Jump to content

The Grasshopper Generative Guitar


GenerativeGuitars

Recommended Posts

I don’t use CAD at work that much these days. So I can pick whatever suits my personal needs. That is Fusion360 for technical stuff and Modo for everything else. SubDs are great for quick sketching. I use Illustrator for 2D drawing and in many cases I rather draw shapes there and export to 3D software. Old habits die hard.

  • Like 1
Link to comment
Share on other sites

10 minutes ago, henrim said:

I don’t use CAD at work that much these days. So I can pick whatever suits my personal needs. That is Fusion360 for technical stuff and Modo for everything else. SubDs are great for quick sketching. I use Illustrator for 2D drawing and in many cases I rather draw shapes there and export to 3D software. Old habits die hard.

I still use ASPIRE for my 2d CAD. It works fast and accurate. I guess because I started using it at version 3 now at 11. It makes since to me. LOL!!!!

MK

  • Like 1
Link to comment
Share on other sites

  • 3 weeks later...

So a key concern designing/modeling this part was

- a 'random|any' guitar part i got from Amazon, thomann, etc is going to be different!
- doesn't match the part on the ikea caster
- doesn't match the part on the client's bridge...

BUT now I know how to position the screws relative to the body/bridge. Next is making the scale/distance to the nut right.
- I could model a AI system to tune guitars right from the start.  Dont hold your breath, idea in the plan since at least 9 months...


This is why generative modeling is a key to ease the placement of the parts on the guitar according to theory (the key neck to bridge distance which is called the scale which can be divided into x frets for any tonality of each string (damn this is complex). Again I learned that where was an offset from A (xyz position) to Z. But Z is depending on P which is depending on D which is depending on A (mid top of body). Who cares....

This is crazy, every offset seems different for just about every part of a guitar! I care!!! Im not building an Ikea guitar!!!!

 

 

image.png

Link to comment
Share on other sites

So one big thing i learned modeling this part was...

The distance to nut for minimum distance offset (from the nut) is on the thinnest string - the longest screw on the bridge - damn this is complicated!

I got 3 different bridges and all 3 are different!!! One dimension i did not come across was the distance from the pickup to the bridge... Is there any theory on this?

Link to comment
Share on other sites

33 minutes ago, GenerativeGuitars said:

One dimension i did not come across was the distance from the pickup to the bridge... Is there any theory on this?

There is a theory but does it matter is another question. Take a look at the nodes picture I posted on the previous page. You can play harmonics at about 1½ and 1.75 fret, the tones being a fifth apart. And for some reason that's where the bridge pickup usually sits. Or close enough given how close to each other those nodes are. Also, you can play the 5th fret harmonic on the 22th fret or over the neck pickup as well.

  • Like 1
Link to comment
Share on other sites

1 hour ago, GenerativeGuitars said:

The distance to nut for minimum distance offset (from the nut) is on the thinnest string - the longest screw on the bridge - damn this is complicated!

Yes, and as far as I understand that’s the only distance you need to care about. String length is never shorter. You place the bridge based on that minimum distance. Which is the scale length. Other strings are then corrected to proper length when setting up intonation. Of course some bridges,  like the LP bridge, is set on angle and there you have to set the maximum length as well.

  • Like 1
Link to comment
Share on other sites

  • 1 month later...

Happy 2024 everyone.

I took a while to reply because I have been cleaning up my Grasshopper model, and started the tolerances testing - which meant I had to clean and reoil the cnc!

The Grasshopper model got really big - version 1.5.8 now! Yes, that's 158 save as... The green part on top is the newest - the bridge...
This morning i revamped and improved the lower mid-left orange blocks which are the neck to body interfaces. And I got to test that this morning!

image.thumb.png.0da44bf27dac4a4332c38883843fbd33.png

Link to comment
Share on other sites

During this morning's 2 cups coffee ritual, I redid the Neck-Body interface.
It was too simple (although it's already a second evolution of the 'basic part' based on the original plan...)

The original plan is just a top down curve. I saw some guitars have a neck taper - seems logical - and there is the (possible) taper across the length of the neck...

image.png.67b1052e66d60913df3391707c35be71.png

This is my biggest worry still but the top to bottom taper of the walls of the neck was the most immediate worry. Maybe there is no tapper but i included it also into the model.

image.thumb.png.88c64748481cc2889a7047dc02df8d32.png

So now you can dial top and bottom taper and with any precision. The top to bottom of the neck (whole part as opposed to just the body-neck interface) will react automatically if you change the neck in a future version. This model is body centric vs Nut centric.

If any Luthier can confirm this? Do you build the guitar starting around the nut or the bridge? Or what would determine which? Guitar weight distribution? My first guess...

Anyway, I fixed this taper - x-y-z taper on tap!
I did the CAM jobs after 'baking' a small test model to see how it fits...

image.png.0b845e0d45913a2def429ca7a7954983.png

and 30 minutes later, I had the full result - fits PERFECTLY!!!
image.thumb.png.20834540780a811ed3a907588d977f4b.png

image.png.036d628107b50ea29ca284cffe6946e9.png

Edited by GenerativeGuitars
  • Like 2
Link to comment
Share on other sites

  • 3 weeks later...

I have been impressed with this approach using Grasshopper. I too was working on a solution at one time. My health has made it so I cannot do wood working anymore so I stick to metal now. If you would like I could look at your grasshopper code and layout and make suggestions if you are willing to share with someone. I have no desire to sell this or use for my purposes. If you are interested let me know and I will PM you.

Mike

Link to comment
Share on other sites

3 hours ago, MiKro said:

I have been impressed with this approach using Grasshopper. I too was working on a solution at one time. My health has made it so I cannot do wood working anymore so I stick to metal now. If you would like I could look at your grasshopper code and layout and make suggestions if you are willing to share with someone. I have no desire to sell this or use for my purposes. If you are interested let me know and I will PM you.

Mike

Hello Mike,

That's very kind of you. Thanks.

Alas Im not looking forward to do support for the code yet as I enter the CNC testing phase.

I dont have issues "coding" in Grasshopper although I have to admit that making the neck joints (between body interface and headstock) was a real pain - luckily a friend gave me some valuable tips and now I have 3 versions with varying results. Im re-writting the code as I go back to adjust tolerances for the CNC'ed body right now.

Actually, where I could use help is whether making cavities in the body could help and the trussrod design...

 

 

  • Like 1
Link to comment
Share on other sites

16 hours ago, GenerativeGuitars said:

3 days spend backing up the PC - Either my RAM or my M2 C boot drive was corrupted. Anyway, long story short, it wasn't hard to reinstall the OS, most Rhino3D plugins or RhinoCAM softwares (Rhino7 and Rhino8) - a 2 day ordeal while still tracking the crashing bsod component. This between changing many little things in the GH model/program/workflow. So now, back to productivity! 64GBs instead of 128... Still not feeling the impact speed wise.

To improve the final dimensions in the model, I isolated the bridge/neck interface cavities to make testing cheaper and faster.

CNC time is about 15 minutes with manual changes for drills...

 

image.thumb.png.1a0e4184143495d1337fe295047ccc8a.png

image.thumb.jpeg.cf87540a9924795c27e9d1b51f8baad8.jpeg

 

 

 

 

 

 

Link to comment
Share on other sites

23 hours ago, GenerativeGuitars said:

Hello Mike,

That's very kind of you. Thanks.

Alas Im not looking forward to do support for the code yet as I enter the CNC testing phase.

I dont have issues "coding" in Grasshopper although I have to admit that making the neck joints (between body interface and headstock) was a real pain - luckily a friend gave me some valuable tips and now I have 3 versions with varying results. Im re-writting the code as I go back to adjust tolerances for the CNC'ed body right now.

Actually, where I could use help is whether making cavities in the body could help and the trussrod design...

 

 

I am not sure I would worry that much about the pickups and truss rod. These are mainly done using a 2d strategy and subjective to the individual. Neck pocket, placement and bridge placement are the real show stoppers. The rest is just glitz.  If you were to add them, use standard CAD drawings and just move them around using GH. LOL!!! MK

 

  • Like 1
Link to comment
Share on other sites

On 2/18/2024 at 9:07 PM, MiKro said:

I am not sure I would worry that much about the pickups and truss rod. These are mainly done using a 2d strategy and subjective to the individual. Neck pocket, placement and bridge placement are the real show stoppers. The rest is just glitz.  If you were to add them, use standard CAD drawings and just move them around using GH. LOL!!! MK

 

The whole guitar is made based on 2 CAD curves from an old plan (I can add, change these at will). I used the plan's shapes to make the body and head-stock shape and set all the dimensions.

Doing this as a total guitar idiot, I learned a lot about build techniques, guitar types, scales, scarf joint shaping with GH. And between the time I started encoding all the plan's measures - and today (about a year after) - I'm adapting the model using planes for references, automating the CAM job as much as possible. Created a nice system to test cavities to avoid wasting valuable wood.

This model is based on that 1950's plan's dimensions and some are completely off like the control box is wrong size, screw holes dont work with this Fender pro neck, different bridge. Thanks to GH, I just change a number in a dial and fixed. Sometimes, I have to rethink how a part is supposed to fit offset to this or that point (nut, bridge, neck pocket etc. Adding one or two more pickups is easy. Will it sound right, I dont know yet...

Im 2 features away to be cnc tested for tolerance - will the HW fit? Bridge cavities and positions already had to be edited which was just 10 minutes to prep for next cnc test. Now it's 2 minutes to test...

I still dont have a pretention to become a luthier or a CNC god - but I love the details of the craft! I just love design :)

My guitar playing ability = Im still learning the James Bond theme. Im getting the sound at least!  :)

  • Like 1
Link to comment
Share on other sites

1 hour ago, GenerativeGuitars said:

The whole guitar is made based on 2 CAD curves from an old plan (I can add, change these at will). I used the plan's shapes to make the body and head-stock shape and set all the dimensions.

Doing this as a total guitar idiot, I learned a lot about build techniques, guitar types, scales, scarf joint shaping with GH. And between the time I started encoding all the plan's measures - and today (about a year after) - I'm adapting the model using planes for references, automating the CAM job as much as possible. Created a nice system to test cavities to avoid wasting valuable wood.

This model is based on that 1950's plan's dimensions and some are completely off like the control box is wrong size, screw holes dont work with this Fender pro neck, different bridge. Thanks to GH, I just change a number in a dial and fixed. Sometimes, I have to rethink how a part is supposed to fit offset to this or that point (nut, bridge, neck pocket etc. Adding one or two more pickups is easy. Will it sound right, I dont know yet...

Im 2 features away to be cnc tested for tolerance - will the HW fit? Bridge cavities and positions already had to be edited which was just 10 minutes to prep for next cnc test. Now it's 2 minutes to test...

I still dont have a pretention to become a luthier or a CNC god - but I love the details of the craft! I just love design :)

My guitar playing ability = Im still learning the James Bond theme. Im getting the sound at least!  :)

Then I would add the basics, Single coil, Tele style single coil and Humbucker. Make the parameters changeable and location changeable, they all would be good. Also the ability to use one or more of any type or combo of those on the guitar.  If also possible make it so that they could mount from the rear as well. MK

Examples of Tele bodies that would require or might go against norms.

The black one may be used with a TOM bridge or String thru Strat bridge. It would also require a set neck. The other allows for any approach.

 

orgtele1.jpg

S-Tel_2.jpg

Link to comment
Share on other sites

5 minutes ago, MiKro said:

Then I would add the basics, Single coil, Tele style single coil and Humbucker. Make the parameters changeable and location changeable, they all would be good. Also the ability to use one or more of any type or combo of those on the guitar. MK

Please excuse my enthusiasm but all those options are already implemented! I want to implement 12, 8, 9?, 6, 4 strings, all fret boards and insets possible, do the whole 9 yards!

But making the right sounding guitar first is the priority. Getting this far working weekends and nights since last 12 months and finally with just 2-3 more cnc carves to figure out where the right screw positions are and make sure the controls of the guitar fit in the cavity has me stoked to start carving the whole thing out!

Tomorrow im getting some cheap downcut end-mills (6 mm shanks) to start and hopefully improve chipping on the top surface of the wood. Hopefully I can source some Amana end millls. Europe is not well furnished for end mill varieties I have to say.

 

 

 

  • Haha 1
Link to comment
Share on other sites

6 flute is a little much, 2 flute and three flute will serve you better in wood. MK

Ooops I misread that as 6 flute when you said 6mm LOL!!! Im Bad and my eyes are old without my glasses. MK

PS. end mills are better than router bits most times. :) I have used the ebay cheap ones as well with no issues. Just make sure you are trammed and the steps are correct. Check your TIR on the spindle when using very small end mills mk

 

 

Link to comment
Share on other sites

So coils - my friend changed his mind about humbuckers and switched to something texas single coils - he knows his guitars :)

He's providing the bridge and neck - So i made my model adapatable to screw positions, neck size... And the last 3 weeks were spent doing test for those (including distance from nut to bridge. I had to re-measure everything (measure twice and cut once) and gladly except for the screws, it's all aligned! The control box is too thin, some depth tests to come later before i carve the big block and flip it to carve the other side (no cavities - telecaster) but totally possible for Lespaul styles etc.

I didn't go light on this model - for me it's an industrial design attempt(in Grasshopper strictly) with strict measures which I barely did before - I did some electronic panels and boxes but not this precise. 1=2 weels before i carve it! yeah!!!!

By then i get my first "to be measured against" guitar...

image.thumb.jpeg.adf73b4d97281b4e96a72f1b277f10fe.jpeg

Link to comment
Share on other sites

7 minutes ago, MiKro said:

6 flute is a little much, 2 flute and three flute will serve you better in wood. MK

Ooops I misread that as 6 flute when you said 6mm LOL!!! Im Bad and my eyes are old without my glasses. MK

PS. end mills are better than router bits most times. :) I have used the ebay cheap ones as well with no issues. Just make sure you are trammed and the steps are correct. Check your TIR on the spindle when using very small end mills mk

 

 

Sorry, Im in mm land... I started doing woodwork in 7/16's measures so I totally understand!
I use mostly 2 flutes - 1 flute if plexiglass (lessons learned)...
For feeds and RPM speeds, I listen to the sound of it believe it or not! Get's tricky in wood knots (for the pun!)
 

Link to comment
Share on other sites

29 minutes ago, GenerativeGuitars said:

Sorry, Im in mm land... I started doing woodwork in 7/16's measures so I totally understand!
I use mostly 2 flutes - 1 flute if plexiglass (lessons learned)...
For feeds and RPM speeds, I listen to the sound of it believe it or not! Get's tricky in wood knots (for the pun!)
 

Yep I get it and yes sound is a GREAT indication for wood routing. I wish that was true on metals all of the time. I also understand the one or what we call a zero flute end mill here in the states. They also work for Aluminum. :) MK

Link to comment
Share on other sites

19 hours ago, MiKro said:

Yep I get it and yes sound is a GREAT indication for wood routing. I wish that was true on metals all of the time. I also understand the one or what we call a zero flute end mill here in the states. They also work for Aluminum. :) MK

we also call them zero flutes or one flutes. Best result in ALU or plexi.

I got my compression end-mills via the mail today but still resolving this pesky BSOD issue on my workstation.

  • Like 1
Link to comment
Share on other sites

Downcut end mills are awesome, no more edge chips!!! Highly recommend. No issues on the bottom side either!

So came finally the time to carve the big block and hmmm, I forgot something…

Screw positions and centering/fixing the guitar block setup on the cnc for facing...


How to set up the block right in reference to the table. 1 problem is that no 2 blocks are the same size (xy)…

The first problem comes when you want to “face” the block. I can’t fix the block just any how… I wont screw it in the wasteboard because center is off, glue it = unprecise...

I would like to reference the screws on the block to the waste board/cnc table xyz = center of the board on the neck end.

I can’t screw it because the cnc will do that with reference to both sides of the block to flip it dead on center.

SO I have to center the piece to the table center. I know of different jigs used for dovetails but these just wont work with this kind of surface (although there might be a way)…

So I devised this kind of contraption to hold on the board and center it. Once set, I can devise the center - but not to the table. And the holes to hold on the jig are never the same given any board size! Connandrum!

Any suggestions?

image.thumb.jpeg.81cc5ef8ce58fe5cc6e3691b5aeb6163.jpegI thought of this jig

 

Link to comment
Share on other sites

  • 4 weeks later...

So a month has passed since last update, Apologies as things evolved a lot...

Got more wood and a pickguard blank...
Remodeled a lot of the scripts to improve speed and overall easy the setups. Finally figured out how to center, flip the board also.


image.thumb.jpeg.381dba5147778b2acaa52c7a20caa054.jpeg

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...