I Think My Frets Are In The Wrong Place....

[kingfisher]

I think my frets may be in the wrong place - and I think yours may be too!

First let me briefly introduce myself.

I have been on this board a good while, although I don’t post much. I have been making guitars for a good while longer (over 45 years) and have made around 30 instruments, plus innumerable repairs, refrets, etc. - so I am not exactly a beginner.

Back to fret placement. I have always calculated the fret positions using 1/17.818, which gives you the twelfth fret at half the scale length.

i.e. if the scale is 26”, then the octave fret comes at 13”. Seems logical, but then, when tuning,

the bridge has to move back a bit, to compensate for the string stretch when fretting. or to put it another way , the twelfth fret is not in the right place. So, it occurred to me, maybe rather than moving the bridge, which is all well and good for fret 12, but not for the other frets (the higher up the neck you go, the flatter the note), I should move the frets a bit in compensation.

This would be achieved by using a smaller divisor than 17.818 in the fret calculation. All the frets would then be moved by the correct amount. - well at least more correct than moving the bridge. The actual number is easily calculated by the way.

Am I on to something here?

I welcome your comments.

[supplebanana]

yeah but if you compensate this way then how do you compensate for different string gauges?

[kingfisher]

yeah but if you compensate this way then how do you compensate for different string gauges?

You don't. But at least you are in abetter position than no compensation at all.

Different strings ie E A D G B E need different compensation in the first place, all I am saying is that it gets you nearer to the ideal than "standard" fret spacing, so less movement of the bridge will be necessary.

[Metalhead28]

yeah but if you compensate this way then how do you compensate for different string gauges?

You don't. But at least you are in abetter position than no compensation at all.

Different strings ie E A D G B E need different compensation in the first place, all I am saying is that it gets you nearer to the ideal than "standard" fret spacing, so less movement of the bridge will be necessary.

But which string would you base your revised calculations on? The string which requires the most compensation, or the least?

If the least, the difference would probably be unnoticable - if the most, would you not have trouble intonating the strings which need the least?

[Workingman]

All tunings (except for instruments like natural horns which can only play in a few keys) are compromises. over the years various attempts have been made to "improve" on the existing standard methods (such as Buzz Fieten). I think to a degree it is just a matter of making choice and what you can live with.

[SwedishLuthier]

I think my frets may be in the wrong place - and I think yours may be too!

You bet. This is where the frets ended up on my last guitar...

[WezV]

yeah, onto something but i dont think fret placement is the way to solve it unless you go the whole hog like the true temperament neck above

string gauge, tuning and action affect how much compensation you need so changing fret placement is not ideal

I am trying out an earvana nut when it arrives. generally compensated nuts mean you need less compensation at the bridge and help no end with intonation across the whole fretboard

[SwedishLuthier]

I am personally a bit of a fan of the Earvana system. It isn't the final solution for intonation issues, but it goes a long way in the efforts to correct the biggest issues with playing cords with both fretted and unfretted strings. Its cheep, retrofitable and you only need a standard tuner to adjust things if you change string gauge

And the original suggestion would only change the de facto over all scale length. Like vintage Gibson's that had a slightly different scale compared to modern once (a true 25.75" scale including compensation if I'm not mistaken)

[Crusader]

I think my frets may be in the wrong place - and I think yours may be too!...(the higher up the neck you go, the flatter the note), I should move the frets a bit in compensation...Am I on to something here?

I welcome your comments

I do believe you are on to something but no matter what anyone thinks of it has already been done I can assure you!

The Gibson 24 3/4 scale is based exactly on what you are talking about

I don't know the process they use to come up with thier fret positions but after countless measurements and thinking and figuring things out, this is what my conclusions are

From the nut to fret 8 it is 24 9/16 (With about 0.1mm compensation at the nut)

From fret 9 to fret 17 it is 24 11/16

From fret 18 to fret 22 it is 24 3/4

I found that this process matches the measurements I took from two guitars I have owned to be accurate within 0.1 of a millimetre

[kingfisher]

And the original suggestion would only change the de facto over all scale length.

No It wouldn't. Thats the whole point.

The calculation would include the overall scale length. I am talking about the Gibson idea, as above, but rather than splitting into three sections, into 22 (or however many frets you have)

[Crusader]

Talking about compromises and string guage

Standard fret spacing only suits light guage strings (ie 1st and 2nd) because you have very little compensation at the bridge. The thicker the string, the more compensation you need. Part of the problem is already solved for you by the thicker strings being wound. Ever notice your 4th string usually has less compensation than your 3rd? Testing the intonation on a standard scale I found it got very sharp above the 12th fret on the 6th string. A higher action made it work better though

I found the Gibson scale (with a close action) provides good intonation on the 6th to 3rd strings right up to fret 22, but becomes a little sharp on the 1st and 2nd above the 12th fret. My thoughts are that it is more important to have good intonation on the 1st string all the way to the 22nd fret rather than the 6th. But this scale was probably designed to suit heavier guage strings

My remedy was to use a standard scale on the 1st string side and the Gibson scale on the 6th. In other words a "Multiscale" but I still find in necessary to have a little compensation at the nut on the 3rd string and there are some areas where the intonation is a bit 'wonky'

So I agree that changing the string guage would affect intonation but this would apply to whatever fret spacing you use. Any idea would be a compromise but I think kingfishers idea is good and why not give it a go? You might come up with a winner!

[SwedishLuthier]

And the original suggestion would only change the de facto over all scale length.

No It wouldn't. Thats the whole point.

The calculation would include the overall scale length. I am talking about the Gibson idea, as above, but rather than splitting into three sections, into 22 (or however many frets you have)

OK if my statement isn't true, please explain this to me in as simple terms as possible. You say that you should move every singe fret a tiny tine bit to wards the nut. Every fret should be move the same percentage to wards the nut. And then the bridge would end up in the place were the theoretically correct place for the bridge based on the original scale length were. If that isn't exactly the same thing as changing the scale lenthgtI clearly have not understood one single thing about frets and fretting (and I'm not exactly a beginner either...).

If you start playing around with one of the on line fret calculators that is available (one is stewmacs http://www.stewmac.com/fretcalc.html) and keep some track of your number you will clearly see that you are effectively changing the scale so that the bridge end up in the "right" way, but if string stretching compensation is eliminated you have effectively shortened the scale of the instrument.

Also look at the first reply from Crusader and do a bit of digging about old Gibson fretboards. You will see that it is the exact thing that you have described.

And another thought:

What is exactly the purpose of your idea to move the frets/change the scale? Better intonation? How would that be achieved if you move both the frets and the bridge? You still just slide the whole enchilada (with individually increasing distances) to wards the nut. Is there any other purpose of this whole exercise? Please be more specific with your thought and we might have a better chance at grasping your ideas.

And please explain what you mean with the comment about the Gibson scale being "split" into three sections. The frets on vintage Gibsons are moved individually in exactly the manner that was suggested in your original post.

[kingfisher]

Firstly let me say that i put this out for discussion, i could be wrong on this. I usually am , but this is how I see it.

Suppose we fret a guitar, and first we need to work out the fret distances. OK, bit of maths. start with the scale length and divide by a number to find the first fret. Take away the first fret from the scale length and divide by the number again for second fret... and so on

So what number shall we divide by? lets say 10.

Oh dear the frets are way off, so we move the bridge a bit (a LOT) to try to cover up our error (i.e.using the divisor 10) No matter what we do , even if we get the 12th fret right , the rest of the frets are still wrong.

Bugger. Shoot the board and start again. lets try 15. not as bad but still nowhere near. Get the n0. 5 plane out again (board is getting a bit thin by now) and try 17.818. This is the number most of us use by the way.

Now it is still not quite right (we still have to move the bridge, but not by much, so we live with it.) What we need is a magic number to divide by that doesn't need bridge adjustment - that is the holy grail.

Of course due to string gauges , maybe different scale lengths, guitar or bass etc, this is not possible, but maybe we can get closer than the legendary 17.818. and get better intonation.

That's it.

I didn't know Gibson did this already, and my comment about "three sections" came from post #9 above.

[WezV]

Oh dear the frets are way off, so we move the bridge a bit (a LOT)

the issue isnt with the maths.

we dont compensate the bridge to make up for dodgy fret placement or dodgy maths.

we compensate the bridge because you have to bend a string slightly to fret it.

some players prefer a higher action so have to bend the string further sharp to fret it and therefore need more compensation... how does screwing up the fret placement to compensate allow for players different set-up preferences :?

[erikbojerik]

And the original suggestion would only change the de facto over all scale length.

No It wouldn't. Thats the whole point.

Yes, it would.

[kingfisher]

OK then I'm wrong - and apparently Gibson are too. Brave man.

[WezV]

well i stayed out the gibson thing so far - but my own research has not found the differences crusader talks about... on new or vintage instruments.

I am well aware crudaser feels gibson have some secret technique for this, but i have not found it. They have however used slightly different scale lengths over the years and even had some years of known faulty fret placement - so its not unusual to see different measurments on them

you may want to look into historical fretted instruments if this interests you though.

From about the 16th century the fret position was found with the 'rule of 18' Divide you scale length by 18 and put your fret there. Some claim that these old instruments didnt need compensation at the bridge as the simplified maths was actually compensating each fret postion. worth noting that not many of these instruments play nicely in tune across the whole fretboard, and they have very little room for adjustment

[kingfisher]

Just one more thing - was I right to use a no. 5 plane, or would it have been better to use a no.7?

Lots of tricky questions in this guitar thing.

[westhemann]

Lots of tricky questions in this guitar thing.

Not really..it's all in the research.

There ARE alot of guys trying to reinvent the guitar though,and most with no success.

Gibson is not wrong,but gibson is not doing anything too wierd..they are just sliding the bridge back enough to keep the saddles closer to the middle of the TOM bridge,which otherwise may not have enough..nothing fancy,just the same old compensation thing...

The whole PLEK thing is only to recrown the frets in "the exact right spot" after fretting variances "might" put the crowns slightly off kilter...still follows the same original mathematical formula.

[SwedishLuthier]

Just one more thing - was I right to use a no. 5 plane, or would it have been better to use a no.7?

Lots of tricky questions in this guitar thing.

A bit grumpy eh? Just because your idea wasn't received with standing ovations by the crowed here there is no need to act like that.

You have had a few really experienced guys trying to make you see that your idea was really not neither new or addressing the problem the right way or addressing the right problem. And if you choose to receive those opinions like above it is not beneficial for either you or this community. Anyway, please do try your idea and get back to us with the result. I love being proved wrong

[kingfisher]

Ha -No not grumpy at all - quite the reverse in fact - It was meant to be a lighthearted attempt at humour. probably lost in translation.

As I have stated more than once, I put this idea out for comment in the first place, all which have been gratefully received.

Thank you.

Having said that I still think if I put 25 into StewMacs calculator, and have to make it 25.2 to make it work, there is something wrong somewhere.

Thanks again to all for your input, at least it has got us thinking !

John

[WezV]

put 25 in the fret calc and it will determine all the fret positions for a 25" scale length without error.

but the actual string will need to be slightly longer- 'compensated' , especially on the bass side. Again this is not an error in the maths but caused by the physical act of pushing the string sharp as you fret

therefore If you wanted to have a compensated scale length of 25" you would need to put in a number slightly less than 25 to allow for the extra string length needed for compensation. This is what gibson has done with its 24 3/4" compensated scale length, the actual sale length is usually around 24 9/16" - but they still refer to it as 24 3/4". Very few other guitar companies do it this way - most will just tell you the number that gets used for the maths as the scale length - and since every string need slightly different amounts of extra string length it makes sense to do it this way

[Woodenspoke]

Just one more thing - was I right to use a no. 5 plane, or would it have been better to use a no.7?

Lots of tricky questions in this guitar thing.

I didn't see the humor anywhere on the OP, maybe why I let this play out without my comments as the whole thing seemed quite silly to begin with...

I am not sure where this plane gem I am responding to fits in to scale but here is my two cents. Planes come is all sizes, the longer planes such as a #7 or #8 are used for joinery hence why they call them jointer's. These will give you the flattest surface provided they are setup properly. If you have a #7 thats what I would use over a #5. If you are removing wood a shorter plane with a slight curve on the blade is best such as a #4. A #5 is the most common plane sold and you can find these at the cheapest prices used, they span between a #7 jointer and a #4 smooth plane and were bought to try and save money rather than buy two planes. They have there uses but were not really suited to either task.

[SwedishLuthier]

Ha -No not grumpy at all - quite the reverse in fact - It was meant to be a lighthearted attempt at humour. probably lost in translation.

OK, glad to hear that

[Crusader]

Still going with this are we?

Has anyone actually tried the idea proposed by kingfisher? I find it works fairly accurately at least for the Gibson 24 3/4 scale

Taking the nut to 22nd fret measurement for 24 3/4"

Gibson 450.1mm

Standard 452.2mm

The Standard calculation is as follows

(628.65-(628.65/1.059463094^22) = 452.2

Rearrange the formula

22(need a square-root symbol here) (628.65/(628.65-452.2) = 1.059463094

Replace 452.2 with 450.1 and you get your new number

22(need a square-root symbol here) (628.65/(628.65-450.1) = 1.05888229

(628.65-(628.65/1.05888229^22) = 450.1

Calculating with this number 1.05888229 for all the other frets gets very close to the Gibson 24 3/4 scale

Measuring to the 12th fret, doubling and using it in the standard formula is no-where near as accurate

This may not be what you want though because as I said the Gibson scale is a bit sharp above the 12th fret on the first and second strings

Finding what you want may take a lot of trial and error

Happy hunting!