The PSG and Properties of Materials

ed packard · Post by **ed packard** » 5 Jul 2002 9:32 am

The PSG is made up of a body, a neck, a nut/roller, a changer, cross shafts, bell cranks, rods, stops, pedals/levers, legs, strings, tuning mechanism, and a pickup. When these are joined into a unit we want physical and tonal stability for variations in temperature and applied operating forces.

The body should be of such a stiffness that it does not bend enough to appreciably change the pitch of the fininshed instrument when a weight is applied to its approximate center when the ends are suspended by the legs.

The body should be of a thermal stability sufficient to minimize the change in pitch due to change in operating temperature over reasonable playing times in a variety of playing environments.

The neck should also accept the physical and thermal stresses without adversly changing the pitch.

Possible choices between body and neck characteristics will depend upon particular instrument design practices re how much the other mechanisms can apply stress changes to them. Different designs may require different material and dimensional choices.

The rest of the mechamisms should be such that that any applied stresses, ..thermal, physical, operational, ..do not adversly affect them or the other components with which they come into, or with which they are in contact.

Acoustic impedances of the materials, and the interfaces between the various materials should be such as to not have an adverse effect upon the desired tone and sustain.

Particular points of irritation in instrument designs have been where the legs join the body (long term wear), string breakage, detuning when pedals and levers are activated, detuning with temperature change, non return to pitch when pedals and levers are deactivated, and pickup microphonics.

To control (reduce) these irritations we have the choices of materials, dimensions, and construction practices. Ignoring weight, we can often trade materials and dimensions to reach a given compromise.

There are "properties of materials" that allow for calculating/estimating how these tradeoffs might be made and how they might affect performance and sound. These "properties of materials" may be found in various reference books, by doing a search on the web, or by contacting the makers of the various materials.

The properties of the materials of most concern are:

MODULUS OF ELASTICITY = MOD E; This property has a "numerical" value, generally expressed in pounds per square inch (PSI) times 10^6. This property is an indicator of how much the material will change its shape for a given applied stress, and then return to its original condition when the applied stress is removed. Strings, bodies, necks, cross shafts, changer parts, bell cranks, stops, all make use of this property.

At some value of applied PSI, The materials will not return to their original condition when the applied stress is removed. At a further value of applied PSI, a material will fail completely. The G# string is a good example of this, ..it tends to need retuning a lot just before it breaks. When it needs retuning it has reached the point of "plasticity".

HARDNESS is the next property of interest; It is stated in various terms for various materials. Hardness is tested by pressing a known shape into a material and then measuring the resulting depth of penetration. This property value allows estimating the effect of having two materials pressing against/into each other, ..such as a string against a changer radius, or a leg into a socket/thread.

DENSITY is the material property that may be used to calculate the weight of a part. It is commonly given in Grams per cubic centimeter (gms/cm^3). DENSITY and MOD E may be used to determine SOUND VELOCITY in the material. SOUND VELOCITY and DIMENSIONS are responsible for ACOUSTIC IMPEDANCE which is responsible for tone and sustain.

LINEAR THERMAL COEFFICIENT OF EXPANSION (TCE1) is a measure of the amount of expansion in a given direction a material will have for a given amount of increased temperature. This property is generally expressed as PARTS PER MILLION (PPM) per DEGREE C, or DEGREE F = a value X 10^6; Aluminum is given as 24 X 10^6 C, and as 12.6 X 10^6 F, ..meaning that for every added degree of C the material will expand by 24 PPM of its previous value. Note that the expansion per degree C is about twice that of per degree F, ..this can ease unit conversion.

STEELS may have A TCE of 15 PPM per DEG C. This means that steel strings will not expand the same as aluminum for the same temperature change = pitch change.

HARDWOODS have two LINEAR TCE values, ..one with the grain, and another across the grain. With the grain, the values are in the order of 3.39 PPM per DEG C, ..across the grain it is in the order of 25.7 PPM per DEG C.

LAMINATED woods will have LINEAR TCE values that will be a function of processing methods and specific materials.

Structural and sound properties of wood are grain direction dependent. Humidity absorption of unsealed wood generally causes more dimensional change than does thermal change for common playing environments. For a sound comparison, go play your Martin acoustic outdoor on a cold dry Autumn day, and on a hot humid summer day, ..what a difference.

CARBON COMPOSITES have varying material properties, depending upon the direction and size of the carbon fibres, and the fabricating process; In other words, it
is basic materials and process dependent.

Some Comments:
A low value of TCE is not necessarily a good thing.
A high value of hardness is not necessarily a good thing.
A high value of density is not necessarily a good thing.
A high value of MOD E is not necessarily a good thing.
A high value of sound velocity is not necessarily a good thing.
It all depends upon what you are trying to achieve in the part/mechanism/interface.

If there is an interest, I can post a chart of some materials and their properties, ..a call to instrument and string manufacturers MAY get you some material definitions and dimensions for their parts as well as how the parts are attached to each other.

ed packard · Post by **ed packard** » 6 Jul 2002 1:27 pm

Experiment of the day, ..re TCE and pitch shift:
1. 10:55 AM; Tuned up the instrument in the garage.
2. 11:00 AM;Moved the instrument to the back porch, but in the shade; checked tuning, temperature, and humidity.
3. 11:15 AM; After instrument had set for a while, retested.
4. 11:30 AM; Retested.
5. 11:35 AM; Moved instrument into sun; checked tuning, temperature, humidity.
6. 11:45 AM; Let instrument sit for a while, retested.
7. 12: noon; Retested.
8. 12:10 PM; Moved instrument back to the shade, retested.
9. 12:50 PM; Retested.

Results for 1 above; Temperature = 73 degrees F, Humidity = 27%, Tuning = +/- a cent

Results for 2 above; Temperature = 90 degrees F, Humidity = 18%, Tuning = -10 to -30 cents, ..wound strings worse than plain strings.

Results for 3 above; Temperature = 90 degrees F, Humidity = 18 %, Tuning = +/- 1 cent.

Results for 4 above; Temperature = 90 degrees F, Humidity = 18%, Tuning = +/- 1 cent.

Results for 5 above; Temperature = 92 degrees F, Humidity = 18%, Tuning = +/- 1 cent.

Results for 6 above; Temperature = 94 degrees F, Humidity = 17 %, Tuning = +/- 1 cent.

Results for 7 above; Temperature = 94 degrees F, Humidity = 17 %, Tuning = +/- 1 cent.

Results for 8 above; Temperature = 94 degrees F, Humidity = 16 %, Tuning = +/- 1 cent.

Results for 9 above; Temperature = 94 degrees, Humidity = 15 %, Tuning = +/- 1 cent.

The instrument is a 14 string E9/B6/13 series PST, made using Sierra Session series hardware. The neck is 25", ..it is keyless/gearless, ..common string gauges, and SIT strings. The body is an aluminum extrusion for the front, top, and back, 8" wide on top, bolt on end plates and cross shaft holders (ribs), ..top, front, and neck sides are covered with black glued on laminate. Fretboard is black aluminum and glued on. Neck is hardwood (maple), 3/4" thick.

Tuner(s) are Seiko ST747 and Korg DT-1.

Discussion of the results:

The tuners are limited to about the accuracy of +/- 1 cent with the signal supplied from the instrument for normal the soft picking I use for tuning. One tuner was left with the instrument, the other was brought out from the garage each time. Temperature effect on the tuners was determined to be nil via this method.

In step 2 of the test, a variation/deviation detuning range of -10 to -30 cents was found imediately upon the change in ambient environment, ..wound strings varied most. The reason is seen as the differences in thermal conductivity of the string material and the body/neck/changer/tuner assembly; Of greater importance is the differences in time taken by the low string mass and larger body mass to reach thermal equalibrium. After about three minutes the tuning was restored.

In step 5 the instrument was moved into the direct sunlight; now the color can make a difference as black tends to absorb more heat than white would. It got uncomfortable to touch. It would take some small (minutes) amount of time for the heat on the laminate to transfer thru the glue and into the body material, hence the apparent absence of detuning during this step.

This combination of body, neck, changer, tuning mechanism, and strings appears to be thermally stable re normal picking environments (inside to truck flatbeds in the direct sun).

[This message was edited by ed packard on 07 July 2002 at 02:39 PM.]

Rick Collins · Post by **Rick Collins** » 6 Jul 2002 4:00 pm

Ed,

Even the speed of light, which is treated as a universal constant, is not, in reality, a universal constant. It slows from 186,284 miles/second, depending on the medium through which it travels.

What are we to use as a measuring device to determine what tolerances we as musicians will tolerate from our instruments? ...in my opinion, our ears.

You do bring up some interesting points. I've often wondered if a steel guitar would sound appreciably different at a very high altitude; for instance, at the airport in La Paz, Bolivia, which is above 13,000ft. MSL, as opposed to 10ft. MSL at Santa Barbara, CA. Since sound travels better through a more dense medium, would the properties of the same sound in a much less dense atmosphere change; so as to be detected by our ears?

Rick

Bobby Lee · Post by **Bobby Lee** » 6 Jul 2002 4:06 pm

Ed, when you say "checked tuning" (as in step 2), do you mean that you retuned the instrument, or that you simply observed the tuning?

Jim Smith · Post by **Jim Smith** » 6 Jul 2002 4:35 pm

sound travels better through a more dense medium

That's why music sounds better in a smokey honky tonk, the air and most of the patrons are dense!

BTW, I believe the speed of light is only defined to be constant in a vacuum.

ed packard · Post by **ed packard** » 6 Jul 2002 6:22 pm

Rick, ..I was worried that the "speed of light" constant was a double ten constant, not a Universal one!

Our sonic sensors (ears), and the analysis mechanism (mind) are the final personal judge of acceptable tuning/sound, ..problem is that all are not equal and acceptability is subjective. Quantifying the changes/differences and defining what causes them makes repeating and refining the performance "goodness" (cloning?) easier.

b0b, ..the object of the test was to see the degree of pitch change as a function of thermal change, ..the tuning was set once in the garage and only monitored (checked)thenceforth. The components that could cause a shift are the body, the neck, the strings, the changer, and the tuning mechanism.

The linear TCE is about 12 ppm per deg C for the steel strings, 24 ppm for the aluminum in the body/changer/tuning mechanism, about 4 ppm for the maple (in the direction of the grain. As the temp goes up, the strings expand in length and diameter, thus lowering the pitch; on the other hand the body/changer/tuning mechanism material also expands. The body is lengthened, and the neck is lengthened thus tightening the strings and raising the pitch. By proper dimensional choices one effect can be made to cancel the other, thus reducing pitch shift (detuning) vs temperature change. The tested design is well balanced for this purpose as the test data supports.

William Steward · Post by **William Steward** » 7 Jul 2002 12:25 am

Ed:
This is fascinating stuff and as you know I am interested in physics. I did my own 'uncontrolled' experiment on my 25 year old MSA guitar last Thursday.

7:00 am - packed up tuned guitar in its 25 year old musty case; location: Living Room; temp: 78 degrees F. approx. rel. humidity 70%

8:30 am - 5:00 pm - left guitar and other equipment in back of Explorer during work day; temp: 110 degrees+ rel. humidity 85%

6:00 pm - pm set up guitar on makeshift stage for outdoor July 4th party at hotel bar precariously balanced by the swimming pool; temp./rel. humidity - approx. constant; tuning: 3rd string- (- 5) cents

11:00 pm - drunken tourist knocks guitar over and I catch just before it bounces into pool; temp. 90 degrees 90% rel. humidity

8:00 am Saturday - take out guitar from musty case now smelling strongly of malt beverage; location living room; temp/humidity: 78 degrees/75%; tuning: pretty much on the money.

I will try to repeat this experiment under more controlled conditions but did I think it spoke volumes about the quality of the instrument to have so many dissimilar materials and yet be so stable under wildly different thermal conditions. Hopefully the new ones will be just as robust.

Reece Anderson · Post by **Reece Anderson** » 7 Jul 2002 7:40 am

Wil....Were you to be playing an MSA Millennium you would need not be concerned about it falling in the pool or playing it in direct sunlight because the guitar does not recognize moisture, heat or cold.

ed packard · Post by **ed packard** » 7 Jul 2002 8:31 am

Reece;
Not sure what you mean by "the guitar", ..the strings, and any metal parts, all are affected by changes in temperature (is this what you mean by "recognize" heat and cold); and the pickup will not like water too much if it is not sealed. Since the pickups are interchangeable, I assume that the contacts might oxidize (unless precious metal coated/plated). Your body material will have a "THERMAL CONDUCTIVITY" value so it will transmit heat/cold at some rate; If its TCE is about zero, then the strings will change pitch assuming that the rest of the structure does not compensate for the strings TCE.

Aluminum has a very low moisture absorption figure, as I would assume (I do not know the particular composition) your Carbon Composite does also, ..wood is another story.

Any place that dissimillar materials come together, moisture and temperature change tend to do damage.

Wil's "alcohol test" and the "fat lady/man falling on the guitar test" should be combined if the most results for the least testing are to be obtained.

What are the values on your instrument for the unchanged string detuning as a function of activated changes (A & B pedals etc.)?

What are the values of detuning experienced for the simple test outlined in the thread?

What are the detuning values for applying a downward load to the center of the body or neck?

There are four types of instrument (like people), ..those that look good, those that play well, those that sound good, and those that wear well; It is hard to get it all in one package so one usually has to "pick their poison".

In my opinion, which does not count for a whole lot, it is about time that manufacturers gave some numbers, test results, and test methods, not just sound bite statements, ..why not start a trend?

Reece Anderson · Post by **Reece Anderson** » 7 Jul 2002 10:58 am

Ed....IMHO formulating an agreeable process of determination procedures and subsequent agreement of test results when completed would be extremely difficult to achieve, and even if accomplished, would appear to many to be subjective, bias, and therefore non conclusive.

However, with that said, were those things to be accomplished, I would be happy to cooperate in your search for conclusions which I consider to be very worthwhile, but lofty aspiration.

Rick Collins · Post by **Rick Collins** » 7 Jul 2002 12:49 pm

Rick, ..I was worried that the "speed of light" constant was a double ten constant, not a Universal one!

Ed,

You are closer to defining the real problem than you know,___"worried" (your word); especially over those things, many of which are trite.

Don't get me wrong;___I think the "scientific method" is the best way to solve physical, chemical, and even biological problems. I can appreciate your work and the measurements you took. When trying to come up with a "recipe" for combining materials of different resonance to blend into a single unit that will produce something as subjective as musical sounds, and at the same time eliminate our detuning problems, we very quickly reach the point of diminishing returns. If we do refine the steel guitar to the point that is "near perfect" to us, this "near perfection" in all likelyhood will not be noticed by the audiences of our music.

Your turn, Rick

ed packard · Post by **ed packard** » 7 Jul 2002 1:35 pm

Reece, I think that some basic pitch stability tests that relate to real world situations are easily defined and executed. An example is the simple one that was described above. Anyone with a PSG and tuner can do it and see how stable the instrument string combination is. A few degrees difference, or humidity differences are a minor issue, ..either the instrument will be pitch stable or it won't. Same for detuning as a function activated changes, or for the load suspended from the center of the body/neck.

If the forum folk were so inclined they could do the above with their own instruments, send the info to the forum (or post it), and to the instrument manufacturers, ..who knows, the instrument makers might even appreciate getting the results.

Rick, what do you look for in a PSG? What do you listen for in another player? Do you care if your instrument shifts pitch when the environment changes? Do you care if the middle strings (or others) detune when the A/B pedals are activated?

For the scale/lick/two strings at a time player detuning means less (as long as open strings and open string harmonics (chimes)are avoided) than to the three strings or more at a time player as the first can compensate by bar position, the latter can not.

That music is subjective does not mean that it cannot be analysed, defined, and that the mechanism that makes it cannot/should not be optimized re the playing function; that way the player gets to tune less and play more.

We will hit on the string breaking issue soon, ..sure would be nice to reduce that problem to a set of parameters that could be applied across the board for a solution/improvement, ..the principals are straight forward but the devil is in the details.

Sage · Post by **Sage** » 7 Jul 2002 5:51 pm

Ed- some food for thought before you launch into string breakage- If you have the strings passing over a zirconia nut, even with the strings at a low relief angle and a linear pull, won't the string windings cause a problem with accurate intonation?
And folks, I did send Ed some test results of spectral analysis that we did on our lap steel compared with a traditional design, as well as TCE, MOD E, and other specs for our carbon fiber. I'm not posting them all up here for 3 reasons- I'm not good at putting pictures up here, Reece is probably right that it is hard to get good controlled test information and even then it may look suspect and finally, there (relatively speaking) aren't that many folks out there who'd know what to do with MOD E specs if I gave it to them.
Nonetheless Ed, I do appreciate the level of care that you display in coming up with these questions and tests. As a builder, I would welcome such outside testing to inform the improvement of my product. So, thanks.
T. Sage Harmos
Harmos Steel Guitars
(updated web site) http://www.harmosmusic.com

ed packard · Post by **ed packard** » 8 Jul 2002 8:45 am

Sage; Thanks for adding some thoughts to the thread, ..you are about as "different" design wise as it gets today (your 3D Matrix body concept).

That any test data might be suspect and inaccurate at first is , to me, no excuse for not starting down the trail. Sooner or later it would clear up and certain tests become "standards" in the industry, ..each instrument delivered should be certified to meet certain parameters re thermal/pitch change, and change activation related pitch change at least.

Concerning the Zirconia nut question, ...At present my lowest string is an 0.80 gauge. It crosses a .0625" dia roller on the nut end and a 3/4" dia changer curvature. The materials are steel, and aluminum respectively. There are no noticeable issues with this fat string (or the others) re tuning, or intonation. I have several bars, ..A 1" Dia BJS, .. Two Zircs, one is 1" and the other is 7/8". The BJS is noisier on the wound strings than the Zircs are. Placing the 7/8" Zirc under the strings next to the nut does not cause the wound strings to tune in jumps (due to the windings). This leads me to accept that a Zirc or other bar material of reasonable diameter will work as either a nut component, or an under the string capo component.

Sage · Post by **Sage** » 8 Jul 2002 9:40 am

Ed, that is cool about the Zirc nut- maybe the old idea of a non-roller nut could successfully be revisited after all, given new materials and the understanding of how to do it.
As to standardized tests, we are recording an impulse test of each steel that we make, so that it's resonance can be a part of it's production record. The sound is still the best test for me, but seeing it's spectral profile can help make sure that there is nothing "out of whack". They may all come out looking about the same in the end, but there may also be some value in having that record. Of course timbre changes a little over time, so it is hard to say what the long term value would be. Anyway, it is a step down that road that you are talking about.
T. Sage Harmos

Al Marcus · Post by **Al Marcus** » 8 Jul 2002 10:22 am

Sage- you mentioned the non roller nut.

Probably not a bad idea. I see that the new GFI has just a straight bar over the nut and held there, by the strings.
It seems to work real well.....al

ed packard · Post by **ed packard** » 8 Jul 2002 11:02 am

Nuts, with and without rollers: The end that moves (stretches) the string is the changer end, ..the other end of the string does not move. The amount that the string moves over the nut (once the guitar is tuned) due to changer activation, is a function of how far it is from the nut to the tuning mechanism. On guitars with tuning keys, this distance is reasonably long for the middle strings; For the keyless/gearless changer it is very short. Rollers are less valid for keyless than for tuning key designs.

Take a 25" neck with a 3 halftone raise, and 1" from the tuning mechanism to the nut. The changer end of the string will stretch less than 0.100", the tuning mechanism 0.000", the over the nut movement will be < 1/26 of the total string stretch, or less than 0.004". Not much motion over the nut, ..You probably won't see the a roller move under these conditions, an it may take that much force to break the "sticktion" of the roller on its axle.

A straight bar for a nut has two problems, ..first, is that the tops of the strings will not be in plane at the nut (the bottoms will), and second, is that the lateral position of the strings must be determined in some manner. Gauged slots (re depth) are a possibility if the tops of the strings are such that the bar can be passed over the nut while touching the strings. V groves would center the strings, but add friction to the string/nut interface, and be a home for burrs, dirt, and other string breaking elements. A step machined, high value of hardness, polished to a micro finish bar of 0.5" to 1" diameter with some sort of guide (pins/trough) behind the nut should do the job.

Why do not the makers of PSG with keyless/gearless do this? I would think mostly because it is not traditional (folks buy ala tradition), and because it would increase the manufacturers part count; Plus the rollers work fine when gauged to put the tops of the strings in a flat plane, and allow the bar to pass over the roller without hitting it.

Everett Cox · Post by **Everett Cox** » 8 Jul 2002 2:40 pm

---"A step machined, high value of hardness, polished to a micro finish bar of 0.5" to 1" diameter with some sort of guide (pins/trough) behind the nut should do the job."

Yes, Ed-- that is pretty much what I do for the less involved lap steels I build. Except that, using 1/4, 3/8, or 1/2 inch bar stock, my nuts and bridges are hand filed. (Can't afford a machine shop for these 'freebies' I giv away.) They're made from brass and filed to have a fairly sharp edged parabolic profile. Brass pins in another piece of stock maintain string allignment with the strings angled outward about 20 degrees to the key heads and ball anchors. Works fine and gets good tone and sustain. Has the problem, though, of requiring a different set for different gauged strings.

I've been working on an adjustable nut with individual pieces for each string. Carl Dixon and I have each come up with a similar idea. Much more time consuming to make this type of nut. I also would like to do away with the rollers. Are you saying that rollers might not really be required???

Thanks --Everett

ed packard · Post by **ed packard** » 8 Jul 2002 2:56 pm

Everett; to the degree that the string does not move after it is tuned, why does one need rollers? keyless/gearless approaches on the PSG can make the amount of string motion at the nut so small that rollers serve little purpose, ..the convenience is that they are accepted as required for the keyed tuner guitars with lots of string behind the nut.

Smiley Roberts · Post by **Smiley Roberts** » 8 Jul 2002 3:31 pm

In response to Ed's original posting:

HUH???

ed packard · Post by **ed packard** » 8 Jul 2002 5:07 pm

To Smiley, Aw, keep your shorts on!!! Everyone knows that you are a Yankee, who else would wander around with their underwear outside their pants?