Knife edge changers - question

[Bent]

Georg, Richard, any interested parties:
What is (are) the main reason(s) for making a knife edge changer as opposed to a conventional all-pull changer?

[Georg]

My interest in them comes from...

1: knife edge is the coupling with lowest friction - ideally (although probably not achievable here) zero friction.
- low hysteresis.
- works best without lubrication.

2: lends itself well to near-linear action.
- reduces bending-stress and subsequent string break compared with regular roller-on-axle based changer.

3: fixed balancing point taking up string tension can be as large/strong as I want without affecting action, while a regular changer axle's OD can only be so large.
- I intend to make the fixed balancing point part of, and thereby as large as, the horizontal plate steadying the keyhead - around 100x200x8 millimeter - all load taken up horizontally.

4: pull-rod travel vs action (change in string tension) can be optimized for each string in the changer itself.
- IMO better and easier than to try to achieve the same action-control at bellcrank level.

5: changer scissors based on knife edges can be inserted/removed independent of their neighboring scissors.
- initial tuning of mechanics made easy.
- ease of maintenance.

[richard37066]

Bent -

I must agree with Georg on his first point: - the reduction of friction. You and I have discussed the mating surfaces problem and have concluded that the fabrication of razor-sharp knife edges is most likely not the way to go. Instead, a very short radius seems to be the most practical avenue of approach. Ideally, the contact area is but a single, fine line. Compare this with the traditional round axle/finger configuration to be found in most changers. Now, I realize that - if one wants to get truly finickey about things - a traditional round axle/finger also has (in theory) a single line of contact. Picture an axle of lesser diameter than the hole into which it is placed and forced to one side or the other. A theoretical single line of contact. However, the larger radii involved (essentially equal) invariably promotes a broader contact surface hence more friction. The prime culprit? The inability or the lack of desirability to produce precision surfaces. The knife edge obviates that necessity.

I have done additional research into the friction problem and, were I to craft a typical axle/changer, I would carry the surface configuration problem but one step further. Most would agree that surface finish at the macroscopic level leaves much to be desired. No matter how ya hack it, it is a rough finish - friction waiting to happen. A simple process called "electropolishing" would go a long way towards at least reducing those miniature "mountains" and burrs which are inevitably left after the machining process. "Electropolishing" can be thought of as the converse to electroplating, i.e., material is removed in an acid-based bath with the aid of an electric current. The surface finish is reduced to almost the molecular level with no sign of surface irregularities after having undergone the process. The basic equipment required to perform this "reverse plating" is no more intricate than the simple plating process, itself. It does require a bit more attention to detail but certainly not beyond the acumen of the average person. Further, when one views photomicrographs of the surface finishes attainable, one can only wonder why it isn't more universally accepted. I can hazard but one guess - the machined surface is "good enough". This is not an attitude without merit since a jillion PSG's have been made where things are deemed to be "good enough" and they work just fine. I, personally, am unwilling to stop at the "good enough" point when one further step would dispel all doubts as to whether undesirable friction has been, essentially, eliminated. The 'net has a wealth of information available for the asking.

The only other point that I can make in favor of the knife-edge approach to changer design is that of simplicity and maintenance. We have all visited GooglePatentsOnline and viewed a plethora of changer designs. Endemic to virtually all of those designs is what I like to call "monkey-motion" - "this" presses against "that" which, in turn pushes "the other" and on and on and on. If one eliminates the levers, springs, tabs, connecting rods, etc. from those designs then one is left with a basic changer consisting of four parts: - Changer axle, finger and two-piece actuators. Contrast that with the simple two-piece configuration of the Fields changer fingers in the Fender PS210 and the attraction becomes quite obvious - utter simplicity. Georg has posted a photo of those fingers in a recent post. It becomes readily apparent that no maintenance is required. I must issue a caveat here: - In our discussions we have seen that the fabrication of the knife-edge finger can be a little more tedious but not really objectionable.

In sum, it might be said that the extent to which the builder is willing to go is dictated not only by their perception of necessary precision but by economics, themselves. Georg and I are attempting to build "one-offs" for our own personal use. We, therefore, take license and adopt those things which, mechanically, we deem to be "better". If one is building for resale then the perception of the buyer must be taken into account. If the buyer is expecting to see gleaming end plates, monster changers and a forest of tuning machines then, by all means, give them what they want.

I have digressed and for which I apologize. I think that I've responded to your question in good fashion. I certainly hope so.

Oh. My knife-edge changer IS an all-pull changer. A modification of Fields' design.

Richard

[Georg]

Want to add that although my changer most definitely is a conventional All-Pull changer - minus rollers and axle, I am making room for a Push/Release action circling around the lower-return spring-loaded push-rod. Still tweaking details there, so my earliest sketches only show a push-rod.

One obvious reason for designing in a knife edge changer in my project-PSG, is that rollers on axles don't make sense the way I want to move the tuning gears - linearly. I would have to add even more parts to change circular movement into linear, so instead I chose to eliminate circular movement there. Hooking up regular changer scissors only isn't really possible without an axle, but the knife edge solution fits perfectly.

Regarding the knife edge ... I want them as sharp as I can get them made. "Blunt knife in a perfectly cut 90 degree seat", is the closest I can describe the ideal knife edge connection. Wonder how to harden the steel enough in just the right areas...



As for the conventional "rollers on axles" changer ... my version (if I were to build one) would be "loose rollers on too small (OD) axle". Making the axle OD around 1-2 millimeter smaller than the roller ID, will IMO change the way the roller moves on the axle enough to reduce friction to a minimum by introducing what I think Richard would call "more monkey-motion". Actually, IMO at least, sufficiently loose rollers will make conventional changers act more like knife edge changers, kept in position by string tension.

I will live as I preach when it comes to my regular PSGs, but not for the changers since there is no apparent friction-caused problem there. Keyhead rollers will have their axles' OD reduced though, to eliminate hysteresis and need for lubricants there.