amateur David Rosales' Bench

[Davide Sora]

3 hours ago, MikeC said:

That drill guide looks good.  You'll have to find some way to position it over the scroll/pegbox. Shouldn't be too difficult to do.

Davide I like that one!  I like antique tools! 

I too love ancient tools, unfortunately the one in the photo I posted it's from the web, not my property

[David Rosales]

On 1/31/2022 at 4:02 AM, Davide Sora said:

The original G form is 350.4 mm long when measured on the center line by reconstructing the curves of the blocks.

The 347 mm measurement is the one indicated by Pollens in his book, and is taken on the corners of the block recesses (left side, to be precise) because it is an accurate measurement that can be taken with a caliper, while the one on the center line is only theoretical because you have to add the curve of the blocks, which does not physically exist in the form.

I also realized that the bottom pin hole I made on the template is slightly off center to the left making the outline 1 mm thinner on the upper bout and 1 mm thicker on the lower bout.  I think I'm going to leave it as is.  

[HoGo]

It's safer to drill the pinholes first and then trace centerline right through their centers and follow with the outline.

[Davide Sora]

2 hours ago, David Rosales said:

I also realized that the bottom pin hole I made on the template is slightly off center to the left making the outline 1 mm thinner on the upper bout and 1 mm thicker on the lower bout.  I think I'm going to leave it as is.  

Well you can't do much about it now, call it personalization

I do not use this system with the half template and the pins, precisely because if it is not done with extreme precision (holes made with the drill press and perfectly centered, holes in the templates made at the same time and without any play with the pins) it easily gives rise to this kind of problems.

1 hour ago, HoGo said:

It's safer to drill the pinholes first and then trace centerline right through their centers and follow with the outline.

Great Advice!

And I would add making the holes for the pins with the sheet for the template  (before making the outline) and the wood of the form clamped together, obtaining the pins from two drill bits identical to the one you use for the holes.

[HoGo]

That's why I never throw away broken drillbits. They make perfect pins for alignment during gluing.

[David Rosales]

Update to Making the Mold: 

After practicing with the coping saw a bit, I felt comfortable enough to try to cut out the mold. I left more material outside the line than I wanted to but I also wanted to end up with a squared surface. Tradeoffs. Setting up the saw for a pull-stroke (the blade teeth angled toward the handle) made it easier for me.

I then started to file the edges but it was slow-going. I tried different filing techniques that I read about including placing a squared corner from the plywood I cut off under the file and abutting the form as I filed to keep it square. All methods are all very slow and I also got some tear out on the edges. 

I'll finish filing the edges and cut out the block recesses next time.

 

[chiaroscuro_violins]

/

[David Rosales]

Just now, chiaroscuro_violins said:

I'd use a knife to cut the edges square and to the line, then use a file just to smooth over the knife cuts.  Files take forever and it's darn near impossible not to round the surface.  Check the squareness with a machinist's square, then take off wood where you need to.  Make sure your knife is very sharp!  With plywood, you'll always be cutting against the grain.  

I think you're right. I need to buy the squares and the knives now! I thought I could save those purchases for later steps.

[Wood Butcher]

You can also try a largeish thumb plane, for rough shaping the convex sections down to the line.

[avandesande]

37 minutes ago, David Rosales said:

I think you're right. I need to buy the squares and the knives now! I thought I could save those purchases for later steps.

Just think of your hobby as being a tool collector

[David Rosales]

Just now, avandesande said:

Just think of your hobby as being a tool collector

Haha, yes, soooo many tools! I just ordered a pfeil 12mm knife and 2- and 4-inch engineer's squares. 

[HoGo]

I would make sanding blocks with square edge and finish the outline of mould while holding it on a flat surface like laminated particle board (old cabinet door or such). Similar to blocks that Davide uses for shaping edges of plates. He uses the "sharkskin" abrasive layer but good 80 grit Norton belt sandpaper will do the job excellently here and keep it square without thinking about it.

[David Rosales]

Just now, HoGo said:

I would make sanding blocks with square edge and finish the outline of mould while holding it on a flat surface like laminated particle board (old cabinet door or such). Similar to blocks that Davide uses for shaping edges of plates. He uses the "sharkskin" abrasive layer but good 80 grit Norton belt sandpaper will do the job excellently here and keep it square without thinking about it.

Oh yeah good idea. I’ll try that while I wait for my knife and squares  

[Davide Sora]

Yep, you can't work without a fairly accurate square, although extreme precision is not strictly necessary in this kind of work, it is quite useful and desirable.

To work the side of the form I use the same system that I use for the edges of the plates, i.e. abrasive sheets glued on squared blocks of wood, using a glass surface as a support and reference. I use these blocks to set the square, but if there is a lot of wood to take off I use good quality rasps instead of files. They are definitely more expensive but much faster and easy to control. Then I finish the work with the wooden blocks and with the sanding sheets, keeping the surfaces squared becomes really a trivial thing, and without having to continuosly check with the square.

Edit: I posted this just before seeing HoGo's post, which suggests the same thing

 

NB: it is important that the form (in this case the back) is placed on a cork sheet with a thickness of about 2 mm to keep it raised from the glass, for optimal contact with the abrasive surface of the sanding block.

 

[Jim Bress]

1 hour ago, Davide Sora said:

Yep, you can't work without a fairly accurate square, although extreme precision is not strictly necessary in this kind of work, it is quite useful and desirable.

To work the side of the form I use the same system that I use for the edges of the plates, i.e. abrasive sheets glued on squared blocks of wood, using a glass surface as a support and reference. I use these blocks to set the square, but if there is a lot of wood to take off I use good quality rasps instead of files. They are definitely more expensive but much faster and easy to control. Then I finish the work with the wooden blocks and with the sanding sheets, keeping the surfaces squared becomes really a trivial thing, and without having to continuosly check with the square.

Edit: I posted this just before seeing HoGo's post, which suggests the same thing

 

NB: it is important that the form (in this case the back) is placed on a cork sheet with a thickness of about 2 mm to keep it raised from the glass, for optimal contact with the abrasive surface of the sanding block.

 

Davide, How are you transferring your scribe/pencil line to the outer edge surface of the plates to keep from having to continually lift the plate up to check? If that's what you're doing. Thanks

[uncle duke]

4 hours ago, David Rosales said:

Update to Making the Mold: 

 

A router table is the best way I have found to get the final shape of a mold.  It can be the most dangerous too.

Is that an Amati model you are trying to make?

[Davide Sora]

10 hours ago, Jim Bress said:

Davide, How are you transferring your scribe/pencil line to the outer edge surface of the plates to keep from having to continually lift the plate up to check? If that's what you're doing. Thanks

Hi Jim,

I do not transfer the mark to the outside, but I start by making a bevel with the files by approaching the mark on the inside, and I use this bevel as a reference to see how much and where I should or should not remove wood with the sanding blocks. While I work with the sanding blocks I look at this bevel that progressively reduces in width, being careful not to touch the points where it is reduced to zero and insisting more where it is wider. Of course you have to check quite frequently, but this speed up the work a lot because it is much more immediate to evaluate the width of the bevel than to see the distance from a thin pencil line, which is more difficult to evaluate accurately.

Here you can see how I proceed, note that I use my finger to keep the mark of the point where I have to remove, or where I must not remove (depending on the case):

https://youtu.be/JnYPrF2_v-Q?t=358

To stay in the topic of the OP, I do the same when I make the forms

[Jim Bress]

Simple and brilliant, thanks Davide.  

[HoGo]

Good quality sandpaper will be comparable to a fine rasp both by ease of cut and fine surface.

I've been using Norton sand paper sold by a meter for large belt sanders and I suppose the 80 grit will remove the 1-2 mm of wood here quite effectively.

Lifting the wood from the surface a mm or two with a spacer of some kind is good tip from Davide, it will also create space for the dust.

[Bodacious Cowboy]

Not really relevant to Mr Rosales, who is cutting by hand, but with a well set up bandsaw, a sharp blade and a little practice, you can cut the mold right to the line without any clean-up. The ribs don't mind the saw marks.

[David Rosales]

17 hours ago, uncle duke said:

A router table is the best way I have found to get the final shape of a mold.  It can be the most dangerous too.

Is that an Amati model you are trying to make?

Hey Duke. No this is based off the Strad G form. It’s about 1 mm off on the bouts though. 

[JacksonMaberry]

If it's not too late, I'd encourage you to reevaluate the corners of your template. To my eye they look quite long, and once you telegraph that shape to the blocks and trim them, even if you fully remove the pencil line you will end up with comically long rib corners. Ask me how I know! Getting corners to look good was the most challenging thing in my early making. Best of luck!

[David Rosales]

57 minutes ago, JacksonMaberry said:

If it's not too late, I'd encourage you to reevaluate the corners of your template. To my eye they look quite long, and once you telegraph that shape to the blocks and trim them, even if you fully remove the pencil line you will end up with comically long rib corners. Ask me how I know! Getting corners to look good was the most challenging thing in my early making. Best of luck!

They looked long to me too, but I'm fairly certain they match Addie's drawing. I'm going to try it as is for the first instrument and see what I end up with! lol

[JacksonMaberry]

10 minutes ago, David Rosales said:

They looked long to me too, but I'm fairly certain they match Addie's drawing. I'm going to try it as is for the first instrument and see what I end up with! lol

I've never trusted the corners in those drawings, which I think are otherwise good, despite that they are allegedly drawn from artifacts in the museo Stradivari. It's easy enough to extrapolate corners from other measurements of the form using a compass and straightedge. 

But since you're going ahead, do they match the inside of the line or the outside? It may seem like a minor difference, but because of the compounding effect and pencil creep, what starts as a 10th of a mm ends up being akin to a mile. 

Ultimately the shape of the corners has no important impact on the sound and performance of the instrument. But they are critical aesthetic elements and, when they look off, can really effect a players perspective. Just a tip as you continue to gain experience. Cheers!

[MikeC]

If you can get a Strad poster, trace the corner blocks from the CT scan.   It's hard to beat that for getting the corner length right.