducing more errors.

So, is 1/64th accuracy good enough? It would seem at this point that I have a decision to make.

1. Use the template, and let it be what it is.
2. Make another template and keep beating this dead horse.
3. Write off the $175 I have invested in cutting my own slots, and spend another 80 bucks on some pre-slotted boards.
4. Admit failure, give up, and walk away.

Any advice appreciated, because I am now stalled but good on this project, and if I can't get past this, my first two will never be finished.

Thanks for the help,

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

-for me, .015 is just not good enough. perfection is my curse.

-are you using an odd scale? why can you not buy a ready made template? i picked up a 24" scale template from lmii on sale for $18. it had the centerline positioned off center -but who cares? lmi will make you a custom one as well. maybe sm will too. i assume that since you are considering pre-fab boards your scale length isn't unusual?

-i've actually made them myself on a ProtoTrak milling machine. very easy. i suspect to your machine shop it just wasn't worth the time.

1/64 th is a little bit too much IMO.
You could buy a pre slotted board and find a way to use that as a template. If you cut it length ways into strips you should be able to get 4 fret rules/templates. You could even build a mitre box out of two of the strips.

When I cut slots by hand I use +/- .2mm (.008") as the standard. Some may feel that's too lax, but I'm not sure you can hear it. A lot of old guitars I've seen were much worse than that. At any rate, I'm not sure you could get much closer by hand and eye.

MichaelN's idea is a good one: We all have cutoff strips from fretboards around. They would, of course, wear over time...

You have to ask yourself how much error exists in your fretwork with respect to the break point of the string over the crown of a fret versus the slot underneath, and how that point moves away from the nut as the fret wears.

I would seriously consider buying from fret scale templates from LMi or Stew Mac. They are not expensive when you consider that you get 2 scale lengths and don't have to worry about accuracy. A properly setup miter box or a simple sled on a table saw combined with these templates will remove this frustration.

Josh

_________________
Josh House

Canadian Luthier Supply
http://www.canadianluthiersupply.com
https://www.facebook.com/canadianluthiersupply?ref=hl
House Guitars - Custom Built Acoustic Instruments.
http://www.houseguitars.com

I dont make that many instruments and seldom use the exact same string length so I use a print out from WFret. Most decent laser printers will print to less than 0.3mm (Sorry you will need to translate that to inches) so that is a lower cost option for you. Dont forget that factors such as string height, tension and mass will affect the intonation as well.

Thanks, I appreciate the feedback. You are telling me what I expected to hear, that 1/64th is just not accurate enough. To answer some questions:

The guitars are Antes Parlors, the scale length is 24.500". Lmi makes a template for this scale length, SM does not. The template made by Lmi will not work on the SM miter box, since they use different styles of index pins (the Lmi uses a pointed pin, SM uses a straight round pin. The straight pin will not fit in the pointed slot in the template.) SM will not make me a custom template, I already asked. I could purchase the Lmi miter box and saw, but it would be cheaper to just pay them 9 bucks a piece to slot a couple of boards for me.

Looks like it's time to start over, and see if that old horse still has a breath in him.

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

Thanks, I tried WFret, the problem with it was the same as with everything else I've tried.

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

Build the guitar with a 24.9 scale length since you can get templates, .4 difference will not change the guitar that much. Just move the bridge plate to the proper position for that scale length and build. The Antes plan is nothing special as far as designs go and all I would use is the outline and modify the neck and bridge plate for the slightly longer scale. As for accuracy .015" probably will not be noticed by anyone as long as you do not compound the error. Lots of builders use the printed template glued to the blank to cut the slots with good results.

Fred

_________________
Fred Tellier
http://www.fetellierguitars.com
Facebook page http://www.facebook.com/pages/FE-Tellier-Guitars/163451547003866

These guys have a huge range of templates in clear acrylic and will do customs
They fit the SM mitre box nicely, some just needed the width eased to fit.
http://guitarpartcenter.com/slottingtemplates.html

I've got a little different idea here...
I'll agree that 0.015" is too far off for a template, but a single board with that much error is probably OK.
Some quick calculations indicate that amount of error translates to about 4 cents off in the worst case, which is the highest fret.
Above the 12th fret, the error should be less than 2 cents. Average ears can barely hear 3 cents.

_________________
A man hears what he wants to hear, and disreguards the rest. Paul Simon

I have several of the Guitar parts center templates and they are excellent and their # 4 template gives you a very close scale length to the plan you are using.

Fred

_________________
Fred Tellier
http://www.fetellierguitars.com
Facebook page http://www.facebook.com/pages/FE-Tellier-Guitars/163451547003866

Fred, looks like I can get a 24.562", which is only .062" longer than specified. That looks like a better fit. So, what would I have to do to change the scale length? You mention modifying the neck and bridge plate, but Keep in mind that the box is closed and done, and the necks are made except for profile carving and fitting. Everything was made to the 24.500" scale, so would the 24.562" work within that?

Also, I measured everything one more time, and will do so yet again. Looks like I have 4 frets as good as they'll get, several off by half a 64th (around .008"), a few off by 3/4 of a 64th (around .011"), and only one by a full 64th. That one, fortunately, looks to be too far, meaning I can carefully file back the leading edge and bring it closer to where it needs to be. Not sure of the others that are off yet, some may be far and adjustable, others may be too close.

Rodger, I need to check again, but I think the one that is off by a full 64th is fret number 14, not so good.

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

you could also buy the lmii tapered index pin, and install it into the side plate of the sm mitre box and use the lmii 24.5 template. provided the widths of the templates are the same or close, you could make the sm tool more useful by accepting both company's templates

pin $10
template $36

good luck!

You could use on of the other scale length templates Stew Mac sells that are close to 24.5"...
Gibson 24.625" and 24.750"
Gibson 24.562" and 25.300"

I agree with Fred. The difference between the gibson short scale (24.562") and 24.5" is very slight. Probably, wouldn't even have to move the bridge patch location. Just measure the bridge location for the alternative scale length or use a saddlematic to find your bridge location. Otherwise, if you don't want to buy a template only for this design, buy a scale length of a style you may want to make more of in the future (maybe 24.9" as Fred suggested and be sure your bridge and bridge patch is in the right location. I bet it would work fine...

_________________
Steve Sollod (pronounced sorta like "Solid")
www.swiftcreekguitars.com

You're right Al, they would wear. I thought about making a mitre box from a bought in slotted board about a year ago. Then I became concerned that it would wear out pretty quickly. So I took a piece of Walnut and cut a shallow slot in it. I then proceeded to simulate the cutting of a fret slot, carelessly putting the saw in and out of the slot. I was deliberately being 'rough' with that saw slot. I think I repeated near 100 times before measuring the slot. It had worn by 0.1 mm.
With a harder wood, a bit more care, a little candle wax, perhaps a bit of CA on the end grain of the slots - you might get 200 boards out of it.
BTW. I've yet to build it!

I bought a SM Gibson template with four scales on it. I've used the 24.3 long scale twice and find it is a tad over-compensated for low action on an electric with light strings. It is a high quality scale and has circles for the dot locations (helpful, but it is missing the last dot). I made a sliding miter jig that fits over the board (I cut my boards to match). I can line up the miter cut with the scale marks using a magnifier.
I'm happy with the SM scale.
Dan

_________________
wah
Wah-wah-wah-wah
Wah

Back when I started out you could get neither miter boxes with guides or pre-slotted boards, so I just learned to cut them freehand to a mark. I still do them that way when I want an odd scale or wood (like persimmon or hornbeam). I worked out a routine that allows me to be accurate enough; a variant of the old 'measure twice and cut once'.

When I calculate the fret scale I print out both the 'nut to fret' and 'fret to bridge' distances. The fretboard is dressed down so that it has parallel sides. Using three clamps, I clamp the steel metric rule to the fretboard with the '0' at the nut location: three clamps allows you to move one without any risk of the rule slipping. I use a marking knife and magnifier hood to transfer the measurements to the fretboard, using the 'nut to fret' distances. Remove the clamps, and scribe the lines across the fretboard using a square.

Re-clamp the fretboard and rule, this time with the '0' at the bridge location, so that the 12th fret is in the middle and the nut is at the scale length (so, for a 650mm scale, the 12th fret would fall on the 325 mark and the nut on 650). Check the fret locations using the 'fret to bridge' numbers. The idea here is that if you use the same numbers to check yourself you're likely to make the same mistakes. It's not uncommon for beginners to be off by 5mm through reading the rule wrong, but since reading from the other end generates different numbers you're likely to catch that.

I cut with the saw centered on the scribe mark, and the saw slot is .5mm. If any of the scribed marks checks out as off by more than .2mm, but less than .4, I draw a little arrow to show which side of the line to cut on. If the line is off by more than .4, I re-mark and scribe a new line, making sure to note which is the right one! If I'm going to use a compensated nut I mark the location to cut the fretboard end off there.

I've compared my fretboards with the pre-slotted ones, and they seem to be good. I've had some pretty fussy players remark on how in-tune the guitars are, so I guess I must be OK. I will say it takes some practice to get this, but once you do you need never be tied down to what the suppliers will sell you again. I do, BTW, buy pre-slotted boards when they have what I want: they're accurate, and the time saving is worth it, but I consider the ability to cut my own a useful skill that any luthier should acquire.

Thanks everyone, those are some good suggestions. I was talking with my cousin who builds on Sunday, he said he always used preslotted boards, couldn't justify the cost not to. I may do that in the future, but i would like to be able to make my own. This is after all a learning experience for me, and I have twice had to suffer the indignity of having some moron slap a chromatic tuner on one of my instruments at a show to "see how good you did."

I checked my template one more time today, and all the dimensions that are off are off by too far. This means I can gently file / sand the leading edge, and get them closer to where they need to be.

It is looking like my main problem is introduced error from vision. My eyes are not what they used to be, and I simply cannot see the gradations as well as I would like. And using a magnifying glass seems to introduce some parallax error. I believe that is also my limitation with using WFret and CAD paper templates.

So this ain't beat me yet. It may be taking forever, but I will have a pretty good idea of how to build one once I'm done.

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

I kept trying different fretting approaches until I finally broke down and got a table saw and the circular fretting blade from Stewmac. Sounds like you are on the same path.

If all the "errors" are "too far" then you can probably adjust the saddle such that the "errors" will be a little less "too far" and the frets that are "right on" will become a little "too close". Making everything a "little off" may get you to where nothing is noticeably in error.
Equal tempered tuning is a compromise to allow instruments to play in different keys, and as such intentionally "detunes" some intervals to allow for this. Slight errors cutting frets may enhance the sound of some notes at the expense of others. It might be possible to adjust a fretboard to play some notes closer to "just" intonation for specific scales.
As Roger Knox pointed out, the frets on the neck can tolerate more error than those on the body, and as someone once said "There's no money beyond the fifth fret". Other errors come into play where the strings rise over the body section of the fretboard, but they are usually ignored by most builders.
I don't get hung up on the mathematical perfection of a fretboard, but try to make it close enough everywhere to sound good.

Thanks all, sounds like real good advice. I think I may be oversensitive to this, for a couple of reasons. First, as I mentioned, I've had to endure two shows where I had to defend fret position to some idiot knowitall with a chromatic tuner. Saying "I think they're fine because it's the best I can do" doesn't get you very far in your own defense.

Second, I have a former pro at my church who keeps reminding me over and over about intonation, and can't understand why we can't use adjustable bridges like on a Strat on an acoustic, and just put a pickup inside. If the weight of the bridge dampens the volume, big deal, amp it, but get the intonation perfect. He tells me this on a regular basis, repeats it like a mantra. As he puts it, he likes to make complex chords up the neck, so any guitar that doesn't play in tune for such has poor intonation, and ain't worth picking up a second time.

It all reminds me of a couple of times at work when we would visit a customer who tested our product against a competitor, then tore both of them down to compare wear on the parts. I will never forget one day when GE showed us a near perfect part with no wear, which happened to be our competitor's, then showed us ours, which had very slight wear. The GE engineer then simply looked at us, and did a "good bye" wave. Point taken.

So, in the end I have ordered a template from Kevin Waldron. Spoke with him yesterday, super nice guy, and he tells me his templates are laser cut for higher precision than a standard CNC. That is something I can defend, if and when it ever comes up again. May be overkill on my part, but the peace of mind is worth something to me.

Besides, i tried over and over yesterday, and I am simply not confident of my ability to measure the locations that accurately to begin with. It would be interesting to know how they are measured for QC purposes in the big factories, and what tolerances are allowed. If it were me, for new school I'd use a CMM, old school I'd use a shadow graph.

_________________
Phil

http://www.oleninstruments.com

"Those who tilt at windmills are only considered insane by those who can't see the dragon."

I don't have the experience the rest of those who replied have, but I cut my own slots using a method very similar to A. Carruth's, except I measure from nut to fret, and check accuracy with fret to fret measurements. If there is error, I re-measure from the nut, and make any corrections necessary.
I just had to jump in and say that if I had some bozo slap an electronic tuner to the headstock of one of my guitars to "check my accuracy" I'd tell him and all his friends standing around to go pound sand. I, and any other player, can change the pitch and accuracy of any note simply by how much pressure I use on the string. Everyone plays with a different pressure and technique, so what might be dead on accurate for one guy might be a different story for another.
I read a story somewhere where a builder at one of these shows actually paid a guy to go around and do exactly what you are describing to a few select builders he considered competitors. Again, tell them to "go pound sand". I would.

_________________
Mike

The only thing nescessary for evil to thrive is for good men to do nothing.

'Evil tempered' tuning is a compromise, of course, but given the reality of it, some compromise is unavoidable (that's what 'temperament' means).

Trevor Gore has a really good treatment of compensation in the books he co-authored with Gerrard Gilet. Basically he walks through all the things that cause the pitch of a string to depart from the desired temperament, and shows how you can compensate for most of them. This doesn't eliminate the compromises of equal temperament, but it keeps you from piling any more problems on. It's worth the cost of admission. One thing he said is really true: once you've heard a guitar that's compensated to play in tune, it's hard to go back.

As for Phil's friend, the 'former pro', there's one in every crowd. I'd be interested to know what he thinks of as an 'in tune' instrument. Very few acoustic instruments of any sort are really 'in tune' on every note (if 'in tune' refers to the theoretical pitches of equal temperament), and some, such as small pianos, are far worse than most decent guitars. It always gives me a chuckle when somebody says that their new whiz-bang guitar gizmo will enable to you play along with a piano: why would I want to be that far off? You could, of course, say that my guitar sounds out of tune as compared with the piano, and since you can't do much to make the piano work better, I'm the one who has to compromise. Does 'pound sand' cover that? And don't get me started on organs.