Page 1 of 2 12 LastLast
Results 1 to 25 of 34

Thread: F4 design evolution/solutions

  1. #1

    Default F4 design evolution/solutions

    As I’ve been modifying my own templates lately I've become more and more curious about the way the F4 developed from the introduction of the 2 point model around 1912(?) and specifically the adaptations to the narrow neck and the new drastic offset of the fingerboard in the Loar era. I’m also curious about the specific manufacturing procedure they followed (if anyone actually knows). It is clear there was lots of active designing and redesigning of the templates and molds, the glaring example being the shortening of the upper point after 1915 (or is it that the cutaway is less deep?) Of course this is ultimately about the F5, but I think it is important to study its genetics.

    To preserve the balance of the scroll, the offsetting of the fingerboard relative to the center of the head block was a problem they were acutely aware of and had a number of ways to solve, with varying degrees of success. I see examples with what appear to be:

    A tighter radius below the button/a narrower bell shape of the head block, leaving less meat on the treble/problem side (were there variations in the forms and is there proof of instruments with different shaped head blocks?)

    The introduction of a parabolic shaped button rather that the previous ogive that screams CENTER

    The steering of the ends of the binding into the button creating the illusion of harmony

    The concealing of the misalignment of the center joint and button in the black of the sunburst

    The introduction of a one piece neck with no stripe to suggest a true center

    And sometimes no adaptation at all and a button that looks like it’s sliding off into the scroll.

    A few things stick in my mind in terms of procedures (correct me if I’m wrong, of course): the mortise was established with 3 bandsaw cuts earlier on, and by the Loar era had a dedicated router setup. And the necks were fitted to the rim sans plates, then the back was glued on, and finally the top—this doesn’t mean the neck was glued in first, but that it was set with templates rather than eyeballing an alignment. And lastly the final piece of binding went around the button, which to me implies the neck was glued in after the assembly and full binding of the body.

    Having built a lot of instruments and having also worked in a small factory setting I know that process is everything, and trusting a more or less skilled worker in the setting of a neck with a stripe down the middle, a pointed and ridged button and a fingerboard set way off to one side doesn’t seem like a very probable consistent factory solution. And that we love to recycle and adapt jigs and forms that already exist, with "good enough" as an acceptable outcome. I can only imagine that there were two independent operations: a jig to fit the (still somewhat wide) neck more or less on center (with the requisite room for error, cutting, shimming), and then a template with the shape of the fingerboard that indexed into the mold. This shape was drawn on the bare neck, and the neck was worked down to final size. I imagine this was also the procedure in the teens era, but the wider fingerboard made for a much more centered looking neck in most cases. In all of this there is the possibility of screwing up the alignment of the neck in the first place, screwing up the shaping of the neck, screwing up the positioning of the back and its centerline, screwing up the positioning of the top and the f holes relative to the bridge, etc. But in the end it’s all about achieving harmony rather than symmetry.

    I would love any input from the resident historians (shoot me down) and also to hear from other builders who have worked through this problem themselves and the adaptations you have made (it looks like everyone crosses this bridge at some point). I have tried both ways, and I see nothing wrong with a redesigning of the head block to allow for a harmonious fretboard position relative to the scroll AND a harmonious button position, working with only one centerline. If you are building Loar copies, I can see the issue.

  2. #2

    Default Re: F4 design evolution/solutions

    Another thing: the F4 is different in that the inner piece of scroll binding wraps up the head block (no riser block) and the fingerboard binding obviously continues this line. To interrupt it would make no sense visually, and I see the offset fingerboard to be an adaptation to the F4 design rather than the F5, in terms of the evolution of the whole thing. In other words the bass side of the fingerboard always needed to align with the scroll binding, whether the fingerboard was wide or not. The F5 seems to take this on as more of a vestigial feature (like the jig was already made) rather than one that makes perfect design sense on its own (though I do think it looks better).

  3. #3
    Registered User sunburst's Avatar
    Join Date
    Nov 2003
    Location
    Kentucky
    Posts
    15,863

    Default Re: F4 design evolution/solutions

    Quote Originally Posted by fazio View Post
    ...we love to recycle and adapt jigs and forms that already exist, with "good enough" as an acceptable outcome...
    Same conclusion here. There are lots of things in the 20s F5 design that look to me like they happened through adaptation of a new(ish) design using existing tooling and procedures.
    I simply sat down and re-drew the whole thing, and in fact, after roughly 30 years of building point&scroll-bearing mandolins I still make slight changes to the shape trying to make it better. As a perfectionist by nature I have to realize that "perfect" is not possible, but it still tantalizes, and therefore I will probably be attempting to improve the thing (to my eye) as long as I can pick up a piece of sandpaper.

  4. The Following 2 Users Say Thank You to sunburst For This Useful Post:


  5. #4
    Teacher, repair person
    Join Date
    Oct 2017
    Location
    Southeast Tennessee
    Posts
    4,078

    Default Re: F4 design evolution/solutions

    I don't build, but I do repair, and I've seen quite a few F-4's. I've even measured a few of them, and found that the body measurements can vary slightly in width and depth.

    The nice thing about building to a general pattern is that you have some flexibility in size, shape, and assembly methods. The biggest requirement is that an instrument has to fit in a case. Knowing what the factory assembly methods were can be useful, but what is most practical for a factory may not be as practical for an individual builder. The bottom line is you get to choose your exact dimensions and your preferences for order of assembly.

    The old violin makers were known to change their patterns and graduations from time to time.

    If you choose to experiment with something, it might result in something that's really good, or it might not work out. This is how we learn, and also how we advance the craft.

    Martin was in the habit of pre-fitting necks to their bodies, but finishing them separately before the necks were glued in.

    There is evidence that Gibson often finished their guitars after the necks were glued in, but I can't say whether or not the mandolins were done that way.
    It is not unlikely that their assembly practices varied depending on the period, the type of instrument, and perhaps even the specific model.

    I will guess that there were at least a few differences between the way an F-4 was assembled in the mid-1910's and they way it was done 20 years later. There is evidence that as time went by, Gibson made attempts to make their instruments faster and easier to manufacture.

    Whatever you do, strive for an attractive instrument that is light enough to sound good, but strong enough to hold together for a lifetime.

    You have a lot of flexibility though. Some instruments have a more feminine appearance, some are more masculine. Either can look good.

    I like hide glue and varnish for mandolins. It takes a bit more time and trouble than Titebond and lacquer. Either can produce good results. How does one choose? How quickly and easily do you want to get the carpentry done, how durable do you want the finish to be, and what kind of a look do you prefer??

    Give two experienced builders the same set of plans to use as a reference, and you'll get two instruments that are different.

    One way I stayed out of trouble during the height of the epidemic was to open up a Goya 12-string guitar that had good wood in it but sounded bad. It needed to be "de-lumberized." I had always wondered what I call the "V-bracing" pattern that Gibson briefly experimented with circa 1929 would sound like, so I pulled out half of the bracing, replaced it with my impression of what the new pattern should look like, and regraduated the rest. It came out quite well and I learned a lot.

    Total dollar cost: $150 + $60 for the guitar, perhaps another $100 for wood, fret wire, bone, bindings, and a decent set of tuners. Total labor cost? Who knows. I did get some sore joints and scrapes and abrasions when I found that my new bracing was still too heavy and I had to re-graduate it through the soundhole, but now I've got a pretty good 12-string and I've learned a lot about bracing.

    Click image for larger version. 

Name:	ts4-front.jpg 
Views:	63 
Size:	1.82 MB 
ID:	201427 Click image for larger version. 

Name:	062020ChuckrowGuitar-001.jpg 
Views:	66 
Size:	651.6 KB 
ID:	201428

    It's hard to tell from the picture, but the original bridgeplate was 3/8" thick, 3" wide, and 14" long, and the ladder braces were very stout. The picture of the new bracing was taken before re-graduation from the outside. I thinned the new braces considerably, tapered the V's and the front tone bar, and took about half of the rear tone bar out. The guitar sounds good, appears to be holding together, and the neck is solid as a rock. We don't see mahogany that stiff any more.

    If I do another one, I'll change the rear tone bar to the opposite angle a la Larson, and perhaps move the wing braces to a different spot.
    Last edited by rcc56; Jun-08-2022 at 1:09pm.

  6. The Following 2 Users Say Thank You to rcc56 For This Useful Post:


  7. #5

    Default Re: F4 design evolution/solutions

    I guess I believe that some of what we take to be immutable design elements are really just relics of manufacturing, but I also know that the overall design was highly thought out, cohesive, and occasionally changed with the goal of improvement (like tilting in the tuners or slimming the neck). The mandolin plummeted in popularity, the F5 was a flop, the advancements stopped in their tracks, and the later changes were meant to make them cheaper and faster if someone just happened to want one. The F5 is like great musicians dying at 27. I think studying the trajectory of the F4 is instructive because it allows us to imagine if the F5 had been given the same resources as the 1920s went on. I think they would have changed the shape of the head block after a few years of twisted alignment, since they clearly had a history of responding to similar issues. I’d love to hear what peoples’ suggested tweaks are! John I think your mandos are really elegant, like the F5 was around in 1912 and had no idea that the blocky Art Deco era was around the corner.
    Last edited by fazio; Jun-08-2022 at 3:22pm. Reason: Clarity

  8. #6
    Registered User sunburst's Avatar
    Join Date
    Nov 2003
    Location
    Kentucky
    Posts
    15,863

    Default Re: F4 design evolution/solutions

    There are design elements in many post-Orville mandolins of the Gibson corporation that puzzled me for a long time pre-internet. When good quality photos of many earlier instruments became available at the click of a mouse a lot of it started to make sense in that I could see how they got there, but they later became more-or-less vestigial because of tooling, accommodating subsequent processes, or whatever manufacturing reason.
    I agree that the F5 looks like work in progress in many ways, and given time and more importantly, desire, it may have been subsequently improved if it had been selling well.

  9. #7
    Teacher, repair person
    Join Date
    Oct 2017
    Location
    Southeast Tennessee
    Posts
    4,078

    Default Re: F4 design evolution/solutions

    Gibson's early periods can be loosely divided into sections:

    1. 1906-1920-- Birth, early development, and mandolin boom. The company is born and grows.
    2. 1921-1929-- Changing public tastes. Mandolins fade, tenor banjos boom and fade, steel string guitars develop and grow. The company learns to adapt.
    3. 1930-1941-- Depression, survival, and partial recovery. Gibson almost goes under. Costs are cut, products and manufacturing are simplified. Gibson starts to recover and once again introduces fancier products, but . . .
    4. 1942-1945-- Wartime. Survival and simplification once again become top priority as shortages in materials become prevalent and the labor force changes.
    5. 1946 and later-- A new beginning.

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

    How does mandolin production reflect these changes? Two A-4's tell part of the story.

    I have a 1917 A-4. By this time, Gibson has figured out how to build their mandolins efficiently and well. A circa 1932 A-4 that went through my hands represented the effects of the Depression. With two exceptions, there was little difference this mandolin and the 1917 instrument except for 1920's developments, such as a one-piece neck with truss rod, adjustable bridge, and ungrained bindings. The exceptions?

    1. The fingerboard was slotted so inaccurately that the instrument could not possibly play in tune and I had to replace it. A likely cause for this instrument leaving the factory with such a glaring defect was that in 1932, Gibson let the majority of their most seasoned workers go to reduce payroll costs.

    2. This mandolin was one of the last A-4's built. The model was soon discontinued and replaced by models that could be built and sold for lower costs.

    By the end of the 1930's, most of the few mandolins coming out of the factory were Kalamazoo models with flat or pressed tops and backs. Gibson spent the 1930's simplifying the designs of their instruments and changing their product line to appeal to the changing tastes of the public. This meant cutting production costs drastically, and offering products that could be built and sold cheaply to a public who was no longer interested in or able to afford the classy mandolins that had been the company's original bread and butter.

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

    The war changed everything again. Like the other survivng companies, Gibson severely reduced their product line and made due with whatever materials they could get their hands on to get instruments out the door.

    By the end of the war, Gibson's only surviving mandolin model is the A-50. They will not build another scroll model until 1949, and these will be very different from the F-5's of the 1920's. Gibson emerges from the war primarily as a manufacturer of guitars.

    Half of America's major early 20th century string instrument manufacturers did not survive the depression and the war. Only Gibson and Martin regained their strength after the war. Epiphone, Vega, and National survived the war, but never really recovered and eventually went under.
    Last edited by rcc56; Jun-08-2022 at 11:32pm.

  10. The Following 2 Users Say Thank You to rcc56 For This Useful Post:


  11. #8

    Default Re: F4 design evolution/solutions

    Quote Originally Posted by sunburst View Post
    I agree that the F5 looks like work in progress in many ways, and given time and more importantly, desire, it may have been subsequently improved if it had been selling well.
    I think the era was important as well, and putting R&D into a flamboyant art nouveau design in the new jazz era made little sense. It’s like tweaking a 20 year old car design.

    So here’s a proposal: a problem is that the button can look like it’s sliding off into the scroll (or that the bell shape of the upper back looks like it’s leaning out to the left). Since closing up the scroll any more isn’t an option, it makes visual sense to me to tilt the weight of the bell shape, to lift the bass side shoulder and have a slightly more precipitous drop on the treble side—without the button it would look asymmetrical, like it isn’t even a bell anymore. Also, lowering this entire shape a bit allows some more room coming out of the scroll and can help to avoid the sometimes sharp bend the binding takes as it rises into the button. It appears some workers were trying to achieve this look if they cared enough, but this was a bottom-up fix rather than the official design coming from the drafting department (which is what interests me more).
    Last edited by fazio; Jun-09-2022 at 2:40am.

  12. #9
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    I'm a bit in a hurry so without reading most stuff posted above I'll delve into this favorite topic of mine...

    Construction details of vintage Gibsons (notably F-5) always fascinated me and while I was researching and collecting material for my drawings I found following (I'm entirely going from memory so some stuff may not follow any time line or logical steps).
    (Much of this I already posted here on cafe in various threads on vintage Loars.)

    The biggest part of the changes in design came hand in hand with new tooling and methods when new body shape was introduced.
    I guess they had several body froms that were not identical and especially around the introduction of Loars new form appeared like you mentioned with a tiny bit less "cutaway". I haven't noticed many oval hole models based on this form, I think I've seen some F-4s made on this form but never paid too much attention to this detail. Perhaps they took one of the (several) F style moulds and dedicated it to the new master model.
    The three pointed had many parts handcarved that later were machined using various fixtures.
    Neck joint that was previously handcarved and not symmetric in any sense was now cut with three cuts of table saw (or similar cutter that left 3/16" or so wide squared cuts - so definitely not a bandsaw) This was most likely done on assembled rims without plates but with the riser block. I think it was not done on blocks before ribs were glued on as the original form has notches for wedges to push ribs to block during gluing right in the dovetail area. Forcing wedge to rib against empty dovetail would possibly lead to cracked ribs so I assume the dovetail was not cut at this stage. I'm 100% confident that backs were glued on first and tops were attached after that so the bare rim was cut for dovetail. I don't know when thy changed for a router (shaper) setup to cut dovetails but none of the Loar dovetails I've seen shows the three cuts. I guess they glued on the riser block using a template/fixture and the flat edge of the block served as reference point for all later processes like the dovetil cutting plate alignment and position of scroll. The dovetail was pretty much centered on the apex of the body and so was the back button both on ovals (pointy shaped button) and Loars/ later ovals (parabolic button).
    The neck was fitted to the body while it was still unshaped square block and I suspect the dovetail served as reference on the neck to shape the "handle" and inlay the center strip (or perhaps vice-versa if the nack cutout and inlay was done first to neck blank and the black strip was reference for dovetail). Certainly the black strip (and maple inlay) and also later truss rod cut were centered with the dovetail.
    later....
    Adrian

  13. #10
    Registered User Timbofood's Avatar
    Join Date
    Mar 2007
    Location
    Kalamazoo, MI.
    Posts
    7,487

    Default Re: F4 design evolution/solutions

    Where’s my popcorn!
    Timothy F. Lewis
    "If brains was lard, that boy couldn't grease a very big skillet" J.D. Clampett

  14. #11

    Default Re: F4 design evolution/solutions

    rcc56 -- good job on the Goya 12 string.

    I'm a fan of 12 strings, both acoustic and electric and have owned many, but finding one with low action that plays like "butta" is a tall order, in my experience. My current favorite is a made in Korea mid-level Alvarez with a repaired headstock that amazingly plays better than most Les Pauls! On the workbench (when I find the time) will be a 60's Harmony Sovereign 12 string that sounds great, but has high action, along with a floating bridge and trapeze that I'd like to convert to a pin bridge. The ladder bracing doesn't bother me as the guitar sounds very rich already, but I know some people like to convert them to x-bracing. Has really nice looking spruce and mahogany, IMHO.

  15. #12
    Teacher, repair person
    Join Date
    Oct 2017
    Location
    Southeast Tennessee
    Posts
    4,078

    Default Re: F4 design evolution/solutions

    Thanks for the kind words. I'll mention that the new braces have been significantly re-profiled since the picture was taken.

    Before you convert the Sovereign to a pin bridge, check to see if the bridge plate is suitable for that configuration.

    Hard maple, 1/8" thick is best. It should extend ~1/8" beyond the front and back edges of the new bridge. For a ladder braced guitar, it is probably advisable for it to also extend a good ways beyond the side edges of the new bridge. If the existing plate is spruce, it should be replaced, otherwise the string balls will tear through it.

    We'll need to start a new thread if anyone wants to pursue the subject any further-- we've gotten a long way away from oval hole Gibson mandolins.

  16. The following members say thank you to rcc56 for this post:


  17. #13
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    Part II:
    I forgot to mention how they prepared the body blocks. The neck and bottom blocks were cut with grain direction parallel to ribs most likely on shaper using templates. The point blocks had grain running perpendicular to body plane, in violin style. I'd guess they prepared long profiled triangular crossection sticks of mahogany (shaped with dedicated moulding planes or shaper cutters) and just sliced them into the blocks.
    Once the neck blank was fitted to the rim they glued on back and top (in this order - you'll see clean back-rim seams inside but you'll see drops of glue on the top-rim seams) they used template to mark scroll shape and cut it on bandsaw all through the whole body. Then they did the binding. When body was ready to meet the neck the neck blank was inserted ointo dovetail and position of fingerboard marked with relation to the crosspiece AND scroll. They always placed the board so that the gap between edge of board and scroll was same as thickness of the cutout within the scroll (~3/16"). That was done on ALL F models, ovals or F-5's. They marked the board on the neck and didn't even care about perfectly centered trussrod on the upper end, often the upper trussrod nut is 1/16" or more off center to get it to project to body center between f holes for good bridge position.
    This resulted in the offset position of neck when viewed from front, but from back the things are quite well centered (the cutout of scroll makes the scroll side of body curve appear asymmetric but the part where neck meets the body is quite symmetric and the button centered).
    The offset may be more visible on Loar era mandolins with narrower boards since the bass side of board was the constant reference point, the treble side moved left a bit more and exposed more of the riser block fillets on the treble side.
    I believe the neck shafts were shaped using shaper before the headstock wings were applied and headstock shape finalized. The headstock was finalized AFTER the neck was glued to body. The procedure was like - square neck blank (on mahogany necks already with maple insert) pre-fitted to dovetail on rim - after body is finalized, block angle and fit checked and marked for position of board- neck handle shaped/ routed for black inlay strip but the headstock part left a bit oversized - headstock ears added - neck glued to body - board surface planed - now the final position of headstock overlay (already inlaid) is marked and top surface of headstock finalized and fingerboard and overlay applied.
    This order may sound chaotic but they were factory (and some clues of workflow are visible when you disassemble the instrument) and single handmaker would carve the parts to closer tolerances before putting them together so he would choose different order.
    to be continued... (perhaps :-) )
    Adrian

  18. #14

    Default Re: F4 design evolution/solutions

    Adrian the neck block interests me most because it has to be perfect every time in order for the production fitting of necks. Everything else on the body is allowed to be a little wonky. Maybe you or someone else has proposed this before. I think the block was bandsawed out with the step down for the rib, the rib glued on, and then this was taken to a waterfall sander with a pin set up to fit in the hole in the back of the block. It was spun on the pin making a perfect radius and flat and square with the rib in place. Is there ever any evidence of the rib being thinner near the neck? When I worked in a furniture production shop I watched them do this with tables all the time. This was an adaptable and repeatable (and easily modifiable) setup. I can imagine another setup where the concave radius for the neck is cut using a similar idea, with holes that take a pin. The neck is held in place, and the jig rotates and swings the heel into the blade, then switch to the other hole and do the other side. This was another table making set up I saw. (I guess they could have also used a shaper for this part but my idea seems safer…)
    Last edited by fazio; Jun-13-2022 at 7:54am.

  19. #15
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    I don't understand pins and holes in your post... They didn't use any pins or holes in the block, just dovetail. The rib went all around the block and into the scroll, didn't end with a step near centerline like some available drawings show.

    If you mean that the neck attachment area was sanded to a circular curve by rotating around a pin while sanding on a sander there wouldn't be enough space for the backing support near the scroll. What kind of blade/tool are you suggesting here? When you assemble the rim on strong form this requires no more than light filing or sanding to smooth out the curve. I was talking about the bare mahogany blocks and their preparation using shapers or pin routers and templates.

    Ribs are typically a bit thinner near their ends which are near the bass side crosspiece. That anly suggests some extra sanding of the scroll cut-out.
    They could most likely use pin routers or shapers for blocks. It's simple and effective.
    The neck dovetail doesn't show much toolmarks inside which suggests the curved machined surface was very smooth and mostly the "tail" part received some adjustments for fitting.
    Adrian

  20. #16

    Default Re: F4 design evolution/solutions

    I’m not saying that I know that it was done this way, just that it seems like the simplest way to machine two matching curves across a whole range of sizes and styles of instruments would be to use a perfect radius right where the neck joins, and the simplest way to achieve that is mechanical, ie just let each part spin on a pin (rather than a series of shaper heads for the neck of A,F,H,K 2-5 styles). I don’t disagree that the blocks could have been done with a shaper but I think there is more to it afterwards. The blocks do always have locator pin holes that look to be about in the right place but I thought these are sometimes slightly off center. But as long as it could index into a jig. And you could definitely go into the narrow part on a vertical sander with the guard removed or a slight modification. I know what you are saying that you could just file it but by the time you get to a K4 with a riser block I think it would take more to ensure a perfect shape and I think they wouldn’t accept that in a factory production setting. And flatsawn maple on a mandolin with water based glue can already cause plenty of inconsistencies even if the two backing surfaces are very solid. Do you know at what point and how exactly they cut the dovetail mortise? Was the entire rim assembled and in the form which then was run through a series of saws, or could it have been the block on its own? (I know what you are saying about the wedges for gluing the rib and worries about cracking the block but they could have easily used a spacer since it was a consistent machined shape). I’m not trying to challenge you at all, I just want to imagine a viable (even if it isn’t entirely correct) semi-large scale early 1900s factory production process for a whole series of fairly complicated instruments and I have a lot of unanswered questions about the neck joint.

  21. #17
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    The female part of dovetail was cut with dovetail router, much like most guitar makers do these days.
    I'ts impossible to tell when exactly but the the earliest logical moment is when the 'S' rib AND riser blocks are glued to the block. That is, BTW, how Mr. Gilchrist does it and he is quite high efficiency oriented guy, IMO. Some folks (and I think Gibson did as well) cut it after the whole rim is together but that requires longer tool. The male part on neck was cut with dedicated head or bit in two simple cuts. Again Mr. Gilchrist uses such setup, though he uses overhead router and not shaper with head. You could use a custom head in table saw as well. Table saw was very likely used earlier to cut the female part but later the three distinctive cuts disappeared that's why I assume routing the dovetail.
    Remember this was decade(s) before handheld electric routers came to factories and many(most) machines were still large units driven by belt from big shaft overhead. I think they did most rough work on dedicated big table routers or shapers and table or band saws.
    With this machinery you need only several dedicated tools (bits/ heads) and series of fixtures with templates for each part. The f style body parts were interchangeable and so were guitar and mandocello parts etc so they didn't need all that many different fixtures and tools.
    Routing dovetail into block before rib assembly is not good was as you would need to cut thorugh ribs and the riser block after assembly to clean out the dovetail and that alone would take more time than just cutting it whole at once after assembly. And especially after the assembly you already know exact position of block in the form and can mark reference points for next steps (I assume the riser block itself acted as reference for neck joint cutting). I'm not aware of any reference pin holes in blocks, just one cut-off nail for gluing back and top (F-5 had no nail in top probably because top is too thin there so it could crack).
    Regarding the precision, they always conted with some (minimal) hand fitting of the parts, at least to get the neck angle correct - it takes just few strokes of file to move neck angle quite a bit and no machine fixture can ensure perfect fit of neck right off the machine. So if you machine to 90% and do the few final file strokes to get the dovetail seat and neck set at correct angles at the same time I think you've got the optimal process.
    Even the most CNC'd folks these days cannot do the standard tapered dovetail without hand fitting. Those who CNC dovetails do modified straight dovetails or just mortice& tenon that can be cut with ease but tapered is whole another level.
    Adrian

  22. #18
    harvester of clams Bill McCall's Avatar
    Join Date
    Feb 2014
    Location
    Forest Grove, Oregon
    Posts
    2,775

    Default Re: F4 design evolution/solutions

    My speculation has always trended toward the necks being cut on a single end tenoner.
    This very machine was familiar to woodworkers in the making of window sash, and the neck cut to match the curved body appears as a cope cut with a tenon, which was an extremely common practice then, and still is today. A tapered dovetail could be cut with 2 passes on the same setting, registered from the fingerboard surface, resulting in a very clean, precise cope with a uniform dovetail.

    To me this suggests the neck block would have been profiled after a plate was glued on to ensure the dovetail was consistently cut to perfectly receive the tenon(dovetail) and allow the cope to fit precisely.

    I don’t believe Gibson did anything new in manufacturing but the area was full of furniture factories with a lot of manufacturing prowess. Nice labor pool for them.

    Never taken an F5 apart or seen one disassembled. Just a best guess.
    Not all the clams are at the beach

    Arrow Manouche
    Arrow Jazzbo
    Arrow G
    Clark 2 point
    Gibson F5L
    Gibson A-4
    Ratliff CountryBoy A

  23. #19
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    The mandolin dovetail is not straight like on furniture but tapered. I already wrote use of special bit or cutter head to cut the dovetail into neck (could be custom made coping bit). You only need correctly shaped bit and fixture that slides the neck at correct angles (in two planes to create tapered dovetail) Because of the other angle to create taper, the curved surface of freshly cut neck cannot fit the curved part of body right off the machine and needs some fitting. No magic, just basic woodworking.
    Since back was glued first, the female part must be cut before that. You could not use shaped bit for this (that would trim the outside and insied at the same time) as that would not create tapered dovetail. Only straight. Tapered dovetail on curved surface throws some of the techniques out of window. Simple dovetail bit would do it nicely without touching the outside of the rim.
    Adrian

  24. #20

    Default Re: F4 design evolution/solutions

    I think they prioritized the fit of the curve over the dovetail itself. If you cut the curve of the neck heel on a bandsaw with a narrow blade, not by eyeballing it but like I said on a spinning jig, all that is left that doesn’t fit is a couple of mm at the narrowest part of the back of the dovetail. If they are both parts of perfect circles, as determined by the swing of the jig, they should fit perfectly.

  25. #21
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    The fit of originals is certainly much better than "off the bandsaw" which leaves quite scarred surface and you would also have to stop before you reach the inside corner. The width of the upper waist of dovetail is wider than back button so you would end up with no curved cut at all at the bottom of the neck and that part is very critical for correct fit and angle of neck.
    The Gibson dovetails had crisp cut right into the very corner of the neck dovetail and it fits the female part quite tightly which is not easy to do by hand - I usually chamfer the sharp edges of the body dovetail so the neck is easier to fit (you don't have to cut ito the very inside corners of neck part when fitting and can concenrate on bulk of surfaces holding the neck in place).
    StewMac used to sell mandolin dovetail cutting jig that included dovetail fixture and dovetail bit and custom made table saw molding head to cut the necks with another fixture holding neckat correct angle. That would get you few tenths of mm from final fit if you are careful. These days you can hardly find images of that setup, it probably didn't sell too well. StewMac still uses such setup for their mandolin kits. I believe they show the jig briefly in the mandolin building DVD.
    Adrian

  26. #22
    Mandolingerer Bazz Jass's Avatar
    Join Date
    Nov 2019
    Location
    New Zealand
    Posts
    347

    Default Re: F4 design evolution/solutions

    This thread would be so much better with pictures

  27. The Following 2 Users Say Thank You to Bazz Jass For This Useful Post:


  28. #23
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    Quote Originally Posted by Bazz Jass View Post
    This thread would be so much better with pictures
    I posted many pictures in older threads that would cover some of the talk above. I'll post some of those and later Imay find some other pics relevant here.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	1911A4.02.jpg 
Views:	88 
Size:	843.1 KB 
ID:	201555   Click image for larger version. 

Name:	Dovetail.jpg 
Views:	75 
Size:	43.7 KB 
ID:	201556   Click image for larger version. 

Name:	07.334.jpg 
Views:	78 
Size:	3.4 KB 
ID:	201557  

    Click image for larger version. 

Name:	1911A4.07.jpg 
Views:	76 
Size:	809.3 KB 
ID:	201558   Click image for larger version. 

Name:	08.jpg 
Views:	72 
Size:	219.3 KB 
ID:	201559   Click image for larger version. 

Name:	07.jpg 
Views:	78 
Size:	29.3 KB 
ID:	201560  

    Click image for larger version. 

Name:	05.jpg 
Views:	84 
Size:	128.5 KB 
ID:	201561   Click image for larger version. 

Name:	04.jpg 
Views:	73 
Size:	84.4 KB 
ID:	201562   Click image for larger version. 

Name:	03.jpg 
Views:	70 
Size:	31.4 KB 
ID:	201563  

    Click image for larger version. 

Name:	02.jpg 
Views:	77 
Size:	52.6 KB 
ID:	201564   Click image for larger version. 

Name:	01.jpg 
Views:	79 
Size:	78.4 KB 
ID:	201565   Click image for larger version. 

Name:	F-4.JPG 
Views:	68 
Size:	91.0 KB 
ID:	201566  

    Adrian

  29. The Following 4 Users Say Thank You to HoGo For This Useful Post:


  30. #24
    Adrian Minarovic
    Join Date
    Oct 2003
    Location
    Banska Bystrica, Slovakia, Europe
    Posts
    3,461

    Default Re: F4 design evolution/solutions

    - you can see traces of toothed plane that planed surface before attaching fingerboard. the lines seem to go all along the surface which suggests the overlay was applied after that. This would allow them to adjust final neck angle by planing away wood and establish nut position which was reference for the overlay.
    - the three point mandolin shows old hand cut dovetail with large asymmetry
    - the earlier black topped A shows dovetail cut with three wide saw kerfs (table saw?)
    - rest of the mandolins have cleanly cut dovetails that look like routed with dovetail bit.
    - the pic of inside shows that backs were glued first and glue cleaned while you can see dried squeeze-out glue on top. Bill Halsey took pics of all his vintage Gibsons and they are consistent in this.
    - they sometimes chamfered the edge of dovetail on body, could be to allow glue to escape - that would suggest gluing neck to body after back was glued on
    Adrian

  31. The following members say thank you to HoGo for this post:


  32. #25
    Mandolingerer Bazz Jass's Avatar
    Join Date
    Nov 2019
    Location
    New Zealand
    Posts
    347

    Default Re: F4 design evolution/solutions

    Quote Originally Posted by HoGo View Post
    I posted many pictures in older threads that would cover some of the talk above. I'll post some of those and later Imay find some other pics relevant here.
    Great photos. Thanks!

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •