By Roger H. Siminoff
June 3, 2014 - 9:00 am
Copyright 2014, Roger H. Siminoff, Atascadero, CA 93423. All reproduction rights reserved. No image or part of this publication may be reproduced, published or redistributed without the prior written permission of the author. Web site: www.siminoff.net
In my five decades of luthierie, string tension has always been a major focus for me. From a structural standpoint, it seems rather obvious that the strings exert sufficient tension to bend a neck, warp a body, or damage a bridge. This, of course, was not a revelation on my part; the structure of all string musical instruments — from pianos to ukuleles — and the compressive loads imposed by the strings' tension have been addressed by countless builders and designers who tried to find a way to control and counter the action of the strings' load, which attempts to compress the instrument into a twisted mess.
But, the acoustical issues related to string tension was one that I believed was sorely overlooked. I sought to learn more about the longitudinal peghead-to-tailpiece loads (or in the case of the acoustic guitar, the peghead-to-fixed-bridge loads) as well as gain a better understanding of how tone is affected by the down-pressure on the soundboards of those instruments with movable bridges, such as mandolins. This was an interest that Jim Rickard (1942-1996, former acoustical engineer at Ovation and columnist for FRETS Magazine), and I shared and spent many hours discussing and testing.
String tension measuring
In the early 1970s, I performed some tests to measure string tensions at the bridge of fixed-bridge instruments, and the associated lateral down-pressure loads on those instruments with movable bridges and tailpieces. What I found was that the relative tension of each string in a set of strings was critically important to the timbre; amplitude, sustain, clarity, and most importantly, the string-to-string balance.
Antonio Stradivari was obviously keenly aware of how energy was driven through the bridge to the belly (soundboard) of his violins. He is credited with the design of the modern violin bridge whose kidney and volute shapes does not allow any of the four strings a direct route to the belly. All strings are over openings, and the energy from each string is modulated through the bridge's waist.
Consider for a moment the traditional mandolin bridge with two posts and adjusting knobs (based on Gibson's January 1921 patent) is designed so the two outer pairs of strings are close to the posts and the two inner pairs are closer to the center of the bridge's saddle. In this design, the outer pairs produce a different timbre than the inner pair because of the flexibility of the saddle and the string pairs proximity to the posts (and to the soundboard).
To counter this design flaw, the down pressure of the center two pair of strings should be slightly less than the outer two pair, and this calls for a set of strings whose load, or tension, is calculated and adjusted accordingly.
A similar unequal distribution of energy can happen on acoustic guitars with fixed bridges (i.e., Spanish and conventional steel string). If one or more of the six strings exert a greater pull at the bridge than (an)other string(s) in the set, the strings with lesser tension will not be able to activate the bridge as readily because the tension imposed by strings with greater tension will have to be overpowered first.
It is important to note here that fixed-bridge instruments are not driven by the strings' down pressure. In fact, there is virtually no down pressure exerted by the strings on the soundboard at the bridge.
Guitar bridge torque
Fixed bridge guitars work on a torque or twisting moment in which the strings' tension (A) causes the bridge and soundboard to be twisted toward the peghead (B) as a result of string tension (load) at the bridge. This twisting effort is readily visible in the hollow or depression normally found in front of the bridge, and the hump or raised area (C) normally found behind the bridge. As a result, total string tension coupled with the longitudinal energy sent to the bridge is critical to the amplitude, balance, and timbre of acoustic guitars. This is an important consideration that both guitar luthier Richard Schneider (1936-1997) and Jim Rickard took into consideration when developing bridge and bracing systems for fixed-bridge guitars.
In many articles of Pickin' Magazine, we discussed string loads and tensions, and often suggested that it would be beneficial for musicians to select single strings rather than sets. In the 1970s, Gibson made its strings available to purchase individually by the gauge but, unfortunately, string loads (tension) were not listed on the packages so musicians could not make well-informed choices. In several articles, we prompted that string loads should be published on string packages.
By 1979, when I started FRETS Magazine, I became more focused on bringing the subject of string tension to the forefront because I believed the issue was critical. We started a monthly column in FRETS called "String Clinic" in which we measured the gauges and loads of every string in a set, and we reported the results to our readers. In addition, the editors (Jim Hatlo, Rick Gartner, and I) would play an instrument with the tested strings and comment on what we heard.
The first six or seven issues brought some comments from readers, but the column went seemingly unnoticed by the string manufacturers.
After about six or seven months, we began to get calls and letters from various string manufacturers whose tone ranged from, "This is interesting; can you tell me more?" to "All you are doing with this column is confusing the musicians!" A few were more emphatic: "Please stop this column! Do you realize what a mess you're making!" One manufacturer complained that monitoring, measuring, and reporting string loads, as well as printing new packages and ads would impose an enormous financial burden on his business.
I'm pretty stubborn, especially when I believe strongly in something so the FRETS String Clinic column continued. More than once, Jim Crockett (publisher of Guitar Player, Keyboard, and FRETS Magazines at GPI Publishing) called me into his office to read a letter from another string manufacturer - who was also an advertiser (yes, read: $$$$).
Jim wanted to double check that I was still on the right track. "Are you convinced that your findings are correct?" he'd ask. "Can you substantiate and replicate your tests?" And, each time I would look him in the eye and say, "Yes, Jim. I'm absolutely sure we're on track here and are doing the right thing!" With that, Jim would write back to the advertisers and politely say, "Sorry, but the column stands."
A few string manufacturers saw merit in what I was doing. I was a consultant to Gibson at the time, and Bob Lynch, president of Gibson's string division in Elgin, Illinois, engaged me in the process of measuring the loads of every string Gibson made and analyzing the relative loads of strings in their sets. We then prepared new string sets with balanced tensions and sent them to Bruce Bolen (Gibson's head of R&D and customer relations at the time) for him to personally test as well as for him to distribute to prominent musicians who were using Gibson strings. Within about eight months of our work, Gibson began producing string sets with balanced loads and began reporting its string tensions on the their packages.
The concerns from other string manufacturers didn't go away. In fact, it got downright testy at times. On one occasion, two key members of Ovation's string division made the trek from Connecticut to California to visit with the staff and me to see how we were measuring the strings loads, and to challenge why we felt this topic was important. Things were a bit uncomfortable in the morning, but by the time we got through with lunch, described our rationale, showed them our testing methods, and demonstrated a few instruments with balanced and non-balanced strings, they had calmed down a little, but were still frustrated that we were "stirring the pot."
With the hope of easing tension and getting some consensus, we thought it would be beneficial to gather all of the string manufacturers (something that had never been done before) at a meeting during the 1982 NAMM (National Association of Music Merchants) Show in Anaheim, California. The meeting was scheduled for February 7, 1982 at the Inn at the Park Hotel, and invitations went out to all of the prominent domestic string manufacturers; most said they would attend, but a few could not.
In attendance, representing FRETS Magazine were two editors, Jim Hatlo and Rick Gartner, Jim Crockett, our publisher, and myself. Industry attendees were Dave Holcomb (GHS), Bob Lynch (Gibson), Ernie Ball (Ernie Ball), Chris Campbell (Dean Markley), Jim D'Addario (D'Addario), and Paul Damiano (Kaman). John Dusinski (Martin) responded saying he wouldn't be able to attend, that he "believes there is already standardization between manufacturers," but he did think the subject should be pursued. Neil Lilien (Guild) had a meeting conflict and could not attend, and Stan Rendell, former president of Gibson, and Dick Sievert (both of Sterlingsworth strings) wrote back that they were "not attending this NAMM but very much wanted to be involved in future string tension efforts."
Lunch included the normal casual-but-guarded conversation among competitors, small talk, and tech talk. And, Jim Crockett was occasionally reminded, half-jokingly, that FRETS was financially supported by many of its advertisers sitting at the table. And there was some chatter about how FRETS could not possibly support its findings about string tension making a difference.
After lunch, armed with a bunch of flip-chart drawings and some photo enlargements of our String Clinic column, I reviewed much of the same findings that you read earlier in this article and went into some areas of string tensions and download pressures with greater detail. Using two new "identical" Mossman guitars (Stuart Mossman [1942-1999] brought what he felt were identical guitars to NAMM for us to borrow for the test), we measured a string's tension (load) right there to demonstrate our procedure and the equipment we used for measuring. On one guitar we installed a set of strings with balanced tension, and on the other a standard set of strings (I'm intentionally omitting the string brand here). Our editor, Rick Gartner, an accomplished guitarist, demonstrated both guitars and put on a good show playing identical numbers. He mentioned that he was making every effort to apply the same attack and emphasis in both performances. Just about everyone agreed that the guitar with balanced strings sounded better. At that point there were all kinds of subjective comments along with some objections. The strongest common thread that emerged was, "Yes, those might be two 'identical' guitars, but everyone knows that no two guitars are alike." "The more balanced one," some said, "is most likely just a better sounding guitar!" after which you could hear the rush of a soft "YESSS" and mumbles of agreement and nods of heads coming from the room.
It was something we hadn't planned for, but I believed in our data and was willing to take a calculated risk, so I turned to Jim Hatlo and Rick Gartner and asked, "How quickly can we move the strings from one guitar to another?" I told Jim Crockett our plan and he promptly got up to say a few things and provide some cover while Jim, Rick, and I feverishly swapped strings. When we were done, and the guitars were up to pitch, Rick performed again.
There was now a hush, and within a minute or so, most sheepishly agreed that the balanced sound moved with the string set from one guitar to another. We said nothing and just stood there and looked at our guests, allowing them time to reconsider.
In the year following our meeting, there was a lot of follow-up discussion. We continued to do our FRETS String Clinic column, and the interaction with string manufacturers turned from negative comments to increased dialog about our findings. Some manufacturers sent us sample string sets to evaluate while a few others still pushed back hard. It took almost a year before we noticed that many string manufacturers began to print string tension information on their packages, and today we find that most string manufacturers not only provide the data, but some have become highly proactive promoting string tension information. As of this writing, D'Addario has a String Tension 101 document on its website, and the company provides tension data on their packaging for all their string sets. Other manufacturers, like GHS, report "high tension," "medium tension," etc. (which is somewhat helpful).
Having string tension information available enables you to make better choices about the total loads you want to subject your instruments to, as well as provides you with data to achieve better string-to-string balance from your instruments.
If you are a luthier, you'll want to consider how the overall string load or down pressure affects the structure of your instrument. The goal is to have strings with tensions that can evenly and effectively excite the bridge and soundboard system. There are no specific rules for this since much depends on the bridge design and how you build and brace your instrument. But, being sensitive to string loads, coupled with a bit of experimentation, will get you a long way down the path to great-sounding instruments.
For the musician, being aware of string load data provides another method for you to know more about the strings you are selecting. String gauges alone do not tell the whole story — especially on wound strings. Your wound strings are made of a core or inner wire wrapped with a covering or "wrap" wire, and there are many ways to achieve the same overall gauge. For example, a .024" (twenty four thousandths) string could be made of a .012" core with a .006" wrap wire (.012" + .006" + .006" = .024"). Or you could make one with a .014" core wire and a .005" wrap wire (.014" + .005" + .005" = .024"), and there are many more possibilities. Each combination you come up with will have a .024" result, but the tension and playability of each string will be greatly different.
Here are some general guidelines: We learned that for mandolins and mandolas with two-post bridges, the string downward loads should be at least 10% to 12% greater on the outer pairs than they are on the inner pairs. However, the key for us to arrive at the ideal loads was finding the right core-to-wrap wire combinations through extensive testing.
For banjos with three-footed bridges, the three strings over bridge feet should have at least 10% to 12% more load than the two strings that sit over the bridge's arches. For fixed-bridge guitars, the string-to-string tensions should be either very similar or linear (example only - not real values: 16, 16.5, 17, 17.5, 18, 18.5 pounds), or mapped in a somewhat parabolic curve.
During the few years that followed our initial work on string tensions at FRETS, I was working on deflection-tuning systems at Gibson for the F5L mandolin project, and it soon became apparent that the issue of the loading of the bridge on moveable-bridge instruments was as important, if not more important than the strings' linear tensions. After all, it was the amount of down pressure that dictated just how much a given soundboard and bracing system would deflect, and how that deflection interacted with the energy being sent by the strings through the bridge. To this end, we actually measured the amount of deflection on the bass and treble side of the bridge of a Loar-signed F5 mandolin in the fully relaxed (strings fully slack) and up-to-pitch states. This testing was the basis for the "deflection tuning" we did instead of using an audible tap tuning process.
Since the down-pressure increases as the string-break angle increases, we also needed to measure downloads at various string break angles (for more on string break angles and relative loads, see http://siminoff.net/string-break-angle-loads/).
String break vs. loads
It has since become abundantly clear that managing the amount of download each string presents on the bridge structure on movable-bridge instruments, and taking into consideration the proximity of the strings' position on the saddle relative to the location of the bridge post(s) is vitally important to the instrument's balanced tone and timbre. The conventional two-posted mandolin bridge has a design flaw in which the inner A and D pairs sit on a relatively flexible portion of the saddle and are further from the bridge posts than the outer E and G pairs. This anomaly can be compensated for by adjustment in the strings' downloads; it is something we've been focusing on for the past two years, and will announce our solution in mid-2014.
An example of the communications from strings manufacturers indicating their opposition to our findings on string tensions, and our recommendation to provide musicians with data.
Jim Crockett’s column in the January 1982 issue of FRETS was published in advance of our NAMM Show meeting with the string manufacturers.
You may leave a comment if you have a Mandolin Cafe Forum account. Clicking "Post a Comment" below will take you to the forum where you can complete this action. Please note that once you have, your comment will appear both on this page and on our forum. YOU MUST BE LOGGED IN to your Mandolin Cafe forum account to comment.
Special thanks to the Mandolin Cafe's primary business partners.