MIMF

Back in September, Alan wrote this about that (see below copy and paste). I want to shout out thanks, because I just went through a neat experiment with my red guitar. By a quirk of design, the first tail piece I put on resulted in a string break angle of 20 odd degrees. Calculated 53 lbs of downforce, which I reckon is way too much for a lightly brace archtop (mind you it stood up to it for a few days with no damage). Tone was bright and arch-toppy, very direct, lacking in bass response somewhat. I just built a mount to raise the tailpiece by around .7" from where it was, reducing the break angle to around 15 degrees, and bringing the downforce to around 35 lbs, coincidentally the same angle and force as I measured on my 1946 Epiphone and my 1957 Hofner. So very traditional. Improvement in tone is quite dramatic, a lot more upper harmonics give a shimmery tone, bass response is improved, overall volume is up. Very happy with this. To go higher would require a modification to the mounting method, which I will probably do anyway, and then going up another 3/8" to get 10 degrees will be possible. Once I get where I think is optimum, I have the option of modifying the tailpiece itself to improve the mechanics of the deal.

Anyway, thank you Alan, I appreciate your wisdom and your willingness to share it.

Here is his post from September:

I've done a few experiments along those lines. Back when I started making archtops I thought that more break angle equated with more tone, so I tried to maximize it. I made several guitars with wooden 'hook' tailpieces: from the side thay looked like a letter 'L', with the short arm pivoting on the side. The line of the strings from the bridge top takes an angle that goes to the pivot point of the tailpiece; with that low on the side you can actually get the strings top effectively go 'through' the top.

On one of them I tried adjusting the pivot point, to see what would happen to the sound. It was interesting. The sound was fine with the pivot all the way up near the edge of the box, and stayed good for a while as I moved it down in small steps. At one point, though, as I got it pretty far down on the side (no, I don't remember HOW far) the tone just died. Raising the pivot point brought it back. so there does seem to be such a thing as 'too much' down bearing. Since then I've learned a couple of things that help make sense of that.

One was brought to my attention by a fellow named Joshua Gordis, who taught at the Naval Post-Graduate Institute. He got in touch to ask if putting string tension on a flat top guitar altered the resonant modes of the top. WE made a few measurements and found that, within the limits of our technique, it didn't. He's been using the frequency change in columns under load as a non-destructive test of truss structures in things like aircraft. A column that s fixed at both ends will have a fundamental bending resonant mode at a certain pitch. If you load the column eventually it will buckle, and at that load the bending mode frequency goes to zero. The neat thing is that the relationship between the mode frequency and the load is linear: at half the buckling load the frequency will be half what it was with no load. This means you can do a simple test with a fairly light load and predict the failure load without risking damage to the structure. On something like an airplane you can find out which parts are likely to fail first, and which ones are over built, so it gives you a way to optimize the structure.

When he told us about this, I checked out the modes of an archtop, and, sure enough, the 'main top' resonant mode DOES drop in pitch as you tighten the strings. It doesn't usually drop much, but it does change noticeably. I did not try increasing the break angle over the bridge (lowering the pivot on the side) to see just when the sound went to pot, but, in a sense, I didn't feel I needed to. The point is that it DOES drop the pitch, which should affect the tone.

The second thing I learned was that you don't need a lot of break angle to get the string to drive the top. Basically, the transverse force of the string that pushes the top in and out to produce sound is well defined if you know the string tension and the angle it makes at the bridge before you release it after pushing it down to pluck it. It's quite small, since the angle is limited: after all, once you've pushed the string down to where it touches the frets you can't go any further. As the string vibrates the angle it makes at the bridge goes up and down, but the 'up' angle will never be greater than the initial 'down' angle. Thus, as long as the string makes a break angle at the bridge that is greater than that down angle, it will stay in contact with the top of the bridge throughout it's vibration cycle, and all of the force produced will be transmitted to the top. Even playing very hard very close the bridge you're unlikely to push the string down at more than about a five degree angle, so in theory that's all the break angle you should need.

In practice there can be a difference between theory and practice, however. You never push the string straight down; there's always some sideways force. This can cause the string to roll on the top of the bridge, and that little sideways motion can play hob with the way it vibrates. Basically, it can introduce other frequencies into the sound, causing buzzes or whatever. Still, you don't need a whole lot of break angle to get all the tone from the string that it's capable of delivering.

So; you don't need a lot of break angle to get sound, and too much can hurt. I can't speak to the limits on either side with any certainty: they probably vary a lot depending on things like how thick and highly arched the top is, and so on. My own inclination these days is to stick to pretty minimal break angles, since, aside from the tone effects, download on the top has no good structural outcomes.

I used a ball and socket joint on the tailpiece of my 'Winter' archtop. I made an ebony post with a ball on the top that fits into a hole in the tailblock. The post has a brass tubing sleeve on it, and this telescopes into a thicker walled stainless tube in the block so I know there's a good fit that won't go out of round. I inset a threaded rod into the end of the post, and threaded insert in the block. so it's easy to adjust the tailpiece height. The ball end fits in a cup on the tailpiece; once the post is screwed in you can't knock the tailpiece off by accident. The whole thing was derived from the tailpiece attachment they use on viols da gamba: a notched stick dovetailed into the tailblock that fits through a square hole in the tailpiece. You can easily find the notch for the tailpiece that gives the best sound for the strings and setup you're using. There really is nothing new under the sun..

Thanks for the post. Some place in the 'Lord of the Rings' trilogy, Gandalf says something to the effect that it's nice to not be wrong all the time.

Alan Carruth wrote:Thanks for the post. Some place in the 'Lord of the Rings' trilogy, Gandalf says something to the effect that it's nice to not be wrong all the time.

A great quote that's relevant to parents and teachers too. I've learned a ton from this forum and you in particular. I'm close to needing this information myself so this is good timing.

Thanks for hanging out with us newbies.

-Eric

So, I have read about experiments like this before (it was probably about Alan Carruth's experiments, but I am not sure). Anyway, every time I read this, I can never visualize exactly how the tailpiece looks so that the strings can go below the top plate. Are there any pictures of this arrangement, anywhere? If I see how it is done, it will clear up a lot about the article. Thanks.

If you consider that a typical archtop might have a bridge height of 3/4" to 1" (on the low side) and a 3/4" arch from the side to the top of the belly, and around 9 3/4" from the bridge to the tail, you can imagine that if the arch of the belly extended out towards the sides, and the string attachment was as low as possible, right on the edge of the top plate, the notional "string line" could go below the top plate. Of course, what really would happen is the notional line (the straight line force vector between the top of the bridge and the string attachment point) isn't the actual path the tailpiece would describe. It could be arched to accommodate the flow of the top and still transfer the force between the two points.

So, I dug out my Benedetto recipe book yesterday to see how Sacconi tailguts work (changing my whole tailpiece design as a result of this discussion) and sure enough, he addresses this very topic. I missed it before. He says that the idea angle is 13 -14 degrees for a balance of tone and power. He says that too little downforce/angle reduces power and richness, while too much mutes the instrument - exactly what I found.

I remember reading that in the book, and I do try to build to those parameters, but I also recall that in Benedetto's DVDs that accompany the book, he did a string break angle experiment. If I recall correctly he installed some kind of a metal adjustable angle tailpiece on one of his instruments and changed the angle several times while plucking the strings. His conclusion from that experiment, was that the string break angel didn't really matter to the sound quality. Now, I am not sure how scientific that was, or what the range of angles was that he used (maybe they were within the reasonable range). I was just looking at this topic this week, because the latest issue of Fretboard Journal had part 1 of an interview with Ken Parker, and he mentioned something about minimizing the break angle of the strings, and building the top a lot lighter to increase responsiveness and player comfort. So, this is a very timely topic.

My conclusion so far is that there is a wide range of angles that are "fine", maybe from 8 to 15 degrees, less than 8 you may have strings losing contact with the bridge, more than 15 (recall that I had around 20) the sound got a bit dead. I don't think that vibration transfer is all that is going on, I think that vibration transfer is fairly unitary after sufficient angle is achieved. I think that pre-loading the top is part of this that I don't understand, at this point I think that minimizing the pre-load is desirable and too much pre-load dampens the top and doesn't let it vibrate as fully. At 20 degrees I had close to 55 lbs of preload. That small child standing on the guitar top for a few years cannot be good for it. I have a resonator guitar that I lowered the action height to quite low, it has around 5 degrees string break angle, and it did not lose tone at all, it sounds exactly the same as when it had a greater string break angle. So I tend to conclude that enough is enough when it comes to string break angle, more doesn't particularly help.

Brian Evans wrote: So I tend to conclude that enough is enough when it comes to string break angle, more doesn't particularly help.

Mr. Carruth has been posting that for years, but mostly relating specifically to flattops.

Download has one effect that is fairly easy to measure; it drops the pitch of the 'main top' resonant mode.

Back about twenty years ago (!) I got a letter from Joshua Gordis, who was a professor of engineering at the Naval Post Graduate Research Institute, asking if I'd ever seen a change in the mode frequencies of the top under load. This is related to what happens when a column is loaded. A column that is fixed at both ends will have a fundamental resonant pitch that is relatively easy to find. You could imagine that if you stretched it in some way that pitch would rise. It turns out that if you load it the opposite happens; the pitch falls.

The interesting thing about this is that the change in pitch is linear relative to the buckling stress. if you load the column to the point where it buckles the fundamental resonant pitch drops to zero. If you load it to half of the load that will cause it to buckle, the pitch will be halfway between zero and what it was when the column was unloaded. This makes it a useful non-destructive test, in particular for truss structures, such as the tubular tail of an old fashioned helicopter. You can make up the structure, load it to half the design load, and check the change in resonant pitch of all the elements of the truss. Anything that drops by more than half is too weak, and anything that drops by less than half is too strong and heavy.

I checked this out on a plate by cutting a circular piece of Styrofoam, sticking a broken rubber band around the edge with double stick tape, and checking the modes. Then I put an un-broken band on of the same size and checked them again. Sure enough the 'ring' mode pitch went down.

We checked out several flat top and Classical guitars in the shop to see if there was any difference in the lower order mode pitches with the strings slack as opposed to tight, and found no appreciable change. Apparently the compression in front of the bridge is balanced by the tension load behind it. The very slight changes we saw could be attributed to the distortion of the top due to bridge torque, which changes the curvature of the top a bit and thus alters the local stiffness.

Archtops were another matter. Putting string tension on them dropped the 'main top' resonant pitch noticeably, and also tended to raise the 'main back' pitch a bit. The drop in the top pitch is probably related to the downbearing, although I don't remember that we looked at that. The back is being stretched when there is a string load on the guitar, so it's not surprising that the back pitch would go up.

It might be useful to do an experiment to see how the pitch drop in the top relates to the download, and whether the pitch would indeed go to zero at the point where the top collapses. This is not exactly a simple column load, after all, so you can't expect it to act like one. On the other hand it would be useful to try to relate the pitch drop to something useful like that, or the load that causes the top to choke. I've seen all sorts of vector diagrams in the violin making literature that attempt to determine the 'ideal' bridge height and break angle geometrically. Some of them actually make a sort of sense. Still, that's all speculation; maybe we can get some data? Of course, you would not want to use an actual guitar for this, but it should be possible to make up a simpler model that would give some useful data.

My poor brain is currently trying to figure out the force vector analysis for a Benedetto Sacconi tailpiece. My challenge is mounting one in a not-stupid way that gets me a string break angle that I can live with and is robust and pleasant looking. The issue with my guitar is the bridge is 7.5" from the edge of the tail, the bridge height is 1.1" above the top, and the top is .75" above the edge of the tail. So I have a short distance to work out the break angle in. Right now I just raised the pivot point of the tail piece mount by around .75", but that plinth is just cantilevered off the top and is already failing. Proved the point of changing the break angle, but not a solution. I am going to play with the design of the Sacconi tailpiece by varying the length, the position of the saddle, the height of the saddle and putting a shim under the saddle to raise it up off the top. The saddle is the little pivot point that rests on the top of the instrument and everybody hates for putting a dent in their expensive top.

I have a backup plan for when this doesn't work, basically putting a post up the rim and hooking the tailpiece to it, but I think that can look clunky. It will probably look a lot better if I end up doing it, perception changes based on needs must...

I am going to spend a few hours relearning vector analysis. I knew it in first year college and I am at least half as smart now as I was then...

Brian

I do not think You would need vector analysis just to learn how much to raise a top. Simple trigonometric estimates on the triangle spanned up by the desired reduction in bridge angle and the distance from the bridge to the mounting point of the tailpiece should give a reasonably good approximation. Mostly because changing the bridge angle means lowering the downward pressure on the top but also raising the horizontal tension on the instrument and thus causes noticeable but hardly forseeable deformation of the guitar.

The real question I had was "where is the pivot point of the Sacconi style bridge?". The string force flows through the pivot point, so if you know that you know the string break angle. At first you think the pivot point is the foot of the saddle where it sits on the top of the guitar, but then you realize it's the point where the sacconi cable exits the tailpiece and goes down at around 45 degrees towards the top/side corner. Since the design has the cable just under the top surface of the tailpiece at that point, you can assume the string force is going through a point approximately 1/10th of an inch below the top surface of the tailpiece. That lets you understand how high you need to lift the thing up, with the saddle and with a shim that the saddle can ride upon. It also lets you realize that most of any string vibration energy that gets transferred to the tailpiece is directly coupled to the top rim through the saddle, it doesn't just get washed out through the cable. There is some nuance to the design.

This is what I did.
I had the tailpiece raised as much as 1/4" higher than what you see. The bridge moved around a little while playing.
Currently at this level it stays in place even with heavy strumming.
If I were to stay with this arrangement I would use slide a rod inside the threaded coupler this would allow for a thicker diameter shaft and the tension would maintain the height.
Also, with this project I have almost no break angle over the zero fret.

My initial tailpiece mount design is essentially identical to yours, but I did not allow for the need to cantilever it above the top. Hence, the bolt that goes into the threaded insert in the tailblock ends up being almost 3/4" in bending, and it's not strong enough to support that torque without moving. That means the plinth that I have rotates down at the front, up at the back, and I don't like it. I will probably keep the design and use it again taking those things into account from the start, but for this guitar I will revert to another design that I know will work. How interesting it's been exploring this aspect of guitar building, I must say!

Eldon, is your tailpiece just made of wood? It looks great! I would just be too scared that it would break if it was just wood (not saying I think your will), it just looks very delicate.

Thanks,
Because it is curly maple, I routed out as much wood as possible and laid in a curved layer of carbon fiber. Also, the tail block is Basswood with carbon fiber for added strength as well.
I was reading one of Alan's post about taping some lead weight to the tailpiece and changing the sound. So I clamped a bunch of small spring clamps to the tail piece and back off again (not a lot of weight) and I could not hear much difference. Maybe not enough weight or the carbon/wood structure doesn’t move much anyway?

Thanks! Pretty ingenious. It amazes me all the people on her who not only figure out unique ways to solve problems, but who are so willing to share. Thanks to you all!

I used a similar idea on my 'Winter' archtop. It was actually based on the way they attach the tailpices on viols da gamba. Rather than use a gut to tie it on they dovetail a square post into the tail block that sticks up above the top. It has horizontal notches in the back edge, maybe 1/4" apart vertically. The tailpiece has a square hole in it to fit over the post, and the back edge is rounded over to fit in the notches. If you decide you have too much or two little break angle you can slack off the strings and move the tailpiece up or down.

On my guitar I used a round post that plugs into a hole drilled into the top and the tailblock. The top of the post is a round knob, and the tailpiece has a corresponding cup with a hole in the bottom for it to go through. The bottom end of the post has a screw in it, which goes into a threaded insert at the bottom of the hole. It's fairly easy to adjust the post height with the strings slacked off, and the adjustment is continuous. That's probably gilding the lily, but what the heck. It looks OK, and has held up well. The main problems are that any strap button will need to go off center, and the output jack as well.

Alan Carruth wrote:I used a similar idea on my 'Winter' archtop.

Would you happen to have some pictures of this? I am very curious.

Thanks,

-Eric

Allan, i have looked up Your website, and there is one remaining question: Did You use some kind of reinforcement in the tailpiece? And if yes - do You think wood laminated cross grain would keep up with the strain?