To summarize up front what I achieved:

Cents drop    before     after
4th string      14         9   w/AB
5th string      10         7   w/B
6th string      21        13   w/A
8th string      12        10   w/AB
Cabinet deflection as measured at the 12th fret with a deflection gauge calibrated
 at .0001" increments
              before       after
A pedal       .0009        .0005 Lower
B pedal       .0006        .0002-.0003 Lower
AB pedals       -          .0007-.0008 Lower
D lever       .0002-.0003  .0000-.0001 Raise (push lever until full lower)
D lever         -          .0010 Raise (push lever until no more deflection)
F lever       .0000-.0001  .0000-.0001 Lower (push lever until full raise)
F lever         -          .0010 Raise (push lever until no more deflection)
Nut deflection measured in the middle of the nut span with deflection gauge
              before       after
A pedal         -          .0001-.0002
B pedal         -          .0000-.0001
Changer deflection measured in the middle of the changer span with deflection gauge
              before       after
A pedal         -          .0001-.0002
B pedal         -          .0000-.0001

I was able to lower the "cabinet drop" with my modifications. I do not have nut and changer "before" readings. The second set of D and F lever readings illustrate quite clearly part of my observations in that the stops of the knee levers contact the body of the guitar while the pedal stops do not.

Hypothesis:
I believed that adding tension to the underside of the guitar body to offset the bow effect caused by the raised strings would eliminate cabinet drop.

Methodology:
I disassembled my guitar and laid aside the body and front apron which are glued together. I retained all of the other parts except rods and tuning nuts which I had to change because of the customization.

The body
The new body is a solid piece of Chunkeq, a variety of Central American Mahogany. I chose this piece of wood for its "musical" quality. I spent a lot of time at the local hardwood suppliers looking for the right piece of wood. I wanted a piece of wood that transmitted sound well. I would hold the piece of wood off the floor with my ear to one end. I would then tap the other end with a fingernail. Most of the wood I found was dead or dull sounding. This included the various species of maple and Honduran mahogany as well as other species whose names I don't remember. Only a couple of pieces sounded good. The piece I ended up with was the right size. It is not a particularly strong piece of wood because it can be bent with some pressure but it sounds good, almost bell-like.
The front apron is a piece of Honduran Mahogany that I had in my workshop.

The Tension system
I built the tension system from aircraft aluminum and 1/4" steel rod. I got a couple of blocks of aircraft aluminum from the Boeing surplus store. I cut it to make angle brackets which I drilled for screw holes and tension rod holes. I threaded both ends of the steel rods. The tension rods sit on both sides of the strings. My guitar's end-plate was a 12 string end-plate so I could move the 10 strings over one position to accommodate the tension rods. I put double nuts and lock-washers on the threaded rods.

Detail

Body not connected to the front and back aprons
I made one further modification. I mounted the body to the endplates only. There is no connection between the body and the aprons, pedal rods, etc. except for the changer. The other exception is the knee lever stop-screws which contact the body at the end of the knee lever change.

How does it sound?
The sustain is amazing! The treble notes are clear and bright. The bass notes make the whole guitar vibrate. I can even feel it in the pedal bar. The guitar is a real pleasure to play because it stays in tune.

Did I solve the Cabinet drop problem?
The answer is no. I decreased the amount of cabinet drop but did not eliminate it. I still have "cabinet drop" which I can attribute to the nut and changer. And I still have cabinet deflection.

Karlis

[This message was edited by Karlis Abolins on 27 August 2002 at 11:45 AM.]

Are you saying by your observations of nut deflection that the nut is moving closer to the changer when pedals are activated? (As a result of the bow in the cabinet?)

Do you have any way to know if the movement you are observing on the deflection gauge is true cabinet movement, or is some of it sag in the strings? I suppose you would have to set up a fixed datum point independent of the guitar to know for sure.

A test on deflection of cross shafts would be some good information. It is my feeling that a lot of the detuning takes place in that area.

These tests are similar to what I do in my line of work, which is aligning internal components of turbines although of course guitars are much smaller and a heck of a lot more fun to work on! We are still looking for movement, and must know if what we think we see is really what is happening. It can be quite maddening, trying to solve a problem like this. To align a turbine, we used to stretch a .017 piano wire about 50 feet through the centerline of the machine housing and use it as a datum line, accounting for sag known at axial distances from the wire support. We use lasers now. Much easier.

I think a guitar would have to get quite massive to completely eliminate cabinet drop. A little deflection in the changer axle, some cabinet movement, some cross shaft deflection, thermal change from stretching the string, who knows what else,it all adds up.

Keep up the good work, Karlis. You can't fix something if you don't try.

------------------

Doug Earnest
The only Zum Keyless U12, Fender Cyber Twin

I measured the cabinet deflection by mounting a bracket to the ends of the guitar and suspending the gauge from the bracket. See the photo of the apparatus used in the before measurements.

It appears that you are making the measurements on the neck; the assumption is that the body and neck (at the point of measurement)behave as one, ..this may or may not be true.

If the steel tension rods are adjusted to eliminate (reduce) the body bowing effect due to activating changes, would you tune first (open strings), then set the tension for least cents change with activation, and then retune or ?.

Assuming that the thermal TCE of the steel strings and the steel truss rods are the same, they are both about 12 times that of the wood, and the truss rods have the larger volume per unit length, hence they will have the greater change with temperature.

The tension on the neck via the tuned strings is in the order of 280 or so pounds; May it be assumed that the tension on the truss rods adds up to about the same value?

If/when the body bows from changer activation, the changer and nut mechanisms will "tilt" if they are attached to the body thus amplifying the effect of the any bowed body dimension to the string length dimension.

Just some thoughts and questions in passing.

The neck and body are firmly screwed together so I have measured the neck only. I would have to assume that any deflection of the body will be transmitted to the neck. I will however have to check this and report later.

As far as the order of tightening is concerned, my method was to tension the neck with the rods before attaching the strings. I then tried to balance the tension experimentally by tightening the tension rods further. A half turn of the tension rod nuts caused approximately a 25 cent raise on all of the strings. I currently have a slight negative bow on the body (as seen visually from the end of the guitar) in relation to the strings, that is it is bowed toward the strings rather than away. I have not yet experimented with the amount of cent detune relative to the tension of the rods.
Your observation relative the thermal TCE is probably accurate and will be demonstrated over time.
The tension on the rods is probably above 300 lbs. since there is a negative bow. If there were no bow, the tension on the strings would approximate the tension on the rods.

Your observation on the nut and changer tilting adds yet another dimesion to the whole problem. I already have determined that the nut and changer deflect due to lack of strength. This was demonstrated by measuring the deflection in the middle of the nut versus the end. The nut deflected the most in the center of the span and the least at the 1st and 10th strings where it was almost unmeasurable. The changer demonstrated the same problem. The issue of tilting is an additional factor that I will leave for the moment.

Karlis

[This message was edited by Karlis Abolins on 28 August 2002 at 03:23 PM.]

When I asked if the nut was moving toward the changer, I was trying to ask about what is now being described as tilt. I'm not very good at saying what I mean sometimes.

I think you did a good job in the way you mounted your measuring device. Rather than clamping it to the endplates or body, you might just try a small amount of weight on it to keep it from trying to walk around. Or maybe you did try that, I'm just throwing out ideas.

I'm very interested in the new MSA, in regard to cabinet stiffness and tone.

If a stiff cabinet at a reasonable weight cannot be produced, we'll just have to live with compensators as a means to fight de-tuning. They work pretty well. I read a long time ago where Maurice Anderson said he was not concerned with cabinet drop. I don't think very many top pro players get real hung up on it, as long as it isn't a gross amount. I know it doesn't hurt my playing although for an entirely different reason.

Keep up the good work!

------------------

Doug Earnest
The only Zum Keyless U12, Fender Cyber Twin

There is another 'solution' that is far less elaborate but 'kinda' solves detuning. I had a guitar sounded great but detuned badly -- Most of all on the big, unwound 6th string. You could not properly tune all A/B pedal combinations to be in tune. When you pressed the A pedal the 6th string (which is already adjusted slightly flat) to drop 18 cents or more.

I was told (by Ricky Davis and others) to try a wound 6th. I raise the 6th a whole step and also lower it a whole step, so there was some question whether a wound 6th was in my future. I was able to get more than a full tone in each direction on my Fessy. Side effects: slightly longer pedal/lever throw (only a minor problem on the B pedal); FAR less temperature sensitivity; string itself lasts longer with good tone; sounds better.

The 5th still drops with the B pedal slightly but, overall, the detuning problems that bugged me the most were gone. (Ricky told me, but I didn't believe him)

Let me clarify
* It ONLY improves detuning for the 6th string and the difference is not because of the tension. Since the core of a small wound string is about 012 or so, it is not as sensitive to temperature and cabinet drop. The plain string is much larger than the core of the wound. As you know, slight mechanical deformations or thermal expansion/contraction make the plain string go further out of tune, compared to the wound string with the smaller core, since the detuning is more or less proportional to the size of the core of the string.
* The wound string sounds better. If you look at the vibration with an o'scope or even a good guitar tuner, you can see that the signal from an unwound 6th is FULL of overtones and is difficult to tune. The wound string has much more fundamental and is sonically much cleaner sounding to my ears.
* The wound string stays in tune better. The problem I was having caused me to have to retune on stage too often. After one song, the sixth string was not the same (within acceptable/audible tolerance) as when I started. When using an electronic tuner, when the note is first struck, the wound string only goes slightly sharp, whereas the plain goes substantially sharp, then takes a second or so to stabilize after being struck. This makes it more difficult, both to tune and to PLAY in tune, than with the wound. (for me -- your mileage may vary)

All-in-all, the wound string sounds better, tunes better, and detuning (e.g., 6th open when A pedal is pressed) decreases substantially (about -6 cents wound vs about -18 or -20 for plain).

Made a believer out of me. Thanks, Ricky, for the suggestion.

------------------
Larry Bell - email: larry@larrybell.org - gigs - Home Page
2000 Fessenden S-12 8x8, 1969 Emmons S-12 6x6, 1971 Dobro

[This message was edited by Larry Bell on 01 September 2002 at 07:04 PM.]

------------------

Doug Earnest
The only Zum Keyless U12, Fender Cyber Twin