Don Noon's bench

[Chris Llana]

Really nice!

[Don Noon]

Back to the shop...

The last of this batch of 3 got its linings installed, encased in counterforms and a forest of modified clothes pins.  I needed to make extra modifications to the clothes pins to clamp around the counterforms as well as the linings and ribs.  This should hold the outline more precisely so I can finish the plate outlines without individually matching them to the garlands.

And, anticipating that this process will work, I made up a baby sanding arbor to finish the plate outline (or close to it), pictured here with its big brother that I made earlier.

[HoGo]

Don, you should upgrade your clamping method, somehow those clothespins are not up to level with CNC and those moulds :-) You need to invent something fancier.

[Don Noon]

3 hours ago, HoGo said:

You need to invent something fancier.

Actually, I don't.  They work perfectly, and are cheap.  I only go fancy when it does something better.

[Don Noon]

Still a lot of time spend in tooling and fixtures for the new models, but a major milestone was cutting a purfling groove and outline of a back plate.  It was a pre-roughed maple blank, so the hold-down method was just for this one.  I'm pleased with how it turned out, so far.

Not obvious  is a huge effort to fix a tiny problem:  the CNC machine X and Y axes were out of perpendicular by ~0.3 degrees.  With the garland made with that skew, and the back cut out and flipped, that would double the skew and the corners would be out of alignment by almost 1 mm... too much for me.  After gnashing of teeth and racking of brain, I finally managed to reverse skew the model geometry, and that is what you see here.  After this batch, I'll precisely align the machine and not have to go thru this again.

And in case you were wondering... it is a torrefied maple back... which required me to cut apart the pre-carved blank, and then glue it back together afterwards.  

 

[jezzupe]

1 hour ago, Don Noon said:

Still a lot of time spend in tooling and fixtures for the new models, but a major milestone was cutting a purfling groove and outline of a back plate.  It was a pre-roughed maple blank, so the hold-down method was just for this one.  I'm pleased with how it turned out, so far.

Not obvious  is a huge effort to fix a tiny problem:  the CNC machine X and Y axes were out of perpendicular by ~0.3 degrees.  With the garland made with that skew, and the back cut out and flipped, that would double the skew and the corners would be out of alignment by almost 1 mm... too much for me.  After gnashing of teeth and racking of brain, I finally managed to reverse skew the model geometry, and that is what you see here.  After this batch, I'll precisely align the machine and not have to go thru this again.

And in case you were wondering... it is a torrefied maple back... which required me to cut apart the pre-carved blank, and then glue it back together afterwards.  

 

Congrats, this will be interesting to see what  your output will be once you get this all figured out. I really have a very good feeling about you and your career in this in general and am looking forward to seeing if you can marry this higher output rough work with that last 10% of the hand work thing. It does look very precise and spot on as far as the shape from here. Don didn't hire an apprentice,  he built a robot assistant. I certainly hope when this is all said and done that the robot gets a proper name.

[scordatura]

When you initially did your machine setup, did you measure the accuracy of the  of the setup (perpendicular) or was it something in the software? I would assume it was in the setup. I'm glad you brought this up!

[Don Noon]

3 hours ago, scordatura said:

When you initially did your machine setup, did you measure the accuracy of the  of the setup (perpendicular) or was it something in the software? I would assume it was in the setup. I'm glad you brought this up!

It's all in the machine; I did attempt to align it at first, but didn't do a good enough job of it.  Now I know more, and will do it better after I get done with these instruments.

I see that now there are machines with linear bearings instead of the round rails.   Far more rigid. If I can justify the upgrade, I would really prefer that.

[Michael_Molnar]

It's a pain to discover this after cutting a lot. 

Bummer. 

[Don Noon]

The body of my small (352mm) new design is together, although final scraping and edgework is left for immediately before varnishing.  Acoustic tests at this point seem to be good, although there's still a long way to the finished state.  At least there's nothing glaringly wrong that needs fixing.  Gotta get working on the neck now.

And, to avoid working on the neck, I've been messing with varnish, testing out new stuff on an older violin of mine, so I will have a better idea of what I'm doing when I get to varnishing the new one.

One thing that has been bugging me is the gazillions of bottles of various varnishes I've cooked up over the years, with none of them quite right.  And, since I hate to throw out varnish (both for the pain of dumping perectly good resin and oil, and the thought of this stuff oozing around in a landfill), I put ~20 bottles of varnish into a large deep fryer, adjusted resin/oil ratios, cooked out the old solvents, and thinned it out with what I like now.  The result:  over a gallon and a half of varnish, enough to last ~100 years or so.  I got one of those overhead stirrers... invaluable for this task.  Those are quart bottles in the photo.

 

[MikeC]

good idea with the varnish.  What does it look like applied to wood? 

[Don Noon]

11 hours ago, MikeC said:

What does it look like applied to wood? 

It looks mostly like the wood.  Although it's dark brown/amber/red in the jar, it isn't that strong of a color to do much when applied thinly.  Then it's at best a pale amber color.  The intent was just to get a good base varnish that dries and handles well, as a (mostly) clear varnish and medium for pigments.

Here it is applied to 2 different shades of bigleaf maple... one a silvery/gray, the other a warmer orange.  Thickened with fumed silica so it doesn't soak in and burn the flames.

[MikeC]

that's the problem with varnish it's hard to get enough color in it.  but at least you've got plenty of clear coat. 

[Guy Booth]

On 5/30/2019 at 5:37 PM, Don Noon said:

That's not the way I'd want to go.  If I can't model it geometrically the way I want, I'd rather just use the CNC to hog out most of the wood (with crude gometry), and finalize the arch and graduations the old way, mostly by eye

Yeah, there was a thread about this group a while ago.  Their arching is terrible, driven more by their method than the actual violin they are "copying".  Although they DO use the same modelling software I have, I don't want to use their method.

Awesome stuff Don!  You’re at the point I’m aiming for - CNC followed by hand finishing. It’s going to take me years to get there so it’s inspiring to see your work in the meantime! I really like your approach.

I’m also not keen on the idea of probing to generate a CAD model, there is no control in design doing it that way. For the same reason I’m also not keen how James Cherry produced violin arcings with Fusion360 in his video. There isn’t enough precision with Sculpt in Fusion as he showed, although I believe there is a ’snap-to’ function in Sculpt that may get the precision - I’m yet to try that. I chatted to James about it, and asking why not build a scaffold with t-splines then use Loft and Patch to complete the surface, but I didn’t understand why he thought that not possible. Maybe it can’t be done that way, but that’s the route I’m going.
Any thoughts on that? 
Anyhow, looking forward to seeing more of your stuff down the line. Cheers. 

 

[Don Noon]

11 hours ago, Guy Booth said:

...why not build a scaffold with t-splines then use Loft and Patch to complete the surface...
Any thoughts on that? 

I think my thoughts at this point are pointless...  until I have actually gone through the process of making a 3D arch or two.  I'm not there yet.

[Don Noon]

A couple of distractions (among many) hindering progress on the actual instruments...

1)  A rotisserie.  Varnishing on my next violin will be coming up soon(ish), and I wanted to take advantage of our plentiful free UV source, but didn't like the idea of getting one surface cooked, and the idea of manually rotating the instrument.  It turns out I have in my stockpile of gizmos some gearmotors that are perfect for the job of spinning the instrument slowly so I don't have to.  Here's the result... at an angle to keep the instrument on the skewer by gravity, but still get good sun exposure.  I also have a stockpile of DC power supplies, and the 9V one works good, spinning at 2 RPM.  The gearmotor is rated for 24V, but will turn with much lower voltage.  Yeah, I'm still an unreformed motorhead.  The violin is something I was messing with, 5 years old.

2)  R&D:  Plate weights and taptones are OK, but don't really do that great of a job in getting a consistent result in signature mode frequencies or amplitude balance, as the mode shapes and stresses are significantly different between the free plate and assembled instrument conditions.  I was thinking that a deflection test at the bridge foot of the assembled body might do better, so I surveyed 7 of my fiddles to see.  The instruments were supported at the endblocks, and a 1-kg mass was applied near the bass bridge foot, and deflection recorded.  The spring rate was then compared to the B mode frequencies:

While there is some significant correlation, I don't think it's useful.  The range here is extremely wide, as the violins #10 and below have been regraduated to excessively thin levels as experiments in getting a "fiddle sound".  It's about the same as weights and taptones... there's some correlation there, too, but also a lot of scatter.

And frequencies aren't that important directly in my view... relative amplitudes are far more important.  And I won't even bother plotting those results, as it's all scatter.

So, another investigation without usable results, as is usually the case, but again proving that this stuff is complicated.

[Peter K-G]

It probably is really complicated, but yet not that difficult, to get plate weights,tap tones and signature modes the same for different wood, or the figures you want.

[Don Noon]

16 hours ago, Peter K-G said:

It probably is really complicated, but yet not that difficult, to get plate weights,tap tones and signature modes the same for different wood, or the figures you want.

I have never been particularly worried about signature mode frequencies for a couple of reasons.  1) If I do my usual thing with tap tones, arching, and plate weights, the signature modes come out in a narrow "normal" range.  2) Even with wildly varying signature modes, as in the super-thin experiments plotted previously, they all are perfectly usable instruments, some of them quite good... and the good or bad has little to do with the signature mode frequencies, something to do with the amplitudes, and more to do with what goes on elsewhere.

[Peter K-G]

I know

But you implied that you looked for some correlation and if you for some unlikely reason would like to know how to get r^2=1 for taptones and modes I can help

[Don Noon]

In re-reading my post, I see that I didn't make clear that my original intent of the testing was to see if I could use the deflection measurements to predict/control the A0 amplitude.  It seemed to me like there should be a decent correlation there.  But when that idea crashed and burned, it wasn't much more effort to look at B frequency correlations, which are modest, so that's what I showed... even though I think it's pretty useless.

It doesn't take much research to find that good (even great) violins cover a fairly wide range of signature mode frequencies... and it doesn't take much effort to make a violin that's within that frequency range.  The hard part is making a great violin, which I think involves balancing the amplitudes of not just the signature modes, but everything else as well... so that's what I prefer to spend my time trying to do.

[scordatura]

On 8/24/2019 at 9:43 AM, Don Noon said:

A couple of distractions (among many) hindering progress on the actual instruments...

1)  A rotisserie. 

Don, is it possible that you are trying to "roverize" your violins?

[Don Noon]

14 hours ago, scordatura said:

Don, is it possible that you are trying to "roverize" your violins?

Naah... Burgess already made one of those.

[Don Noon]

Sadly, not a whole lot of progress in the shop, with continuing self-generated distractions.  And retiredness.

However, there was a milestone:  a CD by Annelle Gregory playing my violin backed by a full orchestra.  I had thought it would be just a recording of her live concert in Kiev, but no... it was a studio recording.

The CD is available on iTunes (with previews) here, or you can order from Annelle's website here.

(Hopefully this doesn't cross over the line of advertising; I'm not pushing anything for myself ...other than ego... but hoping to help out a young talent and all-around nice person who deserves it.)

[Edward Byler]

Congratulations to you Don !! It is a great accomplishment to have one of your violins played by her !

[Three13]

The violin and her playing both sound fantastic...