Official Luthiers Forum!

If made correctly, you can use a thinnish straight sanding block to level. A compound radius FB is a ruled surface.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Is that a single radius setup?
The problem I ran into was with a compound radius.

I've seen a sanding pad that was made from UHMW that was kerf cut that some seem to use. Also a straight edge that you need to follow the nut edge and the tail edge of the board when sanding.

Trevor, could you elaborate on your jig? It seems that this will only do a radiused fingerboard and not a compound radius.

Sounds something like Kim's http://www.luthiersforum.com/forum/view ... 02&t=13639
By the way I ordered the plans from Cool Tools to see what it's about too.

It machines a compound radius, a conical section in this case. If you look carefully you will see that the lower rails diverge as they get closer to the camera. And as the rails diverge, the height of the router off the work piece varies, so the rails sit on ramps to compensate (high at the nut end, but that is really hard to see because of the wood chips).

If you try making one of these, don't underestimate the mathematical analysis and the precision it is necessary to build the tooling to. If you use a conical section with the variation in fretboard width governing the radius change (which it has to, if you're using a cone, because of simple proportionality) typical numbers work out at 10" at the nut, 12" at the 12th and 14" at the saddle for the string spreads that I use.

Above about 14" radius at the nut, the differences between cylindrical and conical are so small as to be inside the build tolerance (for most people) and less than the relief most players want in the neck. Remember, all this is predicated on a straight neck...

There was a paper in AL #101 with an excellent discussion of non-conical fretboard geometries (ruled surfaces) and the limits of usefulness of compound radius boards and there's more stuff in the usual place, though not how to build the jig pictured (although the book does have a similar pic of the jig).

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

It works just how I wrote in my jig idea a few posts before. The long compass rides on the lower two steel rods. Because they converge at the nut, the long compass will draw a radius with a tighter arc at the nut than it will at the end where the guide rails are farther apart.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

I'm not sure they are converging or if that's just the perspective of the picture or if there is something I'm not seeing. I'm interested in seeing what Trevor has to say about it. But I don't think changing the span will change the radius. As the span gets wider the sagita also gets longer in relation ship to the span. You would have to change the angle of the long compass that is related to the sagita and the span. At least that is how I see it and from my attempted prototype and R&D have discovered. If you go to the tables http://liutaiomottola.com/formulae/sag.htm and punch in the #s keeping the sagita the same but changing the 1/2 span it appears at first that changing the span will change the radius but that change in span is changing the angle produced by the sagita to span ratio or rise over run or sine/cosine .
Like I stated before, the logic of how this works is different. But I could be wrong and missing something. But I also thought I had it sort of figured out until I tested my ideas. That's what the R&D was about, testing my ideas which made me see and learn the long compass didn't work like I thought it would. Now if you can figure out how to keep the sagita constant and while changing the span you'd have it. To me it just became to get too complicated when there are other methods that work that are simpler to make.
Don't get me wrong here I'm not trying to discourage anyone and I may be wrong. I'm just relating what I've found out so far as of last weekend. I've thought about this for well over a year and put my ideas to the test and found my logic was wrong. I was also hoping I could work out the bugs but I learned some things and decided to move on since I only have one or two FBs to do.

Trevor and I posted at about the same time, read his post just above mine. It works exactly as I described it.



Changing the span indeed changes the radius. I actually checked at mottola site before I posted. I think you're just approaching it the wrong way. Instead of using the first calculator, use the second that calculates the radius.

e.g. if we chose a saggita of .125 and a span of 3", we get a radius of 36".

Now we change the span to 2" but leave the saggita at .125 and we get a radius of 16"



The thing is that in my first post, I suggested a long compass at each end of the jig - this won't work. What I proposed in my second post and verified by Trevor is to make the contact points of the sled itself a long compass. The final trick is to adjust the "height" of the FB at each end of the converging bearing rods so that the FB doesn't get thinner at the nut.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Yes, thats what I said, when you change the span and keep the sagita the same it changes but when you change the span you are changing the angle it makes. thus a sagita or rise of one over an 1/2 span of 2 produces a different angle as a rise of 1 over 1/2 span of 4 which produces a different radius. I made the same wrong conclusion that you are making at first too.

I see what you mean now. Could be that there's a missing piece of info from Trevor's jig.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Hmmm, I did miss that 2nd. post of Trevor's. Could be the ramping of the converging rails that is the missing piece to the equation. I never thought about that. That might/ must be the key. I wonder if ramping the board would work instead of ramping the rails. Seems like it should. I also wonder if it's cutting more of a section of an ellipse then a true radius since it is ramped? Perhaps that's what is happening . Either way seems like it's working in the end.
Thanks Trevor, I'll look up that artical too.

It's not the ramping. After your last post I went into the lab to run a few experiments to prove you wrong and I couldn't because you're right. The distance between the bearing points of the long compass have no effect on the radius.

Trevor will have to chime in with what we're missing.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Some better pictures of what's behind the router in the pic might help. Found that artical but don't see any pic of a jig in that issue. I'm not sure why the top rails go the full length either unless it's to help keep the router square in the jig or for alinement.

I don't know if these will be helpful to anyone, but this is the layout of the router version. Missing is a spring that would pull the pin upward. There would also need to be a handle to move the cradle. The cradle would hang down and swing, I did not want to take apart my jig to show this.

_________________
http://shastaguitar.com/
http://www.kalimbakit.com/
http://www.youtube.com/user/comfyfootgr ... ature=mhee
http://www.facebook.com/robert.renick.7

I built this rig about 10 years ago and can’t find any notes/calcs from that time explaining how I came up with it. The thing I remember about it (as I mentioned before) is that the geometry was hugely complicated, so much so that looking at the jig now and trying to reverse engineer it leaves me baffled. I must have had a few more synapses firing in those days!

Behind the router is – wait for it – another shipwright’s compass, which you can see in this pic.


Now, this is what I think is going on ( ...and I designed this…). The shipwright’s compass at the end (above) sets the guide rail alignment, so that, except on the centreline, the router travels obliquely to the lower rails on which the router shipwright’s compass/carriage runs. As the router runs down to the far end of the carriage rails, because the carriage rails diverge, the router will effectively lift more as it gets more off the centreline and further down the carriage rails. This means a larger radius is machined at the far end compared to the nut end. I think!

The proof of the pudding is, of course, in the eating. So does this thing work? I was beginning to have my own doubts!

Here’s a rectangular piece of scrap I just radiused, together with my trusty radius gauge.

If you look closely you can see that the longitudinal machining marks are slightly oblique to the axis of the board. They run off the board on the left hand edge, as you’d expect.

Here’s a pic of one end of the board with the 400mm radius gauge on it:

and here it is at the other end:

I didn’t get this board quite centred. The point of these pics is that you can see the difference in the radius of the board at each end compared to the 400mm gauge. So what is the radius at each end?

Well, to me it looks like 350mm and 300mm,


…which happens to be exactly what the jig was designed to cut over this length (half scale length).

So, it definitely works and I think that’s the right explanation. But I have no idea how I figured out the geometry; I just remember it being a real brain sprainer.

The other thing I mentioned in an earlier post was the build precision required, because you can see from the radius gauge pics that you don’t have much tolerance to play with. In fact, on the very first pic of the jig you’ll see a rule on the bench. On that day, that wasn’t a rule, it was a shim keeping the centre of the jig flat (and I built the base from 35mm ply).

I haven’t used this jig for a few years now. Whilst I personally prefer fretboards with quite a tight radius my customers seem to prefer 16” and above. At that radius the difference between cylindrical and conical based on the nut and saddle string spreads is negligible.

If I had to build this tomorrow, I wouldn’t. In the time it would take to figure out the math again and build to that precision it would be easier to knock up a 3 axis CNC router, like Andy has.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

If I had to build the Fox version again I would make it long enough for a bass fb. IMO Charles Fox remains the king of inexpensive effective jigging. It took about 2 hours to make the jig, parts less then $20, with the adjusters I was dialed in quickly. I recently made new wheels to get a different size, that was less then an hour to make, install and adjust. Thanks Charles!
Rob

_________________
http://shastaguitar.com/
http://www.kalimbakit.com/
http://www.youtube.com/user/comfyfootgr ... ature=mhee
http://www.facebook.com/robert.renick.7

I'm showing what I have so far and I'm not sure if it will work until my 80 grit belt arrives, 120 grit just ain't going to get it. There may also be a problem with a roller causing a divot in the 1st fret. Since the bed isn't long enough I have to move the FB foward and back. I may have to sister a longer bed plate to the existing bed. But this might give some others some idea how to convert their 6 X 24 belt sander. I based this on the Somogyi and Fox ideas I guess.

Well it did work with 100 grit belt except for one problem . Not with the jig but with my sander. The drive belt roller sits a little higher then the bed so that it will radius a low spot in the 1st. fret. It could be corrected by sistering a plate to the bed. Improving the leveling of the end beds would also be nice for setup. Possible something along the lines of the Fox setup turnrd 90 degrees.

Started on the Cool Tool design. But seems like there was a problem with it on compound radius boards. I'm hoping if I make the bearings closer at the nut it will work. If not I'll use the carriage for the other jig.

I have been looking forward to your results. I had the same issue, turns out there was an adjustment for that on my machine, never noticed it before. I am also expecting some new belts, I am thinking that heavier paper will help my machine a bit. I suspect that the thinner too loose belt on there now wraps a bit when I push on the belt. With the set up you made, if you had a long enough router table, you could probably get started on that and then move to a finer grit on the belt.

Very cool adaptation Chris, simple and effective,
Rob

_________________
http://shastaguitar.com/
http://www.kalimbakit.com/
http://www.youtube.com/user/comfyfootgr ... ature=mhee
http://www.facebook.com/robert.renick.7

Never thought about that. I actually could throw another top on my router table and probably do that.

From my experience with the Cool Tool design, the key to making it work with compound radius boards is to let each end of the fingerboard move laterally independently of the other end--that is, let the spring hold downs allow the fingerboard's motion under the router to follow a more conical rather than cylindrical path. The spring hold downs lets you move the nut end in smaller increments than the heel end, so the compound radius works comfortably--I didn't have a problem with the bearing spacing. Trying to move the whole carriage along a cylindrical path (with different radii at each end) didn't work.

Dave

I don't know if it will make any differance or not but it seems to me that separating the bearings might help out. I guess I'll find out if my theory is correct or not this week or next weekend.
What I did was find out what the sagita was for a 2.25" bearing span for a 13 1/2" radius (+ 2 1/2" for finish radius)and then what the span would be for a 7 1/2" radius(1.68") with the same sagita(something like .047").
Like I said I'm not sure if my logic is correct or not.

My logic was incorrect and I even put one bearing on the front and I don't think it will be easy to get consistant results for a compound radius. So I think I'm going to move on to Mario's and Dave's style or revisit my sander or both or start checking Craigs List for a cheap used edge sander.