Unfinished Antique - Nut/Scale Questions

[anthonyg]

There well could be errors at both ends.
My 12 3/4 was just a guess, yet to assess by eyeballing it properly you place the rule with the 6 3/8" mark, right on the centre of the 12th fret and see how things line up.

All frets measurements for rule of 18 giving you a nominal scale of 12.722 is getting very close to my guess of 12.75 and that's well within what you could expect from manual measurement errors.

Here is something that I have noticed with scale length variations that I have observed a few times yet maybe not enough to declare it a rule. Yet.
If the nut is placed a little far away from the frets than it should be, then placing the saddle a little further away from the frets than it should be will compensate, to a degree.
If the nut is placed a little too close to the frets than it should be, then placing the saddle a little closer to the frets than it should be, will compensate, to a degree.
I have thought that one day I should make a jig to test this out and see if some variations one way or another, are better or worse for intonation than another variation.
If this makes any sense to the luthiers here and your interested in the concept, then please run with it and make your own testing jig.

Back to MrOldschool's uke. My two bobs worth is correct the nuts position first, string the instrument up and see. You may then need to move the saddle or maybe not, but at least at that point you will have a better idea of the error that needs to be corrected.

EDIT again: Centre to centre is correct for fret to fret measurements. You can measure outer to outer and then subtract the fret width as well.

 

[rmmottola]

I did take all of my measurements from centerline to centerline of each fret, so if that is not correct, I would need to redo all of this. I wasn't clear on that point from the directions.

Great! So basically the fret-to-fret intervals are all reasonably accurate so there is a high degree of certainty that the calculated scale length and the determined fret layout system are accurate. The nut-to-1st fret distance is a bit long, at roughly double the typical fret kerf thickness.

Yes, as mentioned in the calculator instructions, fret distances are measured apex of fret crown to apex of fret crown. Also mentioned are that digital calipers should be used, and that longer fret intervals (for example 1 to 8, 2 to 9, 3 to 10, etc.) are preferable to shorter ones, because it makes measurement errors less significant. But given the results you got those latter two things don't look like they were issues.

 

[MrOldschool]

Thanks for the guidance. When I was noting the scale length as being way off, it was before my subsequent edits, and before I had filled in all of the numbers in the calculator. I hadn't realized at first that it would continually adjust the overall number as new data was added, so I did see that the final number was much closer to what you had said. That all being considered, it does appear that both the nut and the saddle are too far away. Perhaps, correcting the nut distance would be enough to make the variance of the saddle "close enough". I redid the calculation to include the 12th fret and got the following:

Interestingly, that shortens the scale length significantly, and throws the #1 to #2 distance into the orange...

 

[MrOldschool]

The nut-to-1st fret distance is a bit long, at roughly double the typical fret kerf thickness.

Sorry, I don't quite understand what you mean there. Are you saying it is off by double the kerf thickness?

 

[anthonyg]

There is a somewhat easy way to move the nuts position, closer to the frets, without breaking out any files or tools. Place a matchstick on the fretboard, under the strings and up against the inside face of the nut.
I use small Allen keys of 1.3 or 1.5mm size usually.
Hey presto, instant shim, although the nut slot heights will have to be in the ball park.
Check intonation and see.
You could try making a similar kind of shim for the saddle end as well. It's a simple enough thing. Just a matter of sizing it correctly.

 

[MrOldschool]

And just to throw more fuel on the fire, here is when I recalculated taking all measurements from F1:

Again, the scale length varies pretty significantly, but that makes the 0 to 1 reading closer to acceptable.

 

[MrOldschool]

There is a somewhat easy way to move the nuts position, closer to the frets, without breaking out any files or tools. Place a matchstick on the fretboard, under the strings and up against the inside face of the nut.
I use small Allen keys of 1.3 or 1.5mm size usually.
Hey presto, instant shim, although the nut slot heights will have to be in the ball park.
Check intonation and see.
You could try making a similar kind of shim for the saddle end as well. It's a simple enough thing. Just a matter of sizing it correctly.

I like easy!

 

[rmmottola]

Your images of the calculator are cutting off the top. It looks from the results like you are specifying Rule of 18 rather than letting the caclulator figure what fret spacing system is used. I would recommend letting the calculator figure it out.

 

[Bibs]

I have a simple question. What are the variations in cents?

 

[casualmusic]

Checking out mystery stuff is fun.

There well could be errors at both ends.

Yup.

Back to MrOldschool's uke. My two bobs worth is correct the nuts position first, string the instrument up and see.

Agreed about solving the nut first.

I would start by finding the scale length that matches the spacing of the existing frets 12 to 1.

As a visual person I would use that calculator and your guesses to make a few paper test patterns (templates). Draw the positions of frets 12 to zero.

A paper pattern is easier than adding and subtracting numbers. Regular paper is fine but semitransparent is a bit handier (tracing paper from the craft store, parchment paper or waxed paper from the kitchen).

I’d calculate the first pattern for your 12 3/4” guess and the second pattern for 13”. Try both and use the result to make the third pattern etc - sneak up on it.

Lay each pattern on the fretboard and look for a good match from frets 12 to 1.

Starting from the 12th will (a) let you check if the frets are nicely spaced for intonation, and (b) establish the inside face of the nut.

The best fitting pattern will give the preliminary scale length to adjust the saddle for best intonation.

Is the inside face of that nut square to the fretboard? And is the contact point at the end of the fretboard? If a new nut is needed I’d think about widening the slot to fit a stock nut.

Cheers.

 

[John Colter]

If you have determined what the actual scale length is, and the correct position of the nut, then you may use the rule of eighteen to check if the frets are correctly placed. The alternative would be to double the correct nut to 12th distance and use 17.817 - both should give the same results for all practical purposes.

 

[MrOldschool]

I was using rule of 18 setting because the uke is from the 50's and appears to be at least somewhat handmade. Since everything was saying that antique ukes are done in rule of 18, that's what I used. My screenshots are placed where they are because the rule of 18 setting is displayed at the bottom of that area as well. I didn't see a need to show both, and would have had to change my browser display zoom to get it all in. It seems like I could leave all of the fret to fret measurements in place and make calculated adjustments to the placement of the nut by 32nds or 64th of an inch to the 0 to 1 field to guess my way into what would be best, and then the calculator would say exactly where the saddle should line up, at least if I'm grasping the concept well enough.

 

[Bibs]

Y'all realize that the actual sound may not vary enough to warrant any rework if the tuning is centered between the sharp/flat notes. This side of the question has not been considered yet, just the technical, math side, is being discussed.

 

[MrOldschool]

I think I have resolved the issue by working on the nut some more. It looks like I had the grooves for the strings too level, or even possibly canted the wrong way, so that the high point was on the wrong part. I carefully worked each groove to be slanted upward toward the face, and I cleaned up the interface between the bottom edge and the nut groove. It is now sitting much lower in the groove, and the intonation seems to be lining up on the frets just fine. Using my Korg chromatic tuner, each fret position is ringing within spec now. So, in short, it was what I had done after all. The nut I had fashioned needed a little more conscious, deliberate design to do what it was supposed to do.

 

[John Colter]

Well done, MrOS,

It sounds like it has been a great learning experience.

 

[MrOldschool]

Well done, MrOS,

It sounds like it has been a great learning experience.

Truly, and thank you! I genuinely appreciate everybody's insights here.

 

[Graham Greenbag]

I think I have resolved the issue by working on the nut some more. It looks like I had the grooves for the strings too level, or even possibly canted the wrong way, so that the high point was on the wrong part. I carefully worked each groove to be slanted upward toward the face, and I cleaned up the interface between the bottom edge and the nut groove. It is now sitting much lower in the groove, and the intonation seems to be lining up on the frets just fine. Using my Korg chromatic tuner, each fret position is ringing within spec now. So, in short, it was what I had done after all. The nut I had fashioned needed a little more conscious, deliberate design to do what it was supposed to do.

Your post reminds me of an experience setting up a cheap Soprano. I filed the slots in the nut correctly and dropped the saddle height too but the intonation on a couple of strings was still out. Eventually I realised that the nut slots had been poorly moulded - never came across that before - and that the strings weren’t touching and never had touched the front of the nut. I filled the slots with superglue and recut them, the Uke then played in tune. With hindsight the reasons for such difficulties are ‘obvious’, but at the time they are far from clear.