Project: keyhead changer on knife-edges

[Georg]

Richard,

For the pedal mounts I will probably "go home to mama" and more or less copy what's on my Dekleys. Simple profiles I can get anywhere, and each pedal-mount occupies so little space that I can put heel pedals or heel pedal supports between the regular pedal-mounts.

Some counterforce can be added to the heel pedals to make them support the resting foot a bit, but it has to be an even counterforce to prevent "jolting" the heel pedal when pushing down.

Another problem is that it isn't possible to lift the knee while pushing the heel down, and that will cause problems the way I use my LKV - as a "switch". If I mount regular 1st and 5th pedals (for toe-activation) with heel-extensions, that problem should not be any greater than what I have on my 5 pedal Dekley now.

When I play my modded Dekley I always rock/move between pedals while pushing LKL or LKR and activating the LKV "stop switch", and this works pretty flawless with 4 regular pedals and tailored LKV. The 5th pedal causes some problems when I'm "in a hurry", but it works. Have to give pedal arrangement some serious thoughts and testing before mounting any on my project-PSG, as I don't want to relearn everything from 30 years of playing this contraption.

[Georg]

And then comes a tunable 2-pitch change for improved JI tuning (my preferred tuning principle)...


( larger drawing: http://www.gunlaug.no/att/div-2011/psg-var-6.png )

...which is relatively easy to build in as most of the pedals and levers are not connected directly to the bellcrank axles that pull on the changer on my PSG project. The sketch shows the mechanical principle for a 2-pitch change, not necessarily the mechanical solution I'll end up with.

On my E major / E9 PSGs I have found that only one activated change need 2 pitches: the raised B strings. Note that I raise both 5th and 10th string on both A and C pedal, so I only need one pitch-change mechanism to cover both. Note also that I use pull-rods for all raise/lower balancing - also known as "tunable splits", as screws locking raise/lowers to each other is a limited/limiting solution.

- By raising 5th string B to C# 3Hz sharp, and then lower it 3Hz using another bellcrank/axle that is spring-attached to the raising bellcrank/axle, that change goes beatfree into nearly all open and pedaled chords.
- When lowering G# strings to F#, the raised 5th string sounds 3Hz flat and beats with the F# at that frequency. By "eliminating" the 3Hz lowering, the 5th string C# goes 3Hz sharp and sounds beatfree in chords including the F#.

For E9 with F# on 7th string, that and 1st string is compensated in OPEN to stay in sync with the lowered 6th string and raised 5th/10th string, using pull-rods.
If I end up choosing my "Twin Tuned E major with chromatics", I only have to compensate 1st string for such changes since 7th string F# doesn't exist - 7th string is "Twin Tuned" one octave below 6th string.

[Georg]

More of the inner mechanics sketched up...

...as seen from the bridge-end.

This concept keeps bellcrank axles short (stability) and the inner mechanics open (configurable). As the sketch shows, there isn't much that gets in the way when front-and back-cover are removed, even if not all trusses are drawn in. When the neck is removed the inner mechanics will of course be even more accessible.


I will use spring-loaded pull/push-rods between the lever/gear/pedal axles and the bellcrank axles for changer pulls. This allows for precise stops for bellcranks, with as much flex in pedals/levers as I choose when setting the contraption in playable condition. Nylon bushings and a long plastic-insert in the square aluminum "tubes" that make up the frames, will secure all axles once in place.


The light gray pieces at an angle are tubes for inserting legs. Have not decided if I will fix nuts at the upper end, or just weld the leg-tubes closed and spring-lock the legs. These "leg-tubes" act as trusses between upper and lower frames in the corners.
Will need legs of more than "normal" length since they will go a couple of inches up into the tubes/frame. Apart from that there's nothing special here.

[Georg]

Another tidbit:

Looking back at the knife-edge changer scissors and thinking out loud...

...I am pondering whether or not it would be practical to vary the central connection point to, basically, get less "top-travel" for wound strings than for unwound with same travel for pull-rods. Ideally a certain-length pull-rod travel for for example the 5th and 10th string, using same-height holes in changer and bellcranks, should result in identical - in-sync - raise or lower for these strings.

With identical scissors I will have the same balancing "problem" for my changer solution as found on all other changer-end tuned All-Pull PSGs. Not much of a problem really, but especially if I choose "Twin Tuned extended E major" or "12 string Extended E9" tuning I will have to include gear-down bellcranks to achieve in-sync action, and this inevitably means that precise tuning becomes "delicate".

As I wrote: this "problem" is found on all changer-end tuned All-Pull PSGs, and I have pretty much ruled out the "Blanton solution" as impractical on a fully JI tunable PSG. I don't know if anyone intentionally "in-sync balance" changer scissors/fingers on regular bridge-changers.

One advantage of my knife-edge changer is that individual scissors can be taken out/swapped without having to dismantle the changer. So if I find the right formula for "graded" scissors, I may be able to optimize scissors for individual string gauges and tension, number them, and swap them around for optimal action while setting up.

[Georg]

Have continued balancing "wants and needs" for the frame and inner mechanics...

( http://www.gunlaug.no/att/div-2011/psg-var-7.png )

...and found it to very flexible. Since it has evolved past the description a couple of posts back, I'll add some missing pieces here.

Most of the frame-trusses are best screwed together, which makes the PSG easy to assemble and maintain.

The round opening along the backbone is not continuous. By having what in effect will be short tubes with wide spaces in between along the inserted backbone, I can insert one of the meter-long screws with self-locking nuts and adjust the slight bending-effect and holding-force the backbone shall put on the neck. I am pretty sure the backbone won't give much under string-tension with this arrangement, so body-drop should be minimal.

The gray curved inserts in the profiles steadying the bellcrank axles, will be lengthwise-cut soft-plastic water pipes. Have around a kilometer of those pipes on the farm so they are ideal (and cheap). Soft-plastic inserts will also dampen the frame sonically, which is important.

Will raise the lever/gear frame as high as possible, to make the PSG as low as possible along the player's side.



The black covers will be attached to the neck and act as resonators, and I want to be able to "bend" them a little to change the sonic effect while playing. Rounded yellow pieces along the frame front and back will be shaped pieces of rubber to keep the covers from touching the frame, so I can control the resonance by pushing down on the "armrest" while playing. The "armrest" is not a soft pad like on many PSGs, but the pretty slippery and slightly rounded shape of the metal-cover.

By having such a "playable cover" and a couple of extra PU-coils strategically placed under it and along the neck, I intend to widen the body-tone range of my PSG far beyond what is normal for regular PSGs and still have it sound quite normal when I want to.


As mention earlier in this thread, I want the frame and mechanics to be low weight so I can distribute the weight needed at strategic places for tone - mainly on the neck at or near the bridge. Right now I am looking at some thin-walled rectangular aluminum "tubes" outside my window, that are light weight and pretty strong. Must find something similar for my PSG frame.

[richard37066]

Georg -

Break out the graph paper and a BIG compass! As you've no doubt gathered already, the geometrical relationship between the three points ultimately determine the ratio of "raise" to "lower" movement at your movable edge. I have three or four such "designs" sitting in a notebook, each with a different ratio. Will have to make some really critical measurements before I settle on a given one. In my crude experiments, I've found that even a "three half-tone" lower on the 10th string of my C6 tuning is still much less than 1/16" (1.59mm) at the changer. I'm recalling you having said that you entertained some really radical raises and lowers but my guess is that the movement would not be significantly greater than what would be typical at the movable edge. Given this, and given the fact that the pivot points move such a short distance - "normally" - I question whether it is worth the effort to attempt a sort of "balance" regarding the pivot points in the changer. It smacks of a degree of precision which might very well be difficult to obtain in practice and which might well be realized in simpler fashion elsewhere. If we assume that a given copedant will be fixed then I would find it much more expeditious to have a "set" of both changer fingers and bellcranks in slightly different hole configurations such that the "in-between" points can be reached - if desired - adjusted, and then left for all time.


One other thing which has concerned me from day one about your design is the location of the upper "raise" positions. It would appear to be an awfully short lever arm thus requiring greater force to activate. The "lowers" have string tension to aid in the transition. Have you progressed sufficiently enough in the mechanical design that you can calculate the moments to be found at the pivot points all the way down to the pedals? I have - temporarily - split the raise and lower functions into smaller groups of ones, twos and threes (depending upon the design) and staggered along the edge. This should give me a little flexibility in attempting to "balance" things between the changer and the bellcranks. Just because the changer on my GFI has a total height of approximately 4½" doesn't mean that I can't increase that dimension by 1/4 to 3/8 of an inch if it means gaining more flexibility in adding an additional raise or lower in an effort to realize a given leverage.

As to any inclusion of a "Blanton" bellcrank I, personally, think that it's a bunch of hogwash! To reiterate something that you have said previously about balancing things and not being able to hear a difference, I find it a bit ludicrous to be able to make an adjustment to less than 1/32" or, even, less than a MM if one can't hear it! Further, one needs a screwdriver and a wrench to adjust a Blanton crank. Is this a time - and effort - saving feature? NOT. In viewing photos of the Blanton crank, I find one glaring fault with it and one which appears in many other designs. I refer to the inability to run a pull-rod THROUGH the bellcrank - such as is found in the MSA design (and others) so as to eliminate the need to bend a rod in "Z" fashion in order to clear the crank. I know, many folks have done it and it seems to work just fine. I just rebel at the idea. Give me a straight pull! Another thought regarding the "fine tuning" of cranks: - Will all players feel it necessary to include an expensive frequency meter in addition to a Peterson in order to tune up to the Nth degree? There are times when practicality and, even, actuality is thrown out of the window. A wide selection of actuating points at both the changer and the bellcrank should do the trick.

My last concern is one about which I must smile - and that is the possibility of ending up with an undercarriage which resembles a bowl of spaghetti such that one can't get a wrench or screwdriver down into the bowels in order to make an adjustment. But, then, you've already thought about that, haven't you? Smile, anyway. The fun continues.

Richard

[Georg]

Richard,

Thanks for the input. It is important that I get the details right before building, so I appreciate you ripping it all apart at this stage so I can assemble it again in my head and on paper. These processes can only lead to an improved construction.


In my experience with extended tunings, the short changer-movement tend to cause one problem in particular: lack of free-play between nylon nut and changer in neutral for lowest string.

On my "extended E" I have 3 G# strings and 3 E strings. Synchronizing raises and lowers on these "triples" was a PITA - actually impossible, with butchered and free-wheeling gear-down bellcranks. To get free-play on lowest G# and E strings I had to pull the bellcranks over so there's much too much free-play on the higher strings and pretty uneven travel-start. It works, and have for 20 years, but I would not mind having less travel on top of scissors for same pull-rod travel, for the lowest strings.


The upper raise holes do require high force and short rod-travel, but since this is mirrored on the relevant bellcranks the problem should be minimal. May require more pull-force than on most PSGs, but unless the pull-rods give/stretch or nylon-nuts pop off I don't think there will be problems.



I do have an option for fine-tuning bellcranks for sync'ed travel, in that I can position bellcrank axles out of line - place the bushing higher or lower than the others and even high at one end and lower at the other, along the profiles holding them. I can even make this "offset" adjustable if I want to, but since these axles are indirectly coupled to pedals/levers - doesn't matter much where a bellcrank axle is position with regards to where its pedal/lever is hooked up - I plan to have a few offset bushing-holes in the profiles to handle "impossible cases" if they pop up.

I do want straight pull-rods and adjustment via nylon-nuts at the end. Much less risk of skewing and/or stretching with straight pull-rods than with bended ones.


I am not the least concerned with what a tuner or frequency counter shows in real life - my frequency counter is for lab-control. I regularly tune my PSGs up to a quarter-tone off by ear - my "ear" isn't better with respect to accurate frequency without a tuning-fork, and play happily along as much off-fret as I have to to play in perfect tune with whatever tune I hear.

I tune out string-beats, which means I am only concerned with how well strings and changes are tuned together. Pushing strings with the bar to minimize beating is the norm, but I have grown a bit tired of wringing strings or avoid certain chords for certain pedal/lever-combinations. Thus, my 2-pitch changes will be incorporated in my project-PSG - to eliminate unnecessary irritation if nothing else.


As for the "spaghetti" mechanics... I have taken that into account from the beginning, and made provision for as much clutter as I can force-fit into the frame

All PSGs I have seen - including the WBS - have large frames/bodies that get in the way. Most only have access to mechanics from the underside, and all parts/details in them are designed and mounted for such limited access even if that means they can not be mechanically optimized.

My trussed frame with three lines of bellcrank axles can...

1: be accessed from all sides when covers are taken off.

2: access is even better with neck removed, which it will be during initial set-up.

3: two sets of five changers open up a space down the middle for playing around with pulls. Note here that the backbone can be removed separately, so a space of about an inch in width is totally open with a fully playable PSG during set-up and service.

4: slender profiles / trusses, of which most can be dismantled without frame falling apart, improves access.

5: in my pretty open frame all parts can be mechanically optimized, which amongst other things mean I don't have to "play it safe" with regards to shapes and dimensions.

May not be very visible in my sketch, but for instance the screws holding the pull-rod bellcranks securely on their round axles, are inserted opposite the rod-attach holes. Can be done on all PSGs, but those screw-heads sure aren't easy to reach inside a regular body without dismantling all rods. On mine I access them from above, with or without the neck in place.

Another detail is the "two sets of five (or six)" changer scissors. Apart from opening up space down the middle, the narrower changer-sets makes it easier to prevent "give" without oversizing construction.


The pull-rod bellcrank axles will be arranged and spaced optimally for pulls, with no regards to which pedals/levers/gears that operate each. This leads to shorter pull-rods and more exact and reliable pulls in line with the changer scissors.

The entire pedal and lever/gear frames along front and back can be taken off with no adverse effect on action and tuning - the PSG frame just isn't as strong without them. Since I will have stoppers on the pull-rod bellcranks, all pulls can be tested and tuned precisely without having actual pedals/levers attached.


Have probably forgotten something important in here, so just blow my design apart again. I am always ready for another assembly/tweaking process.

[Georg]

Two changer-variants...



...and at the moment I'm not sure which one will work best.

[Georg]

And then...
...I return to the KISS concept...

Simple steel knife edge parts fixed inside the U-shaped aluminum raise/lower scissors (raise inside lower).

The two (colored red) steel parts on top are relative easy to shape ... identical, or in two/three variants to optimize for shorter action for low-octave strings.

The connection points for the "fixed" and "action" knife edges above the aluminum U-shaped raise and lower will be full width (8.5 millimeter), while the lower part of these steel parts will have width for tight fit inside the U-shapes.

The white "holes" can either be used to bolt the parts in place, or I can simply punch in the aluminum U-shapes to make sure the steel inserts stay in the exact right positions during the assembly process. Parts will be permanently epoxy'ed together under heat and pressure when tweaked and ready.


Will add a guide at the bottom, by extending the steel insert for stopper on the lower downwards as a narrow "lip" and add a fixed guide plate like on most All-Pull PSGs.

Lower-return spring will (as previously presented) be in the form of a push-rod from spring-loaded bellcranks. Plan to combine raise-helper and lower-return spring in an adjustable configuration so I can take up a defined amount of string tension and balance out raise/lower pulls.

[Georg]

And then changer parts are taking shape...

...and the "raise inside lower" configuration should be clear.