rollers for keyhead changer...

[Georg]

As I look around for suitable parts for my keyhead changer project, what do you all think about using these ball bearings as keyhead rollers?

[Bent]

Georg,
I am leery of those ones because of 2 reasons: The OD being .678" is on the large side. I believe that the larger the OD, the more string has to touch it, and this is detrimental to sustain. Very minuscule but still there. Also, without a groove in it, the string will be touching only on the bottom. The ideal thing is to have the string touching both sides of the top part of the groove.

I believe in this so strongly that if I had a way of grooving the changer fingers, I would do that end as well.

[Ross Shafer]

I agree with Bent regarding the OD and the groove. There's lots of far smaller bearings available that could be used. That said,since the rotational movement is so miniscule, I'm not convinced there's much (if any) friction reducing advantage in using a ball bearing instead of the standard "plain bearing" method.

I do think there is the potential for what could be an affect tonewise (could be good or bad)...With a bearing at the nut, the strings vibrations are being sent through all the parts used in the bearing. Since the bearings has numerous balls in it and they are not all loaded at any given time some will be loose and could actually vibrate on their own to a small degree which could help dissipate (dampen) the strings vibration (think handle bar weights at the ends of a motorcycle handlebar or dampers on an archery bow). With a standard "plain bearing" type assembly the strings vibrations travel through the roller's single line of contact (if the roller turns it cannot have concentric contact) with its shaft....no extra unloaded parts are subjected to vibrations that they may dampen.

I know this is splitting hairs, but we're talking about very small amounts of vibration to begin with that create the tone from a string. It seems to me that offering those vibrations the most solid foundation with as few damping opportunities (moving or flexible parts) is the way to go to preserve as much of the strings vibration as possible...an extreme of this would be the solid nuts used on some keyless steels that are purported to offer better tone transfer. I've heard it said that the old ShoBud (or was it Emmons...) keyheads that did not use rollers of any kind offered better tone....this makes sense to me, but the friction over those solid aluminum nuts would definitely be an issue with the amount of string movement found int those strings with a longer span from nut to tuner.

more food for thought....

[Georg]

Bent, Ross, thanks a lot for the input.
I agree on all points except degree of movement. Keep in mind that my keyhead changer will pull and release individual strings some +/- 4 millimeter over those rollers, unlike the max 1/2 millimeter movement caused by string stretch/release for regular bridge-changers. In short: think backwards

OK, I am going through all factors in my search for the near-perfect keyhead rollers for my design, and am open for more inputs from y'all.

My thinking (so far) is that as long as the OD of those ball bearings is less that of a bar - 7/8 inch or so, the OD should in itself not reduce sustain or cause buzz more than a bar at first fret does. A small OD matters more at the bridge end on my design, where an effective OD (roll-off for strings) of around 3 millimeter on a fixed bridge is my goal.

I am mindful of the potential for noise from vibrating ball bearings - the outer casing seems to be relative light. Also not sure if these ball bearings have thick enough outer casing to allow for a minimal groove, as a groove will be preferable to even out string-height, guide strings and increase friction.


The fixed nut-bar on my recently acquired keyless GFI surprised me by causing intolerable hysteresis. Greasing that bar eliminates hysteresis for a minute or two - just long enough to determine the cause but not to solve it. With only some 8 to 15 millimeter string-length past the nut-bar, audible hysteresis at that point should be impossible, but it is in fact very audible. That bar has to go - be replaced by rollers, or the entire GFI is just a good-looking, good-sounding piece of garbage that I have no need for.
The experience with the GFI makes me mindful of "miniscule movements" at these critical points as I am designing my own.


I would like to use much smaller ball bearings or rollers at the keyhead end, but that is mostly because I want to be able to adjust them individually both for string-height and string-length(*).
Such an arrangement means the rollers must be narrower to allow for individual mounts, but in my thinking they do not necessarily need to have a smaller OD than a bar.

If I find the right ball bearings, then fine. If not I will probably fall back on my "PackMan rollers on knife-edge bearings", as they will be fairly massive - should support sustain well, can be grooved, allow for +/- 6-7 millimeter movement, have near-zero friction, and can be mounted individually on adjustable bearings the same width as the roller - almost zero space between rollers. Such "PackMan rollers" also makes maintenance easy, as they are held in place by the individual string and can be taken off for cleaning and adjustment when changing strings. On the negative side: "PackMan rollers" can be lost since they are only held in place by the individual strings...


(*) I have found that individualized string roll-off points at those nuts improves overall intonation on PSGs - same as on my regular guitars BTW. This means "tuning the string-lengths by individual positioning of the keyhead rollers." A tunable variation of roll-off points of as much as 3/10 millimeter makes it possible to tune out bending/vibration characteristics for plain/wound strings of various thickness while taking actual tension for the individual string into account.

[Bent]

Georg,
I agree with adjusting string height. But in addition to that I wanted to say that I have looked at and abandoned individually height-adjustable rollers.
I didn't like the necessary space in the hole for each roller lifter. It invites buzzing and potentially cuts down on sustain and clarity. I have started experimenting with making each roller its own diameter to compensate for the different string gauges.

[Georg]

I have started experimenting with making each roller its own diameter to compensate for the different string gauges.

Have you contemplated making eccentric rollers, which can be turned to even out string height? Would guess around 3/10 millimeter eccentricity should do (if OD is reasonably large, half an inch or so), with a "zero" marker at the top edge on the half-way point to see how far they are turned either way for up or down.

I plan to pack the necessary number (8 for a 10 strings) of around 100 millimeter long levers tightly horizontally on the top-plate, and mount whatever rollers I end up using on top of those. Since "top-plate" in my case is the aluminum/steel "T" support for the neck/keyhead, I plan to hide those levers under the hollowed-out visual neck and adjust each lever in height, and length, with screws offset from the strings close to the rollers at the keyhead side. Outer rollers will be fixed in height/length.
May do something similar to get adjustable position/string-length at the bridge-end, without rollers and height adjustment.

[richard37066]

Georg -

I may just have some good/bad news for you.

Read your criticism concerning the GFI and it really furrowed my brow. I've had my D10 for 3 1/2 years now and have never experienced a hysteresis problem. EXCEPT - when I tried to get too fancy and attempted to make mechanical adjustments which were too close for the degree of precision of the instrument. My bad. Live and learn.

Took a break from reading your post, went to the guitar, cranked everything up and gave it a good tuning. From top to bottom, pedals and levers. Just needed tweaking since my instrument has been hassle-free from day one. Excercised the pedals and levers while making note of the return point of each. No more than a couple of cents difference all the way around. I can't hear a difference of only a couple of cents. Maybe others can.

I'm thinking that the solid nut is not your problem. I also know that, should you tend to agree with me, you'll look in all of the right places and for the right things. One of the things that I have noticed, in times past, is that when changing strings and cleaning things up there has been a small spate of hairline scratches in the solid nut - most likely from that minute string travel across it. The material? The plating? Don't know. It shouldn't make a measurable contribution to a hysteresis problem.

In any case, just had to give you input regarding my experiences with the instrument.

Can't make a contribution to your dilemma regarding the roller nut(bridge) except that I'm reminded of some recent threads by Ed Packard on the SGF regarding Zirconium (?) tone bars. It's one tough material. The surface finish on them was immaculate even when compared with expensive "name -brand" bars. You might look into not only the bars - for basic information - but the material as well. Maybe you can use it.

Respectfully,

Richard

[Georg]

Hi Richard,
I have two Zirc-bars - black and white - one for use here and one back in Norway. They're fine but I'm not sure if the process to turn zirconium into anything useful is within my reach.

About the GFI: mine still has strings from previous owner on, and that doesn't help. Especially string 6 gives quite audible body-drop and hysteresis. Got new Jagwire sets the other day so will restring the GFI soon and tune it optimally inside-out and see how it behaves.

The more I think back and forth about those rollers, the more the "PacMan on knifeedge" roller concept seems to be superior to all other options. "KISS" still works best, I guess.

[richard37066]

Georg -

Yer a man after my own heart!

K.I.S.S

Richard

[Ross Shafer]

I've not had any personal experience with a solid nut on a steel and have wondered if indeed they work as well as rollers as has been claimed by some. I am surprised to hear that the performance on Georg's is so egregious....while the performance on Richards same make of guitar is fine. An interesting puzzle.

I'm keen on hearing whether or not fresh strings have a positive effect. Old strings can get weird and cause tuning/intonation/noise probs sometimes...I learned that long ago on regular old 6 stringers.