Official Luthiers Forum!

How many thousands per cent do you move the break point on the saddle for proper intonation? Being able to start with a measured distance would be much more efficient, time wise, than cut and try.

cecil

Depends upon the string type, string gauge, scale length, setup(IE: action), etc.....

Todd, are you sure on that 0.007" figure?
Frank Ford suggests twice that amount (nut to first fret /100 by cents) ie 0.0143
I can't get my head around working it out from basics this early in the morning

First: You can't beat the discussion on intonation in the Gore/Gilet books.

The distance from the nut to the first fret is, by definition, 100 cents for the whole string length. If that distance is, say, 36mm, then .36mm will be one cent, and 3.6mm ten cents, or close enough. (yes, Virginia, cents are logarithmic too...) A cent at the 12th fret will be half that.

If you're only going to compensate the saddle, find out how sharp the _fretted_ note is at the 12th fret *, as compared with the open string, and move the break point back by the same amount. Again, if the first fret is 36mm from the nut, and the 12th fret note is 10 cents sharp, move the break point back by 3.6mm/2 = 1.8mm.

If you're going to compensate the nut as well (which I recommend) it gets trickier. When you move the break point at the nut forward toward the first fret, you are, effectively, moving all the fret back by that amount. If the intonation is ten cents sharp at the nut, and you move the break point up by 3.6mm, you've flattened the 12th fret note by 20 cents, which is probably more than you need. You end up having to split the difference. A 'quick and dirty' way that might work (I have _not_ tried it) could be to move the nut up by about half the amount you'd think you'd need, and then use the old 'moving break point' trick on a flat topped saddle to set up the compensation at that end. It might not be perfect, but it will probably be closer that it would have been with just a compensated saddle. It's also less tedious that actually setting up a beam with movable nut and saddle points, and homing in on the correct distances by trial and error at the bench. Compensating the nut does cut down on the amount the saddle needs to be moved from the 'theoretical' position.

* use the fretted note rather than the 'harmonic', since string stiffness can make the partial sharp relative to what it should be.

If you do it that way Alan you end up with the 12th fret in tune (without retuning) but the open string 5 cents flat
If you add twice the distance, you end up with the 12th fret 10 cents flat and the open string 10 cents flat and then you retune to bring them both up to pitch.

Thanks everyone for your input. I was looking for a starting point. Just a ballpark figure, so I cold adjust for all the other variable from some place close to where I will need to end up.

I have a saddle location that is almost 20 cents flat across the entire width and the break point is very near the center. Not enough adjustment to be had from an 1/8 saddle. I see a filled and re routed saddle slot in my future.

cecil

Wow 20 cents flat at the 12th?
Well for a ballpark figure my formula would require you to bring the saddle forwards by 7mm whereas what Todd is suggesting is 3.5mm.
Better be sure who is right, it makes a big difference.

7mm look to be about right Jeff. It would appear that some dummy may have set his router up on the wrong side of the layout line when he routed the saddle slot. I thought it look a little funky.

cecil

If we look at the 20% flat situation

If you move the saddle contact forward 20% of the 12-13 fret or 3.5mm then that will bring the 12th fret into tune assuming string tension does not change.
However you have also shortened the total string length by that 3.5mm so the open string will now be 10 cents sharp.
Retune the open to pitch and the 12th fret will now be 10 cents flat.

If instead you move the contact point forward 20% of the nut to 1st fret or 7mm then the 12th fret ends up 20 cents sharp(20 flat +40 sharpening)
The open string ends up 20 cents sharp too
So both the open and the 12th are 20 cents sharp, retune and they are both at pitch

So I agree with the Frank Ford approach

I agree that 7mm is a lot, hopefully the OP is exaggerating the "almost 20 cents flat"

You are neglecting the effect of the change of the string length on the open string

Yes in that example you have raised the pitch at the 12th fret by two semitones making it 1 semitone sharp, but you have also raised the pitch of the open string by one semitone by shortening the overall length by 0.711"

Now you have both open and 12th fret at 1 semitone sharp and you can just adjust the string tension to bring them both to pitch

No need for a slide rule, just look at the total picture instead of just what is happening at the 12th fret

I started with using a slide rule too, but this is too simple to need it.

Well I just did it experimentally

E string tuned to pitch, 12th fret reads e
Insert false bridge 35mm forward
Open string reads F 12th fret reads F#
Retune Open to E and 12th fret reads F

So shortening the total string length by the nut to first fret distance gives One semitone adjustment at the 12th fret.

You guys got me thinking, so I decided to do a test with my intonation rig. It was already set for a 24.9" scale. Just did the high e which has .010" action @ 1st fret and .062" action @ 12th fret. Set the rig up with 10 cents flat fretted at 12th. Measured distance from 12th to saddle break point using a 10x loupe and a hook rule so I could pull directly off the 12th fret. Then I adjusted the saddle to get the fretted 12th to 0 cents error, using a plugged in Sonic Research strobe accurate to .01 cent (might be .001 but not sure). Anyway the difference or saddle movement was 9/64" or .141", which means .014" per cent correction.

Also, I did it twice, and hope someone else with a test rig could verify.
In case you are curious here are some pics of my jig before I put the pic up on it;

Chuck

_________________
_________________

Ok I think the Frank Ford formula is well and truly proven both theoretically and practically

So it's 0,014" or 0.35mm per cent correction

Nice test!

Can you repeat that test with whatever you've got for the second string already on the rig, tuned to the same pitch(high e) as was used in the first test, please? Same scale, same action, same pitch, same everything, but a heavier gauge string. Also, tell us what the string gauges used in each test are.

thanks!

Should be Exactly the same result regardless of string guage because you are NOT in this case measuring the amount of compensation required in addition to scale length.

Chuck' jig will do that as well, but that was not the object of this test

We are only measuring how much saddle movement is requred to make a number of cents change

Ok. I've learned something reading this thread today (how did I miss it before?).

Thanks.

_________________
Formerly known as Adaboy.......

Mario;

The first test was done with a .012" string.

Did the same test with 2nd string, .016", tuned to high e. Resulting saddle movement was 9.5/64" or .148" for 10 cents. The action for the second string @ 12th fret on my rig is set a few thousands higher since it is set like I set my guitars up, so that could account for the extra 1/128".

Chuck

_________________
_________________

If your rig has enough travel you might be better using 20 cents to make measurement errors less significant. are they barrel type saddles? must be hard to find an exact break point.

Jeff, I don't have enough adjustment on the saddles to do the full 20 cents. The saddles are barrel type that I ground so as to have a definite break point. It's just a cheap electric guitar saddle that I modified, I guess I could get longer screws to do 20 cents.... I built the rig so I could get accurate intonation for various scale lengths with nut compensation.

Chuck

_________________
_________________

Should be Exactly the same result regardless of string guage because you are NOT in this case measuring the amount of compensation required in addition to scale length.

No sir, a stiffer string should, with all other variables remaining constant, still need more compensation. As the results proved. That's why something as simple as changing string brands can often lead to complaints about intonation; various brands will have different wrap to core ratios, giving the strings different slightly different intonation requirements, despite the gauge being the same. Typically not a big deal on acoustic guitars, but get an electric guy who runs a lot of reverb or has fussy ears, and it will drive him to your door in a hurry.

Doesn't matter if the saddles in this rig are barrels or not, as in this case he's measuring the movement of the whole saddle, not the contact point. Chuck, if you know the thread count for the saddle adjustment screws, you could determine the exact distance changes just by counting the turns it takes. For example, a 10-32(32 tpi) screw would mean 32 turns to move it an inch, or, 1/32" per full turn, 1/64" per half turn, 1/128" per quarter turn, etc... Might be worth determining what the thread count is and making a note of it for future reference, as it will make tests very quickly and extremely precise.

Oh, and thanks for running the tests, Chuck!

Yes sir,
It will be independent of those other factors because they have already been taken care of in the initial measurement

A stiffer heavier string will obviously need more compensation, but not then more adjustment per cent change.

Regarding Chucks Jig. I just question whether the figure in the thousands of an inch would be significant given the difficulty in accurately measuring these small movements.

This is one situation where theory and test do line up very closely.