I've spent the week replacing my old CAD computer that was gasping its last gasps. Got Rhino, MadCAM and all my other design software loaded on it, and it is just so fast! Can't believe what a difference 10 years can make!

Working on my 7 string Hawaiian (that's been underway for way too long), I modeled a pyramid bridge and hope to cut it later this week. I also modeled a 12 fret slot head neck and fingerboard for an upcoming project. I really like to do these things by hand, but needed to get my CAD chops back up. I haven't done much with it in 10 years. I've attached a picture of a trial run in pine. The saddle is straight and the top is flat since it's a Hawaiian Guitar. The real thing will be BRW.

The biggest progress I made was learning how to use "Grasshopper," a parametric plugin for Rhino. It provides some of Solid Works' features to Rhino--like being able to use a single model to build a fingerboard of any scale length, width, radius, etc. just by plugging in some numbers. Very cool indeed now that it's starting to make sense. I'm a total novice at it, but I'm impressed so far.

At any rate, it's been a sitting at the computer week in the shop. Sorry to bore with techno crap. I'm a retired IT guy though. Back to making sawdust soon. Need to cut binding channels on the seven string...

Dave

So if Dave or anyone else wants to chat about Rhino, Grasshopper, CNC etc. please feel free to message me

I

It came in because the guy who commissioned it was disappointed in the sound and wanted to know if it could be improved. It was ordered as a nylon string guitar but looking inside I saw it was extremely overbraced. Torres style bracing layout but the fan braces were 1/2” tall and 1/4” wide!

Nice carvings though.

Just Finished

I ended up declining to work on this guitar but thought I’d post some pics because it had some of the wildest ornamentation I’ve seen.






It came in because the guy who commissioned it was disappointed in the sound and wanted to know if it could be improved. It was ordered as a nylon string guitar but looking inside I saw it was extremely overbraced. Torres style bracing layout but the fan braces were 1/2” tall and 1/4” wide!

Nice carvings though.

Your pics aren't showing. If possible, scale them to reasonable size and upload directly to the forum rather than linking to google. That way they will remain viewable for as long as the forum exists, rather than disappearing into the void whenever google changes their URL format.

I'd love to see what your 5-axis setup looks like. I'm not sure if I could use such a machine if I had it Too poor for CAM software, so I write C++ programs to generate gcode in a slightly more user-friendly way than writing gcode directly. I do want to add a 4th axis, but tying to visualize 5 axis tool paths might be too much

Just Finished

Thanks Dennis, I've shrunk the pictures and they seem to show correctly now.

The 5-axis setup is on my tiny mill. The rotary axes use steppers and 100:1 strain-wave or harmonic-drive gearing with impressively low backlash.

I'll also be generating the G-code myself using the Python component inside of Grasshopper. Fortunately I don't need any fancy 5-axis simultaneous motion, just simple 3-axis cuts but from 5 different sides of the workpiece. Figuring out how to calculate those positions will be an interesting challenge.

Try using a free photo hosting site like IMGUR....

Still no go for me... just small boxes of failure instead of huge ones:

I love strain wave gears. I want to try making itty bitty ones for guitar tuners, but I'm not sure if it's actually possible. With a 0.5mm bit I can mill 24 teeth into a 3/8" steel rod for the flexspline/tuner post, but due to their relatively large size it will need to flex quite a lot between full engagement and moving past ring teeth. Also due to their relatively large size, all that flexing will be concentrated into the 24 short spans of metal between teeth, and probably fatigue it.

I could reduce the height of the teeth so it doesn't have to flex as much, but then the tolerances become even tighter to prevent skipping. Ideal would be to reduce the width of the flexspline teeth to leave wider flexible spans between them, but then the ring teeth need smaller spaces between them and I can't mill that. I could also try going down to 0.3mm end mill, but then tooth face width is limited to 1.5mm (even less after deburring) and that's pretty tiny for guitar string tension even with multiple teeth engaged.

Yeah, that would be much easier just using it to reduce workholding changes. Although figuring out how to hold onto the raw stock in such a way that you can get at 5 sides doesn't sound so easy either... and then how to machine off the portion that was being held.

In my 3-axis programming, I have a variable to rotate the entire coordinate system around the Z axis. Ultimately, all movement is done via functions like G01_XY(x, y, feedrate), which will rotate those coordinates, apply backlash compensation, and then write the G01 command to the output file. So for example I can write a tool path to cut one gear tooth, and then use a loop like for(int tooth = 0; tooth < toothCount; tooth++){ setRotation(tooth*360.0/toothCount); cutTooth(); }

You could probably do something similar for your 5 axis system, where your programs are written in XYZ space, and all the transformations to 5 axis are done "under the hood".

One thought regarding your design: could you put the mechanism in the tuning knob instead of the tuner shaft? Like a Steinberger gearless tuner? Might give you a significantly larger diameter to work within than the shaft.

My way too patient friend asked if he could pick this up in two weeks after waiting three years. I’m finally on the last haul. The body is ready to buff, but the neck needs a little finish touch up before buffing and fretting.


Sent from my iPhone using Tapatalk

Love the crisp, understated aesthetic. Beautiful.

Beautiful James, looks like a great guitar soon to be.