Don Noon's bench

[Don Noon]

33 minutes ago, Bodacious Cowboy said:

Do you have any concerns about possible long term implications of heat treating?

Other than accidental damage (it's more brittle than untreated wood), no.  But there sure won't be any shrinkage cracking.

[Don Noon]

The work for about the next year-ish.  I like to put ground on the ribs early to prevent glue ghosts.

[Rue]

Pretty! 

[Andreas Preuss]

On 2/11/2021 at 10:24 AM, Don Noon said:

The work for about the next year-ish.  I like to put ground on the ribs early to prevent glue ghosts.

Looks like a pretty prolific production for someone who is retired. 

[Anders Buen]

These were nice!

[Don Noon]

In addition to the 4 garlands on forms previously shown, I needed to get another one going for a client fairly high on the waiting list, so I made another one   The blocks on and ribs thicknessed and waiting to be bent.

Looking to some upcoming viola orders, I needed more maple big enough for my violas, and went shopping.  And then torrefying.  This is the box of wood after cooking (photo taken in a very hot blue LED light), and some ribstock that goes with it.  I needed to saw off some of the ribstock to fit my chamber, and those are the blazing white pieces to compare to the torrefied color.

For retirement, I seem to be pretty busy.

[Jim Bress]

Have you not learned that no good deed goes unpunished? 

[Michael_Molnar]

I remember how careful you were to the point of being reluctant to jump into CNC’ing, and I was like the soothsayer, Spurinna, warning you, “Beware of the Ides of CNC”. Well, they have come but not gone. You have taken CNC’ing to new heights. Nice work. The grounds look good too.

[Don Noon]

1 hour ago, Michael_Molnar said:

I was like the soothsayer, Spurinna, warning you, “Beware of the Ides of CNC”. 

And now the 3D printer is here.  It's amazing how many things I can find to print up on that thing.  But it's all good; I think that the initial speedbump of learning the software, generating the models, and making the tooling is nearly over, and soon productivity will increase.  It has been a very long speedbump of about 2 years, but worth it for my specific situation.

[Don Noon]

Although I have finished gluing in all the linings in this current batch of violins, going through the process got me thinking about the perfect lining clamp.  And thinking about that, along with having a 3D printer, got me designing and printing it.  There's not a whole lot of deviation from the modified clothespins that I have been using... only some optimization and precision to work with the linings and counterforms I'm using.  I now bre-bevel all of my linings before gluing them.  The upper and lower bout linings can be beveled before bending, but the C-bouts get all wonky that way, so I bend them first and then bevel them.

It took about a week of trial-and-error-and-error-and-etc. before I got everything tweaked in to where the spring fit right and it was to my satisfaction.  The photo below shows some of the evolution... but not all of it.  I didn't keep the hairballs and blobs and grotesque mutilations that resulted from printbed adhesion failures.

[Don Noon]

Progress on the next violin:  

Just finished installing the bass bar on the next violin; back is done and glued onto the ribs.  Closing the soundbox will be soon, which is always an interesting milestone for me, as it is the first chance to do some tests to get an idea how it will perform.

This is my first time where I used the CNC to rough out the graduations, and the arching curves were done almost entirely with splines.  It's a handy modelling tool to fit complex curves, and keeps the curvature from having abrupt changes.  The arching is basically what I would want to do by hand if I was skillful and patient enough; I didn't let the CAD tool dictate the shape.  Scraping, edgework, and the last couple of mm of graduations all still have to be done by hand the normal way.

I don't know if the arching has anything to do with it, but the M5 taptones were exceptionally high on both the top and back.  Details...

Top - 358 x 207 model (large), 16 mm arch, .37 g/cc (torrefied) density of unremarkable measured speed of sound and radiation ratio.  Top plate done with F holes and no bass bar:  58.2g, M5 358 Hz.  I decided to use a dense, center-heavy bass bar from torrefied cherry (.66 g/cc density, not as stiff as spruce) to keep the assembled B1+ frequency down.  With the bar, the plate was 63 g and M5 381 Hz.  Maximum bass bar width was planed down to 4 mm to avoid getting too heavy, and it might free up some higher plate modes.  The plate is darker in the middle than the edges due to a coating of casein which contained a small (0.5%) of CuSO4 as a mold/fungus preventative measure.  I was very surprised at how such a tiny amount of that light blue stuff would darken the mixture.

Back - the maple was also unremarkable in measured properties... .64 density, medium speed of sound and radiation ratio, also torrefied.  Back arch is 15mm, and I stopped thinning at 100.7 g and M5= 399 Hz.

The M5 plate modes are exceptionally high compared to most of what I have built so far.  The dense, center-heavy bass bar is the only adjustment I'm going to make to keep the assembled instrument from getting too stiff, as I don't think plate modes are very meaningful.  Most likely the B1+ mode of the finished violin will come out in the 540 - 560 Hz range, as seems to be my standard no matter what I do (except a bit higher for the few small 352 models I've made, and lower for violas).

Mass and distribution are my current main goals, after the even more important arching is done.  With the wood I have, 60g or slightly below for the top (without bar), and 100 g for the back seem to turn out best, and I don't want to get too far from that.

 

[Andreas Preuss]

1 hour ago, Don Noon said:

 

Mass and distribution are my current main goals, after the even more important arching is done.  With the wood I have, 60g or slightly below for the top (without bar), and 100 g for the back seem to turn out best, and I don't want to get too far from that.

 

The top wood looks really dark. 

Lower block looks unusal wide, but I suppose this was made on purpose?

Is there any purpose for the cross-grain-stud under the bass bar? (From restoration I try to avoid those studs at any means because they kill the response, though just one might be not harmful.)

Weight looks very good, though I thought torrefying spruce would eventually make it possible to go down to the region of 55g (w/o BB)

My quick calculation came up with a 2.8mm-ish thickness of the top and the border around 3.2mm. (?) 

I would have maybe left some areas of the back for outside graduations and especially the area at c bouts is hard to 'calculate'.

One parameter I am now interested in more than before is the thickness of the linings.

 

 

 

[Don Noon]

1 hour ago, Andreas Preuss said:

I thought torrefying spruce would eventually make it possible to go down to the region of 55g (w/o BB)

I have made about a half dozen violins with tops in the 51 - 55g range, but I'm finding that closer to 60 g is more to my liking, and less prone to an unusual feel, response, or sound.

The top is not quite as dark as the photo... the lighting and intensity are not right.  For reference, the table is blazing bright white melamine.

Thickness of the top is 2.3 in some areas, 2.5-2.7 generally, and a few thick areas up to 3 mm or so here and there.  The block areas of course are thicker, and I use pretty wide blocks for neck and endpin.  I often carve away some of the lower block on the bass side, but not this one (yet).

The crossgrain stud is to prevent print-thru of the bass bar to the top in this flat-ish area.  Above the eyes, the arching is tight enough to resist deformation.  The bass bar is very narrow, so print-thru might be more of a concern.  Bass bar cracks too.

[Andreas Preuss]

6 hours ago, Don Noon said:

The bass bar is very narrow, so print-thru might be more of a concern.  Bass bar cracks too.

Print through is mostly on thin tops by forcing a non-perfect fitted bar to the surface. Bass bar cracks come usually as well from badly fitted BBs, especially when there was too much curvature at either end of the bar.  For both reasons I am fitting my bars without tension.

If you like to add tension to the bass bar, I think Rene Morels way to do it is pretty safe. Tension only in the middle and on the outer quarters of the length no tension at all which means when fitting the bar no more rolling in those sections.

Do you cut away the lower block on the bass side for acoustic reasons?

[Don Noon]

I think springing the bass bar is silly and makes fitting more iffy.  So I don't do it.

The idea of trimming away the lower block on the bass side is to free up movement of the bass bar and get more low end.  I might want to do that on my smaller model, but not so much on the large one.  Realistically, I can't imagine that it would be more than a small fraction of a dB, and totally below any level of perception.  So maybe I won't do it again.

[Don Noon]

I'm often to excited to close the soundbox and do some acoustic tests that I forget to document some of the construction details, and then later I wonder what I did.  This time I remembered. 

Cornerblocks are super-nuked (thus the dark brown color) low-density Engelmann.  It's easy to cut the lining pockets by hand in this stuff, but I tried writing a CNC program to cut them, and it worked fantastically well.  Faster and better than by hand, and one less thing to stress my trashed wrist joints.

Support struts are added to give extra strength to the ribs in case a sidemount chinrest is used, which I happen to think works better.

One thing that annoys me about setting a neck is cutting the side walls of the mortise, where I find it difficult to get a clean cut without resorting to cutting with a knife from the back to the front.  Using a chisel from front to back goes at a shallow angle against the grain, with the inevitable tearout.  So I have glued in some spruce at an angle slightly steeper than the neck taper, and we'll see how that works.  It's not too hard to set up the CNC and cut the angles in a bunch of blocks, but as yet I don't know if this is worth the trouble.

[Don Noon]

First test piece of using the CNC for a few basic scroll carving tasks: cutting the outline, drilling pegholes, and scribing turns 2 and 3 of the scroll.  There's still a lot of the usual carving to do, with fishtail layouts of the turn widths, hollowing out the pegbox (ugh), chamfers and fluting.  But at least some of the layout and bandsaw operations are not needed, and CAD allowed me to tweak geometry around to get the scroll I liked, starting with the Kreisler Guarneri and "improving" it.  Eventually I'll get the CNC to do more, but this will have to do for now, so I can make progress on actual instruments.

 

[uncle duke]

39 minutes ago, Don Noon said:

 hollowing out the pegbox 

I have a cnc'd chinese pegbox.  It appears two bits were used at least, maybe three bits - I can't tell how the bottom of pegbox was smoothed. 

The bits leave machining grooves in the rounded corners telling me two different size bits were used.  The G peg box area, before it starts ramping down towards the E, appears done with the bigger bit hence the almost 2mm bit grooves left in the rounded corners in front of the nut.  My guess is around a 3 or 3.5 mm dia. bit used for the G area.

Next it appears a smaller diameter bit was used starting? at the A peg area and working it's way upwards to the G peg ramp.   I can't tell if the scroll was cut first in it's entirety or afterwards. 

Anyways, with the fiddle I have here lying on an even horizontal plane it appears the small bit starts behind the A peg area and goes straight down 10mm and then beelines 45 degrees? {7mm} down to the peg box floor - like a corner section of an octagon with ample peg clearance.  The  machining grooves from that bit have a 1mm or so spacing left for me to see in the A peg  area box corners.  I can't tell how the bottom of the pegbox was smoothed hence my thinking a third bit but the smaller bit leftover 1mm machining grooves do traverse past the E string to meet the G peg ramp smoothly.   

Maybe save for a rainy day.      

.   

[Don Noon]

Progress has been slow, mostly due to laziness, but now there's a shoulder issue in addition to the other joint problems.

Anyway, violin #33, the first of my CNC design "large" model, is together enough for acoustic testing.  The varnish is semi-done, with at least some on all active surfaces.  Playing, I think it is very good, and the impact spectrum looks good as well.  However, I think in the next revision I'll tweak the top arching to try to reduce the 940 Hz peak, and perhaps strengthen the "bridge hill" at the same time.  Right now, the arching is too smooth, with very gradual radius changes.  Cremonese is more abrupt changes than I have.

[Andreas Preuss]

17 hours ago, Don Noon said:

Right now, the arching is too smooth, with very gradual radius changes.  Cremonese is more abrupt changes than I have.

‘Cremonese’ archings have a history of 200 years and I wouldn’t generalize the form of all archings having abrupt radius changes. Carlo Bergonzi for example made extremely low and flat archings with very smooth transitions. 
 

But this is certainly not your point here. For the abrupt radius changes you might refer mostly to Strad and the typical barrel arch. And supposedly you are thinking more of the top than the back. 
 

I think that it is important to look at the points where convex and concave curvatures meet, but I am not a religious follower of a perfect curtate cycloid cross arching. (With zero aberration) 
 

In terms of pure construction technology I am convinced that it makes a difference if the arching was carved ‘as is’ or if the arching curvature is a result of controlled ‘stretch in’. It is technically possible to make a pretty thin plate and make it warp with water under string tension. (Remind you that original Cremonese bass bars were very tiny and short!) And I think for this kind of procedure you need a wider concave part at the borders . 
 

If you try on the other hand to ‘precalculate’ or estimate the degree of radius changes it might really get freaking complicated. 

——————————

Another aspect might be the total elevation of the length arch versus the elevation of the cross arch. After looking at the experiment of Jansson (in Anders Burns thread on ‘damping of the violin bodies’) there might be more possible in terms of sound than what I would have thought. So if the surface of the rib garland has lengthwise a slight arching, we create a top with a higher length arch than cross arch. (I am preparing an experiment on this to hear what effect we can expect from there) 

———————-

Otherwise, now that I have at least a little experience reading graphs, I would say that the A0 is pretty much separated and not too high. So I would think the G string lacks a bit volume and roundness.

[Don Noon]

23 minutes ago, Andreas Preuss said:

I would say that the A0 is pretty much separated and not too high. So I would think the G string lacks a bit volume and roundness.

... which is why I also take bowed semitone scales in addition to the (more repeatable and detailed) impact plot.  There are differences.  This is bowed, and A0 is very strong, as well as some CBR output:

 

[Andreas Preuss]

25 minutes ago, Don Noon said:

... which is why I also take bowed semitone scales in addition to the (more repeatable and detailed) impact plot.  There are differences.  This is bowed, and A0 is very strong, as well as some CBR output:

 

Hmmm, now the high frequency peaks all of a sudden look pretty low. That’s one thing which still bugs me on the new concept violin. 

[Don Noon]

In bowed scales, my highest played note is just under 1 kHz.  All of the higher frequency stuff is all overtones, which attenuate at 6 dB/octave... so my bowed plot will be weaker on the highs than the impact plot.  These plots are a bit iffy to compare, and much more iffy trying to compare plots taken by different people under different conditions with different equipment and techniques.  For personal record keeping and comparisons, it's a bit better.

[Don Noon]

This violin is taking forever, and now it will take a little longer.  It was together enough to test as a "tone only" entry at VMAAI, where it scored OK (I thought it sounded very good).  But I didn't like how the varnish turned out, so I'm re-doing it.

As part of the re-do, I'm looking into different color options for the varnish.  I went a bit bonkers making lakes out of everything I had in the shop, as well as a couple of other things I though would be good and bought them.

It remains to be seen how these colors will end up once they're dried and mulled into oil, and if they're lightfast.  They might be overly bright, but that can be toned down with iron rosinate or gilsonite.  So many options.

 

[Michael_Molnar]

You may want to have a green to compliment the reds.

Looks like fun.