Official Luthiers Forum!

I have been asked to refinish the top and re-glue the bridge on a classical guitar. The scale length is 650mm. I just realized that the classical bridge already has a compensation of 1.0 mm on the bass side due to the angle of the saddle slot in relation to the front of the bridge. I have not worked much with classical guitars but what I am aware that most bridges do not have an angled saddle slot and that the entire saddle is typically compensated 1.0mm. My question is should I position the bridge so the entire saddle is at 651mm or should the treble side of the saddle be at 650mm and the bass side at 651mm?

A quick answer is 2mm set back for the high e.
3mm for the low e.
I use a 3mm wide saddle & find I sometimes have to
adjust for every single string.
Also I cut 1mm from the nut end of the fretboard.

Go http://www.byersguitars.com for details about this setup.

Mike

_________________
Mike Collins

Much of that depends, too, on how the nut is set. If you aren't going to do a full comp job, then I wouldn't use a slanted slot. Look at David LaPlante's saddle. I copied it. I use a 1 mm set back, but that's with a .5 mm nut set toward the bridge, so technically it's 1.5 mm set back. I use a 2.5 mm bridge, with almost no comp on the "e" sloping to about 1.5 mm on the "g", then jumping forward to virtually no set back on the "D" and sloping to about 1.5 on the "E". It works almost perfectly. Now I'm compensating the nut too, and have been amazed at how well the guitar plays all the way up the fretboard, compwise.



Not a great photograph, but you can see the saddle.

_________________
Waddy

Photobucket Build Album Library

Sound Clips of most of my guitars

Hi Mike,

Not to derail the thread but I just had a read there. Do you have any idea how those "rules" change with scale length? I'm building a small classical with a 480mm scale length and I don't know how I should handle compensation. I'm too inexperienced to trust my instincts and I'm petrified I'll string it up and won't be able to intonate it!

Cheers,
Nick

I'm confused Waddy. If you use a 1mm set back and then move the nut forward .5mm towards the bridge, then isn't it technically only .5mm set back, not 1.5mm? I must be missing something. Wendy

No. The frets haven't moved. They are still in the same position, so it's still 1.5 mm's and the Nut forward by 0.5 mm.
I use the 'old fashioned' 2 mm set back. Not one single player has pulled me on the intonation.

Nick:
The most complete treatment I know of on this is in the Gore/Gilet books. I'm sure Trevor Gore will show up at some point, or you could just search on his name. They're expensive, but worth it.

To summarize with desperate brevity: pushing the strings down to fret the notes stretches the strings. This increases the tension. The tension increase depends on the string material/construction, length, and the displacement, but not on the initial tension. A string that's near it's breaking point won't see much pitch change, because the increase in tension will be small relative to what it's already carrying. A relatively slack string will see a much larger pitch change.

Fretting at the first fret raises the pitch some, and as you go down it goes sharper. Moving the nut toward the first fret flattens all the note on that string by about the same amount: it's the same as shifting all the frets back. If you shift the nut forward so that the first fret note is in tune, the second fret will still be a bit sharp, all else equal, and the 12th fret will be more so; just not quite as sharp as it would have been.

Moving the saddle away from the frets drops the pitches of the notes too, but it has very little effect on the the low positions, and much more as you go up. If you leave the nut where it's 'supposed' to be, and shift the saddle back, you can get the intonation to be correct at, say, the 12th fret, but the first fret intonation will still be sharp, and all the notes above the 12th fret will be flat.

Some combination of nut and saddle offset can usually be found that will yield excellent, if not 'perfect' intonation at all the frets. Trevor gives all the calculations in his book for finding these. Alternatively, you can set up a dummy fretboard on a board, with a movable nut and saddle, and shift things around until you get the offsets that work.

With strings as short as you're contemplating, you might find you'l need a lot of compensation at both ends. It will be problematic,too, in that even a small amount of bending of the string will cause large pitch changes. Even small differences in finger pressure when fretting will be enough. I built a steel string once with a 20" (508mm) scale length, and it took some getting used to to get it to play acceptably in tune. You may find that some special stringing will be in order, and you may want to investigate that before you cut too much wood and lock yourself into some problems. A little study and experimenting now is time well spent.

Thanks so much Alan! I'm still new to forum decorum but I didn't want to annoy anyone by derailing this thread entirely so I've replied on the thread in which I've been getting advice on this build. Hope that's ok with you.

http://www.luthiersforum.com/forum/viewtopic.php?f=10101&t=41311&p=547764#p547764

Thanks, Alan, once again, for saving me a lot of typing. There's not much more to add.

To do it all "properly" for "nylon" strings is a lot of work (properly = optimised, "nylon" = all the different types of plastic that get used in classical strings, which you have to measure the properties of yourself). However, there are a few sub-optimal methods which give a pretty good result for considerably less work.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Even if you stick with 'real nylon' you're dealing with a pretty wide and varied class of chemicals, and the string makers have gotten 'way beyond that. Also, some of the 'constants', such as Young's modulus, aren't really constant, but can vary pretty widely depending on things like how the string was made and how much tension it's pulling. In the end, a dummy fretboard might just be the easiest way to get the info you need to a reasonable accuracy.

I can't quite visualise the dummy fretboard test. If I cut the fretboard for the 48cm scale guitar, could I just use that with a few frets, two e strings, a temporary nut and a cheap electric bridge screwed to a plank? Would it make enough noise to test intonation? Or have I got the wrong end of the stick?

The "Dummy Fretboard" sounds interesting. Can someone direct me to some construction instructions or photos?


Sent from my iPhone using Tapatalk 2

Nick, I think you've got the right end of the stick. I use an under saddle piezo pick-up and plug straight into my computer. You can use a Strat type bridge arrangement if you have a good mic and position it close. Even for saddle-only intonation you should do all 6 strings. The G, especially, can be quite different from the rest. For "nylon" strings, you need to let each string stretch out for 3 to 5 days before you take any measurements. Pick a brand of string that you/your niece can stay with, because there are significant differences between brands.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Thanks so much, Trevor, that makes perfect sense and will give me some serious peace of mind! I'll order the strings (Pro Artre Extra high tension - my attempt, as well as tuning up to [open] G, at making up for the low tension/tiny scale) and cut the fretboard asap so I can do this all well in time.

Might be quite fun to make a dummy neck with rails for the bridge to slide along so it can be used for all scale lengths! Maybe clamping the fretboard in place from the sides. Could have interchangeable classical and ss bridges and headstocks. I might save that for when I'm not trying to make a guitar in time for Christmas.

Such as? Setting the saddle at 652 mm? Giving the G string an extra mm? I won't get your book until Christmas so a quick answer would be much appreciated.

Cheers

_________________
http://coneryguitars.com

Kill your television!

The simplest method is to figure out how much compensation you need using a string test rig, with the strings you want to use set to the action and scale length you will use. There is no "standard" answer because different strings behave (sometimes markedly) differently and people use different set-ups. Measure the saddle compensation needed to correct things at fret 12 then either use that or divided the saddle compensation by 2 and put half on the saddle and half on the nut. Hardly optimal, but better than typical and gets you started.

Roll on Christmas!

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Just for the record, I set my compensation at 2mm (652mm) with no adjustment at the nut end. Then "double" compensate the saddle as per Waddy's description.

Thanks for that, David! What about for a 48cm scale? Sorry, only kidding... I'm going to make a test rig!