Intonation problem: fretted notes all flat

[BlueMountain]

Please help me think this through, as it's a problem I haven't run into before.

I received in the mail today in a trade a lovely '25 Gibson A snakehead, all original, blackface, very nice tone, perfect action, nice frets. I tuned it up and discovered that the intonation between the nut and 12th fret was sharp, so I downtuned and adjusted the bridge back a couple millimeters toward the tailpiece. I finally got the 12th fret intonation right.

Then I started playing a scale. I noticed in seconds that the fifth fret position notes were all flat. Unbearable. So I turned on the tuner and discovered that in relation to the (original and apparently unmoved) nut, ALL of the fretted notes seem to be about ten cents flat. It was just more noticeble with some notes of the scale than others, but they're all out, right up the fretboard. At first I thought that the fifth fret was misplaced, but it's all of them. I tried tuning the D strings to an in-tune D# at the first fret, and the intonation was still in tune at the 13th fret.

Maybe I'm tired or something, because I'm not sure if I'm thinking this through properly. Do I need to put a shim between the nut and the fretboard, or does the end of the fretboard need to be trimmed slightly to bring it into tune with the fretted notes?

Again, the nut, frets, and fretboard look original. I'm wondering if people have been playing this out of tune for 82 years, or whether its being out of tune led to its not being played (and thus the excellent condition).

[Albert Whiting]

Before you go do a bunch of work,try a different set of strings. I have had a brand new set of strings on my mandolin and it note terribly. I threw another new set on and it noted correctly.

[dave17120]

Hi, if you can tell me the distance between the nut and fret 12, (in mm please if possible) and then the distance between fret 1 and fret 12, I can tell you if you have lost anything from the end of the fretboard at some point.

I have repaired several like this, that have perhaps had a new nut, and the cleanup before installation has been a bit too vigorous. Regards, Dave

Once you know, then you can decide what to do.

[BlueMountain]

Thanks, Dave.
Nut to center of 12th fret is 176 mm.
Center of 1st fret to center of 12th fret is 158 mm.

It's possible that someone removed the nut, sanded off the glue, and changed the angle of the end of the fretboard, thus effectively shortening it slightly. I can't tell for sure.

The strings seem to be J74s, not brand new (but sound great), and evidently what the previous owner was using.

[Jim Garber]

Compare that fretboard face-to-face with one you know is right. BION some Loar era (and post-Loar era) snakeheads had erroneous fretboards. My repair guy showed me just that way comparing with a good fretboard. Itr may not be but you never know. In that case you need to have it completely refretted correctly. Hopefully not, tho.

[Glassweb]

Some Loar era (and post-Loar era) snakeheads had erroneous fretboards.
That's absolutely true... I once had a killer-sounding black A1 snake with "The Gibson" stenciled horizontally across the peghead that was unbearably out of tune... had to let it go. I've played many Snakes and block inlay Ferns that were pitifully out of tune. Even some Loars...

[Eric Hanson]

Chris Thile said the first thing he had done to his Loar was to have Lynn D. change out the fretboard. He said his too was way out of tune.

[dave17120]

I checked out the measurements, and I have bad news, whatever else is going on, your distance from nut to fret 1 is a fair bit out.
Assuming that your measuements are good, and the whole thing is in tune if tuned up to first fret, as you seem to suggest in your first mail, then it may just be a problem with fret 1........... why is anyone's guess!?!

Your measurement of nut to fret 12, gives you a scale length of 352mm. At this scale length, your first fret should be 19.75mm from the nut. Your measurement of that distance was only18mm.... way too short. SO if your fret 1 distance is short, we can assume the whole scale length should be longer, which then makes your fret 1 distance even further out. (Hope you're following...) Typically your average carved top mando has a scale length of around 355mm. At this length, your nut to fret 1 distance should be 19.75 (not18mm).

How you sort the problem is a whole other issue, but you probably need another 1.75mm of fingerboard stuck on the nut end to put the rest back in tune. This is assuming the other frets are correctly spaced of course..... which is an assumption you can't always make, as others have pointed out!!

Hope that helps a bit, Dave

[Jake Wildwood]

Hmm. If the nut is big enough there's always the possibility he could just shave 1.75mm of the nut down flat with the fingerboard and solve the problem that way. http://www.mandolincafe.net/iB_html/non-cgi/emoticons/smile.gif

That would be my "on the road" fix!

[BlueMountain]

Thanks for making the effort, Dave. Taking your advice, this afternoon I made an ebony shim 1.75 mm. thick and put it between the nut and the fretboard. Hobbyhorse, rather that slice out a piece of the headstock to make the nut fit, I filed out a couple mm. from the headstock side of the bottom of the nut until it fit snugly. The repair is not very easy to see, given that the fretboard isn't bound.

Result? One tuner (the more skittish one) tells me that instead of being about 10 cents flat, the notes are now from 0 to 5 cents flat. The usual tuner I use says that the intonation is now right on all the way up the fretboard. It's strange that apparently, this is the first time it has been since 1925.

By the way, this morning I measured this fretboard against my '24 A jr. and my '23 F4. The A jr., which is well intonated, has the 19.75 mm. spacing between the nut and the first fret that Dave recommended. However, the F4, which also plays in tune, has the same 18 mm. spacing that the '25 A had before I fixed it. It's also interesting to me that the fret spacing between the A jr. and the A1 are NOT the same. Some frets are the same, and others vary. Yet the scale length is exactly the same. That surprised me. I guess that comes from hand-cutting the slots, but I thought they would have more exact guides. In some cases it's a matter of a millimeter.

Again, thanks very much.

[dave17120]

Glad I could help a bit. I've fixed up lots of old mandos now, and I can tell you, that in Europe at least, fingerboards can be quite erratic, and I've even found a couple where one or two of the frets were on an angle.
I'm not certain, but I think its only relatively recently, that there has been a standard mathematical formula for working out fret distances.
I guess some mandolins are/were well made and others weren't!! Happy playing, Dave

[BlueMountain]

Well, I think I'm going to have to take the Chris Thile option. Even though one tuner said the strings were in tune at all the frets, the other didn't, and the thing just still doesn't sound the way it should: the way my other Gibsons sound.

So it occurred to me that I have a Cumberland Acoustic fretboard in a box of parts. I know they use a special multi-blade setup that cuts all the slots in one pass, I assume they are accurate, and am I right in thinking that probably they made the fretboard that was put on Thile's Loar?

So I matched it up to the frets on this Gibson A1, putting the top of the first fret into the first slot in the C/A fretboard. What a shock! With the shim I added yesterday, the distance from the nut to the first fret was perfect. HOWEVER, the only other frets that were right on in the entire fretboard were 7, 8, and 15. Some were only half a mm. out, but more were 1 mm. out, and the problem 5th fret was 1 1/2 mm. off! Below the 7th fret, they were off toward the nut, and above the 8th, they were off toward the bridge. So if I were playing closed scales starting at the 5th fret, my top notes would be off by up to 3 mm.

I hate to do anything drastic to this otherwise original mandolin, but I think I'm going to be a lot happier with the sound after having the fretboard replaced. It's better now than it was, but it just doesn't sound right. And it's not that I have such a sensitive ear.

[dave17120]

Time for a new board then.......... but if its an old one, check that's ok before it goes on!! Dave

[BlueMountain]

Update taken from another thread:
The top fretboard in the photo is an unused Cumberland Acoustic mandolin fretboard. Cumberland Acoustic has a high reputation for accuracy. The fretboard below came from the 1925 Gibson snakehead discussed here that just didn't sound right.

While the use of a wide angle lense may have introduced some distortion, I shot this from three feet away, so it's not bad. The 12th fret on the Gibson fretboard was exactly right: halfway between the nut and correct distance for the bridge. The 11th fret is also correct. In person, you can see that actually frets 8-12 are accurate.

You can easily see, however, that all of the other frets are off, some by major amounts. On a hunch, I've just discovered something interesting. If I match the 7th fret of the Gibson fretboard with the 7th slot of the Cumberland, the 1st seven frets are close enough to accurate so that no one would know by the ear. If I match the 13th fret of the Gibson with the 13th slot of the Cumberland, all but the top two frets above the 12th are accurate, and the top two are close.

But if I set the intonation for the octave, there's no way for the ones that are out to play in tune (short of bending each note). This makes me wonder whether, as some have suggested, there were several templates, and one was off, or if there was one template but the cutter's hand slipped twice, or whether there were several partial templates. All I know is that it should never have happened. It's what I expect (and generally don't get) from extremely cheap mandolins of the period.

I seem to recall reading an interview with a Gibson worker who claimed that he had such a good eye that he could cut fretboards without a template, but I could be wrong about that by quite a ways. It could well be. Or maybe they used a slotted miter box, and the slots got wider from the saw rubbing against them, and the cutter took to holding the saw against the left edge as he cut the lower frets. Who knows?

[Frank Ford]

In the decades before CNC, fret slots were generally cut on a "gang saw" - a long arbor with one small circular saw blade for each fret slot, with precisely machined spacers between the blades. #That way, each board is exactly the same as the others. #Well, that's the theory.

But, what happens when the blades get dull and need replacing or sharpening? #

Sometimes spacers get mixed up, boards cut, instruments made and sold before anybody notices. #More often, the operator fails to clean the spacers and blades, so schmutz gets in there, providing a more subtle change in intonation. . .

[BlueMountain]

Well, I put a new Cumberland Acoustic fretboard on it and some stainless steel frets that were very close to the size of the original frets (within a couple thousandths of an inch). The results are pretty close to perfect, and the intonation is wonderful. What an enormous change in tone. Where before I liked the tone but could hardly bear to fret it, now everything sounds right (yes, I know it's not actually completely right, but it sounds great. Thanks for your suggestions.

This was my first time using stainless steel frets, and I did another stainless steel fret job last night with much wider frets. (0.039 vs. 0.080). It was a little harder to deal with than regular frets, but not much. It may have taken me an extra ten minutes, or perhaps twenty, with the extra filing necessary. Not significant. One difference is that generally, after beveling the fret ends with a file, I let them round off naturally while polishing everything with MicroMesh. However, the MicroMesh wasn't rounding off the ends the way they do with regular frets. So I had to spend a few minutes, after beveling, rounding the edges of the bevels with a little fret end file. A little extra work, but that made the difference. I don't hear any difference in sound, but I sure like the idea of not needing to do another fret job on this.

I used for the first (and second) time a technique I got from StewMac: radiusing the fretwire (for a flat fretboard). (Well, I did that with one fret job, as the wire was straight. With the other wire, it was in a roll already (1 foot radius, I guess).) Anyway, the Stew Mac folks say, use the wire curved, tap on each side, then tap in the center. It worked really well. Much easier than straight wire. It keeps the tang from rotating and keeps the ends from curling up on one side or the other if they aren't hammered in quite right.