Official Luthiers Forum!

I'm not sure what would qualify as ,"real" data. I am a repair guy and could post every Picture I have of a cracked bridge front or saddle leaning forward at a crazy angle but I don't think it's necessary. A quick search on the UMGF for "bridge saddle crack" or any combination of terms like those brings up 30+! pages of results.

If there is one thing I think we can all agree on, especially repair people, it's that big companies producing hundreds of guitars a day, are no proof of good building methods.

I believe Tangent is what I intended. The idea is what would minimize the change in string length as the strings or saddle were raised or lowered. Strings move on a roughly scale length ark with the center at the nut. A Tangent saddle would minimize length change as compared to one leaning dramatically forward or back. No?

Real data would be resolved "warranty claims" -- from makers that have properly made bridges and saddles that are snug -- you see a lot of these, I believe you - when I did repair work I saw none. The thing is, Martin over time will make changes that prevent warranty problems. Move the X, add the Popsicle brace, cantilever the neck block, don't spray over the pick guard, replace the solid truss rods with adjustable etc. All I am saying is that I personally do not have any indications of an epidemic of failed bridges do to perpendicular saddle slots. And I think maybe the wider saddles I use (1/8 -1/4 which have a larger foot print less tendency to tilt forward -- maybe not?

I read that some think 2 degrees is a good tilt angle and others that say 9 degrees is the magic number -- is more or less better?

_________________
Ken Cierp

http://www.kennethmichaelguitars.com/

Or has strings that are too thick.

I've been using the back angled saddle for years, although it was not my idea, but Rick Turner's. It's really no more trouble top make than a 90 degree slot: I just put a 9 degree wedge under the bridge when I rout the slot. If' you're routing the slot on a bridge that's already glued down put the wedge under the router.

Angling the slot back reduces the tipping force on the top of the saddle. In theory, if the saddle leans back at an angle that bisects the break angle, there's no net tipping force. It also maximizes the the downward force of the saddle on a UST, which is one reason Turner gave for using it. Having repaired a number of bridges with split out fronts back when I doing more repairs (none on my guitars) it seemed to me that anything I could do to avoid that issue down the road would be a good idea.

One issue with using the back angled slot is figuring out where the bridge should go when it comes time to glue it down. you can't just set it so the front edge of the slot is a little behind the 'theoretical' point, since the back angle will move it further back. These days I tend to position it so that the back edge of the slot at the high E location is on the theoretical point, and that works out pretty well.

I think that's exactly right.

Believe it or not, I had to interrupt my internet play to meet a customer in my shop with guess what. A 50's Martin w/ a cracked saddle slot! I simply can't explain how others apparently don't come across this problem as frequently as I do.

Maybe I do have the wrong Idea. Perhaps saddle slots should be built tipped forward so as instruments collapse over time, intonation remains perfect. Why not?

I sense I'm not getting any converts here. That's ok.
I'll just keep fixing the broken saddle slots that come my way. The slots I fill and recut, or rout on replacement bridges I make, have an almost unperceivable 1 1/2 degree back angle that I think makes a significant improvement.

"I sense I'm not getting any converts here."

Some of us have been tilting back saddles for a number of years now. It can make for a stronger bridge and add a little bit of compensation adjustment for different set ups. Like compensated nuts, it is not going to suddenly make every note play perfectly in tune. But if you can get a little bit closer, for so little additional effort, why not?

I back angle 10 degrees. When I show customers, 100% say that makes a ton of sense. So there is possibly some product differentiation gain from it when you are trying to pull a factory guitar buyer over to handmade. Show them the little things, soundports etc., that you can do that most factories won't mess with.

In my experience, 3 factors seem to be at play when I've seen cracked bridges come in - 1) vertical grain saddles provide poor strength for the soundhole side of the bridge, or for possible stresses at pin holes - rift sawn seems a far better option, 2) some (especially pyramid bridges come to mind) have precious little material at the high E end of the bridge, between saddle slot and bridge side of the saddle, 3) humidity swings seem to have the ability to contribute to both of these - for example, I had a 0-16NY that was 50 years old come in with a broken original brazilian bridge, after it spent 1-year in a dry climate. It lived 50 years without that problem.

Andy

I agree with VG not helping bridges survive Andy. I wonder how many degrees of rear tilt it would have taken to spare your rosewood bridge.

Here's some justification for a slanted bridge slot if installing a USP

http://www.fishman.com/files/advanced_u ... lation.pdf

_________________
Ken Cierp

http://www.kennethmichaelguitars.com/

Sorry, fellas...long day in shop and no time to respond.

This shop does both 90 degree and angled saddles, so I have no real objection to either approach. As to the word tangent as applied to this conversation might not be congruent with the narrower geometric definition, but I understood Mr. Farmer's point. What I do not understand is how one would ever have a conventional saddle where each string was not tangent in the sense of that usage.

A few questions that I was asked yesterday when I brought up the conversation we are having:

- How does one determine the angle which cancels all side loading - which seems the entire point of the exercise - for realistic break angles and for the lifespan of the instrument, or at least between neck resets)?

- Given that the standard for a saddle fit in this shop is a near airtight, piston-like fit that still allows finger-grip removal without any apparent movement under load, and that so-called sticktion (I did not make that up...a certain engineer I know used the term to differentiate between static and sliding friction) is more dependent on friction between components (fit & finish) that side loads between the saddle and slot, does side load reduction have a meaningful effect on under saddle pickup performance?

_________________
A constellation only takes shape when one maps the whole.
- Beth Brower

Again, the 'best' angle for the saddle slot that minimizes the tipping force is the one that bisects the break angle. You really don't need all that much break angle:, I've done tests with a break angle at around 6 degrees that showed no audible or measurable problems with string contact. Anything more than that is gravy at best, and harmful if the saddle is not angled back, at least on an acoustic guitar. In practice, a nine degree back angle implies an 18 degree break, and anything close to that will give only a small tipping force one way or the other.

The greater force on a UST from a back angled saddle follows from the same vector analysis that points up the tipping force. Assume the string touches the top of the saddle at a point, and tips down behind at some angle from the horizontal that we'll call the 'break angle', and that it's fixed somehow back there . Assume also that there's no friction over the break point, so that the tension on the string is the same on both sides. The back string has to be pushing down on the saddle, otherwise it would not make that angle. Similarly, from the point of view of the back string there has to be some force over the saddle top to make the string angle down to the horizontal. These two add up to a force vector that bisects the break angle. If you use a vertical saddle then the forces on it are the downward component of the vector, pushing it down into the slot, and the forward component trying to tip the top of the saddle forward and split out the front of the slot. The down component is reduced compared with the bisecting vector. If the saddle tips back at the angle of the bisector, there's no tipping force, and the down force on the bottom of the saddle is greater. USTs tend to like static pressure downward, so they work a bit better with the back angled saddle.

One thing that tipping the saddle does that MAY not be good is that it reduces the force that the saddle exerts on the top that is perpendicular to the top.
I think I have this math right. So there is more perpendicular force on the UST but less perpendicular force on the soundboard.
Dave

I'm pretty sure that the vertical component of the force that the strings exert on the top through the saddle is the same whether the saddle is vertical or angled back since the vertical component is determined by the break angle of the strings (relative to vertical) on the back side of the saddle, assuming the break angle is the same in both cases.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Ok sorry. I'm having a little trouble with the angle logic so I will rethink it. Thanks
Dave

"One thing that tipping the saddle does that MAY not be good is that it reduces the force that the saddle exerts on the top that is perpendicular to the top.
I think I have this math right. So there is more perpendicular force on the UST but less perpendicular force on the soundboard. "

We have to be careful to distinguish between the static and dynamic forces here. The static down force on the UST is increased by the use of a back angled saddle, as is the static tipping force on the saddle top. The dynamic force, the actual string signal on the bridge and top is the same no matter what the saddle tilt (although the angled saddle might increase that on the UST as well); neither the up-and-down 'transverse' or the longitudinal 'tension' signals are affected. These are much lower in any case than the static forces: typically the transverse signal has an amplitude of around 5%-10% of the static tension of the strings, and the tension change force averages about 1/7 of that, so around 1% more or less of the static tension.

I've got to respectfully disagree with you. If some of the force is spent trying to "twist" the saddle out of the slot, it is wasted energy that is not going into tone production. That's the main reason I've always angled mine back. Not to prevent bridge cracks, though that might be a nice bi-product. The closer to perfectly bisecting the break angle one gets, the closer one gets to having all of the string force going into downward pressure rather than any torsional force.

My wording was ambiguous. What I was referring to is the vertical downward force on the top rather than the force against the bottom of the saddle slot. The vertical downward force on the top is determined only by the string tension and the break angles on the front an back sides of the saddle and not by the angle of the saddle. It's the same whether the saddle is angled or not. The force on the bottom of the saddle slot exerted through the saddle, though, is determined by the angle of the saddle. I agree that the maximum force on the bottom of the slot and zero torsional force would be obtained with the saddle bisecting the angle between the strings on both sides of the saddle.

Regarding tipping the saddle, Dave said above that "So there is more perpendicular force on the UST but less perpendicular force on the soundboard." I agree with the more force on the UST part, but not the less force on the soundboard part. The force on the soundboard stays the same, but the force on the UST increases as the angle of the saddle approaches the angle bisecting the angle between the strings on both sides of the saddle. That's because the bisecting angle corresponds to the highest combined force of the string pull in front of and in back of the saddle.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Sorry for misunderstanding your post. Yes, I agree with you on all aspects. I think once the bridge is introduced into the equation, the downward force on the top is the same regardless of the angle of the saddle. (assuming break angle is the same).
Though then a new and improved question can be introduced into this: The further spread apart the saddle and pins are, the more torsional force is introduced. The closer together, the more vertical force is on the top. But that's kind of irrelevant to this conversation.

No problem. Looking back at it, I could have misunderstood my original post. It wasn't as clear as it could have been.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Paul Woolson wrote:
"The further spread apart the saddle and pins are, the more torsional force is introduced. The closer together, the more vertical force is on the top. But that's kind of irrelevant to this conversation."

When I did my string height and break angle experiments, I looked at the torque on the top when I tightened up the strings for three setup conditions. These were:
Case A - strings 11 mm off the top, with a 25 degree break angle,
Case B - strings 11 mm off the top with a ~6 degree break angle, and,
Case C - strings 18 mm off the top with a 25 degree break angle.

For this test I used a Classical guitar, and varied the string tie on the tieblock to alter the break angle, but a guitar's a guitar for a' that, and the results should hold for a steel string as well. Anyway.

I used two tests to look at the top displacement. I stuck a laser pointer to to bridge, and measured the displacement of the spot on the wall. I also used a dial gauge to see how far the top moved 5mm in front of and behind the saddle. Basically, the top did two things when the tension went on: the bridge pushed the top in a bit, and it also rotated around a fixed point. There was no significant difference in the amount of rotation between Case A and Case B: when the saddle height was the same the bridge rotated through the same angle, so I'd say the torque was the same even though the break angle varied. On the other hand, with the lower break angle there was less drop. With the taller saddle the bridge rotated more, and dropped further.

What was interesting was that the location of the fixed point around which the bridge rotated (which I'm calling a 'centroid') changed when the break angle did. That is, the steeper the break angle the closer the centroid was to the saddle. It occurred to me the other day that this change in the location of the centroid could alter the forces at the back edge of the bridge, making it more or less likely to peel up.

Anyway, the bottom line is that, so far as I can tell, the total torque on the bridge depends on the string height off the top, and not on the break angle.

That's an interesting experiment. I like the idea of the reflected laser light on the wall. It seems like that would be a sensitive indicator.

I would expect that Case A would produce at least some increase in torque on the saddle compared with Case B because, with the steeper break angle, the downward force component of the strings behind the saddle is increased and the horizontal component that counteracts the tension between the top of the saddle and the nut is decreased thereby giving a net increase in the horizontal force at the top of the saddle toward the nut. Maybe for the two cases with equal saddle height, the deflection of the top was already maxed out or very nearly maxed out in Case B? Or maybe the difference in torque between 6 and 25 degrees was not enough to detect even with the reflected laser.

Case C makes sense since the taller saddle provides a bigger lever arm.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

After an hour spent being tutored on something called "The Method of Sections" , my memories of why I elected not to study engineering have come flooding back. I cannot relay all of the sketches and thought experiments and doodles made, scanned, transmitted, and misunderstood, but what I can say is that I (think) I agree with Mr. Carruth that the force and the moment that the rest of the guitar will need to counter from string tension on the bridge (or 'reacted' to use a new-to-me word) is a constant for a given set of strings and saddle height above the top. Additionally, I (think I) can also say the way that any individual guitar reacts that force and moment is unique to that instrument's configuration of bridge, saddle, string, and structure.

In other words, making the bridge longer or wider, or thicker, or moving the bridge pins closer to or further away from the saddle, or eliminating the pins entirely has no effect on that total force and moment applied to the rest of the guitar by the strings, but will change where and how they are reacted.

I'm still not precisely sure why this is so, but I can now draw what appears to be a potato with some straight and circular arrows that look convincingly like what I just got done discussing.

On a separate note, is it all instructors of instrument building that feel the need to answer every question with one or more questions, or just the difficult ones? The last question was rhetorical - I know there is no good answer.

_________________
A constellation only takes shape when one maps the whole.
- Beth Brower