Help salvage this mess...

Wood Butcher wrote:

> I think you are making this much more difficult than necessary. I would
> never attempt to do a 3 axis glue up like you are talking about.

I've thought it further. One dowel per side, glue checkered strips
together, then glue checkered strips into a board with the long-grain
dowels going all the way through. It will work, and it will keep
everything from wiggling.

Most critically at this point, it can give me just barely, barely, enough
wood to finish this stupid #$^@#^ thing without *another* trip to buy wood,
because I can use up some short pieces that aren't long enough for another
set of 18" wide strips. I've already paid for two scrap boards, plus all
the left-over that I couldn't use, and my budget for this box is shot to
hell.

I think for _next_ time, I'll try putting maybe three dowels into the
strips. I didn't do dowels originally because of problems getting them
centered _precisely_ on the line. Solution is to clamp two strips
securely, drill holes in both, remove one, clamp another, repeat. That way
every piece is guaranteed to line up with its neighbor, even if there are
other problems. Three dowels will probably hold everything well enough to
avoid glue sliding problems.

> 1. See Rogowski's article on gluing up tabletops in FWW #166 Dec'03.

I'll look for it. Is that one in stores now?

> 2. Use cauls during the glue up to keep the top as aligned and flat as
> possible.

I did. I guess that's what you'd call it anyway. Piece of plywood on top,
with a piece of railroad track on top of that. Yes, my workbench top is
flat too, so it seems like that should have worked a lot better than it
did.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

Frank McVey wrote:

> IME, conventional methods of doing things usually become convention
> because they work.

Yeah, well... :)

> I think a belt sander might be a bit brutal for what you want to do. I'd
> try a scraper and, if you don't have any joy, a ROS. Keep it moving.

Too much wood to remove for a ROS. It needs a lot of help. I figure after
I plane it against the grain with my crap planes, I will need to remove
another 1/16" to get the tearout. I got *bad* tearout on the maple planing
it the wrong way under much more ideal circumstances than what I will face
on this lumpy, off-kilter board.

> For your second problem, I'd simply use a block plane on a shooting board
> to trim your edges.

Don't have a block plane... They're too far out of whack to plane anyway.
Now that I've had another look, there are three blocks of strips that have
slid out of alignment, and the squares don't line up.

This one might be cut up and glued into a turning blank after all.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

Sun, Nov 16, 2003, 4:30pm [email protected] (Silvan) says:
<snip> Learning is expensive, and time consuming.

Often worth the cost.

JOAT
Of course I don't think you're a complete idiot. Some parts are
missing.

Life just ain't life without good music. - JOAT
Web Page Update 15 Nov 2003.
Some tunes I like.
http://community-2.webtv.net/Jakofalltrades/SOMETUNESILIKE/

Silvan wrote:

> Learning is expensive, and time consuming.

OK, lots of good advice from everyone, and tips for what to avoid next
time...

I looked at my scraps, determined that I have enough walnut in bits and
pieces to do my tinker toy idea, and started puttering around setting up
the sled for that.

Then I took another look at the board, after having been away from it for 36
hours. Faced with the prospect of installing hundreds of dowels and
relegating yet another board to the overflowing scrap bin, I decided to see
how bad the board would look after taking a crack at salvaging it anyway.

The board is 15" x 15.25" after I trued up the rough edges. The squares
aren't all lined up correctly, but while you can definitely find the goofs
if you look, it isn't far enough out to just *leap* out at you. I think
it's not far enough out of line to interfere with seeing move lines during
a game, and not worth trying to break it apart and re-glue.

While it should have been a nighmare to plane after the way I flipped the
grain all around, I found this wood was surprisingly tolerant. After a few
strokes, I grew more hopeful, and after only an hour or so I had something
that looked for all the world like a chess board. Especially if you close
one eye, squint with the other, and look at it sideways in bad light.

I wound up with two wavy grained maple squares with some obvious tear-out, a
couple of chatter-like gouges in one corner (from rough attacks), and one
swoopy leveling gouge that I never could plane out. Other than that, it's
smooth as glass, and one face is damn near flat enough not to wobble on a
big piece of granite.

These surface flaws might leap into glossy relief after I shellac it (and
I'm going to shellac it anyway), but I think maybe it's good enough. I
brought order to chaos, and it looks decent. Sometimes you just have to
stop being such a damn perfectionist, accept that this is the best you can
do right now, and move forward.

Now we'll see what happens when I try to do the maple inlays in the frame,
and the contrasting splines on the mitered corners of the box. I suck at
miters, so I should really think about doing more pegged butt joints, but
in for a penny, in for a pound. If I make the box sides too short due to
goofs, I will splice in some maple accents or something. The grain on
these pieces of walnut I have marked out for the sides is just begging to
wrap all the way around.

Anyway, moral of the story is maybe next time I should try it before I start
crying about how badly I've screwed it up.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

Jeremy Brown wrote:

> with the good face down (playing side down). I held the strips together
> just with pressure from my hands, but I was also careful to press each
> strip down against the flat surface. I did that to ensure there would be
> no height differences between strips (takes care of #1). Overall, less

Yup, that's what killed me I think, after doing little for the last 24 hours
but mull this over in my head. Even though I didn't crank way down on the
clamp, the clamping pressure slid everything out of whack. Could have
maybe averted some of that with a few well-placed dowels.

Given my ability with hand planes, I need to keep this realistic, and the #2
attempt is so far off that there are places where I have to lower the
entire surface evenly by 3/16" or more to compensate for problems. No
chance of my getting that right. I could do various sanding options, but
that wouldn't fix the fact that the entire board is skewed like this
slightly / /, and I'd have to knock it apart and try to re-glue it no
matter what.

So I'm going to chuck it in the scrap pile and figure out something to do
with the wood later. These squares are about the right size for train
wheels. I can also cut them into little blanks for little detail turnings
for steam domes, smokestacks, etc. Nothing gets wasted at Chateau Silvan;
not even f-ups.

I can only afford to do this one more time, so I'm going to do my ridiculous
tinker toy idea. I think it will work. Buttloads of dowels, very little
glue, no clamping.


--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

If you do start over here's a couple of suggestions which I learned
the hard way.
1. If you have a crappy saw like I did, set the fence once and don't
touch it. Rip your slats, glue them together, and then crosscut without
moving the fence. This way you maximize the probability of the
squares coming out actually square.
2. When you have the slats ripped, align them so all the pieces have
the same uphill grain direction so that if (when) you need to plane
the finished glue-up you're not planing against the grain on any piece.

Art

"Silvan" <[email protected]> wrote in message
news:[email protected]...
> Frank McVey wrote:
>
> > IME, conventional methods of doing things usually become convention
> > because they work.
>
> Yeah, well... :)
>
> > I think a belt sander might be a bit brutal for what you want to do. I'd
> > try a scraper and, if you don't have any joy, a ROS. Keep it moving.
>
> Too much wood to remove for a ROS. It needs a lot of help. I figure after
> I plane it against the grain with my crap planes, I will need to remove
> another 1/16" to get the tearout. I got *bad* tearout on the maple planing
> it the wrong way under much more ideal circumstances than what I will face
> on this lumpy, off-kilter board.
>
> > For your second problem, I'd simply use a block plane on a shooting board
> > to trim your edges.
>
> Don't have a block plane... They're too far out of whack to plane anyway.
> Now that I've had another look, there are three blocks of strips that have
> slid out of alignment, and the squares don't line up.
>
> This one might be cut up and glued into a turning blank after all.
>
> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

Silvan wrote:

> Even though I didn't crank
> way down on the clamp, the clamping pressure slid everything out of
> whack. Could have maybe averted some of that with a few well-placed
> dowels.

I've been known to use small steel dowels for this. <g> Cut the heads off
tiny brads, drill holes for the shafts on one the pieces to be glued, insert
decapitated brads into the holes pointy side out, press together to mark
holes for the other side, drill other side holes. Dry assemble. If the
holes aren't perfectly straight the pieces won't come together even. If
this happens I can sometimes make the holes for the pointy ends a little
"sloppy" so they can fit perfectly, then use a clamp across the joint to
hold them that way while the glue dries.

The points don't have to go very far into the other piece to keep them from
sliding while gluing.

-- Mark

Silvan wrote:

> I can only afford to do this one more time, so I'm going to do my
> ridiculous tinker toy idea. I think it will work. Buttloads of
> dowels, very little glue, no clamping.

Have you considered splines?

-- Mark

Hi Silvan,

IME, conventional methods of doing things usually become convention because
they work.

I think a belt sander might be a bit brutal for what you want to do. I'd
try a scraper and, if you don't have any joy, a ROS. Keep it moving.

For your second problem, I'd simply use a block plane on a shooting board to
trim your edges.

HTH

Frank



"Silvan" <[email protected]> wrote in message
news:[email protected]...
> I'm making a chess board out of multiple strips of wood. Conventional
> wisdom says use nine strips of each color, rip into strips, slide, then
cut
> off the extra squares on the ends.
>
> I just figured out the hard way why conventional wisdom says that. I
tried
> to save material, and thought I was being clever. I cut them a little
> long, glued them up, then cut them into checkered strips of the proper
> width, and rotated every other one 180 degrees. I have a nice looking
> board with a lot of potential, but now I have to get myself around a
couple
> of thorny problems:
>
> 1) the squares aren't all flat and level by a long shot, so the board
needs
> a good deal of planing... having the grain all going the same way sure
> would have been handy... I think I'm probably doomed to face horrible
> tearout, though I haven't yet tried my luck.
>
> 2) the edges after the second glue-up are a little off. I whacked them
> against a straight edge before clamping, but there were minute variations,
> and when I put a piece of frame against those two edges, the little gaps
> stick out like a sore thumb.
>
> For #1 I'm thinking maybe plane it anyway, get it flatish, then remove the
> tearout by renting a portable belt sander. Any better ideas?
>
> I have a contemporary #4 that's my de facto scrub because it's better for
> ugly work than smoothing. I have a 1960-era #5 that can take a pretty
thin
> shaving, but I don't think it's up to tackling this grain either. Could I
> do this with a scraper? Am I likely to get a scraper to work well enough
> to do this job with no previous experience using them? Or maybe try that
> router surface flattening jig?
>
> My router is a piece of crap that will *not* hold a depth setting, so that
> complicates matters. It might be the best way to go, the more I think
> about it. I've proven to myself that while I think planes are fun to use,
> I really do absolutely *suck* at using them for anything more complicated
> than truing the edge of a board or planing a very narrow face. If it's
> wider than my plane iron, I'm going to fuck it up *horribly*. That's just
> a fact right now.
>
> Dealing with #2 is quite thorny. Owing to minute variations between the
two
> times I set my rip fence, even checking it myriad times and using both a
> combination square and a spacer block to ensure the same setting, the
> "squares" are all 1/64" wider than they are long. The extra size adds up
> to 1/8" of extra length on the board, and it's in the direction opposite
> the rough edges. This means if I trim the edges to even them out the
board
> will become even more rectangular than it already is.
>
> I could trim, say, 1/16" all the way around the board, I guess. It might
> not be too obvious. Another idea is to attempt to put a tiny overhang on
> the frame members so that they cover 1/16" of the board and hide the gaps.
> This again takes space away from the perimeter squares. Of more concern,
> I'm afraid of executing it poorly. I'm envisioning a small, thin
overhang,
> and it would be exceedingly easy to break it off with whatever I use to
> craft it, whether it be the table saw, router, or chisels. Removing 1/16"
> from the board might be safer. Or maybe 1/32". I haven't actually
> measured to see just how small an amount I can remove. I could tolerate a
> tiny gap here, but the current gap is much too obvious. Planing all that
> endgrain is out because I don't have a block plane. The way they line up,
> it would be hard to plane anyway. It's like the edges of the boards
> weren't square going in, so there's a little triangular divot at every
> joint.
>
> Well, anyway, there it is. I'm on my second attempt as it is, and I have
> four days tied up in this. I want to use this one if I can come up with a
> way to save it. If not, I can cut it back up and make more turning blanks
> out of it I guess. At least I took notes on what to avoid next time.
>
> Learning is expensive, and time consuming.
>
> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

In case you do start again, I have a couple (more) tips to help avoid
the problems you're facing now. I am just finishing a chess board using
pretty much the same procedure you are.
In conflict with conventional wisdom (and directions on the bottle) I
didn't clamp my board either time I glued it. Instead, I put a sheet of wax
paper on a large flat surface. Then, I put glue on the edges of the strips
and put the edges together, then set the board on the wax paper with the
good face down (playing side down). I held the strips together just with
pressure from my hands, but I was also careful to press each strip down
against the flat surface. I did that to ensure there would be no height
differences between strips (takes care of #1). Overall, less than 5 minutes
passed between the time the glue first touched the wood to the time I left
it alone to dry. The second glue-up went the same way, but I also made sure
the squares lined up, making perfectly aligned +es (should take care of #2
if your original strips were the same size). When all was said and done, the
glue joints are plenty strong. They've held up to all the stress I've put on
them as I've experimented with finishes. They were level enough that I
easily corrected the small problems with a ROS. I think most of the uneven
spots were from resawing the wood, not gluing, so if you use good flat wood
to start with you should be fine.


"Silvan" <[email protected]> wrote in message
news:[email protected]...
> I'm making a chess board out of multiple strips of wood. Conventional
> wisdom says use nine strips of each color, rip into strips, slide, then
cut
> off the extra squares on the ends.
>
> I just figured out the hard way why conventional wisdom says that. I
tried
> to save material, and thought I was being clever. I cut them a little
> long, glued them up, then cut them into checkered strips of the proper
> width, and rotated every other one 180 degrees. I have a nice looking
> board with a lot of potential, but now I have to get myself around a
couple
> of thorny problems:
>
> 1) the squares aren't all flat and level by a long shot, so the board
needs
> a good deal of planing... having the grain all going the same way sure
> would have been handy... I think I'm probably doomed to face horrible
> tearout, though I haven't yet tried my luck.
>
> 2) the edges after the second glue-up are a little off. I whacked them
> against a straight edge before clamping, but there were minute variations,
> and when I put a piece of frame against those two edges, the little gaps
> stick out like a sore thumb.
>
> For #1 I'm thinking maybe plane it anyway, get it flatish, then remove the
> tearout by renting a portable belt sander. Any better ideas?
>
> I have a contemporary #4 that's my de facto scrub because it's better for
> ugly work than smoothing. I have a 1960-era #5 that can take a pretty
thin
> shaving, but I don't think it's up to tackling this grain either. Could I
> do this with a scraper? Am I likely to get a scraper to work well enough
> to do this job with no previous experience using them? Or maybe try that
> router surface flattening jig?
>
> My router is a piece of crap that will *not* hold a depth setting, so that
> complicates matters. It might be the best way to go, the more I think
> about it. I've proven to myself that while I think planes are fun to use,
> I really do absolutely *suck* at using them for anything more complicated
> than truing the edge of a board or planing a very narrow face. If it's
> wider than my plane iron, I'm going to fuck it up *horribly*. That's just
> a fact right now.
>
> Dealing with #2 is quite thorny. Owing to minute variations between the
two
> times I set my rip fence, even checking it myriad times and using both a
> combination square and a spacer block to ensure the same setting, the
> "squares" are all 1/64" wider than they are long. The extra size adds up
> to 1/8" of extra length on the board, and it's in the direction opposite
> the rough edges. This means if I trim the edges to even them out the
board
> will become even more rectangular than it already is.
>
> I could trim, say, 1/16" all the way around the board, I guess. It might
> not be too obvious. Another idea is to attempt to put a tiny overhang on
> the frame members so that they cover 1/16" of the board and hide the gaps.
> This again takes space away from the perimeter squares. Of more concern,
> I'm afraid of executing it poorly. I'm envisioning a small, thin
overhang,
> and it would be exceedingly easy to break it off with whatever I use to
> craft it, whether it be the table saw, router, or chisels. Removing 1/16"
> from the board might be safer. Or maybe 1/32". I haven't actually
> measured to see just how small an amount I can remove. I could tolerate a
> tiny gap here, but the current gap is much too obvious. Planing all that
> endgrain is out because I don't have a block plane. The way they line up,
> it would be hard to plane anyway. It's like the edges of the boards
> weren't square going in, so there's a little triangular divot at every
> joint.
>
> Well, anyway, there it is. I'm on my second attempt as it is, and I have
> four days tied up in this. I want to use this one if I can come up with a
> way to save it. If not, I can cut it back up and make more turning blanks
> out of it I guess. At least I took notes on what to avoid next time.
>
> Learning is expensive, and time consuming.
>
> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

"Silvan" <[email protected]> wrote in message
news:[email protected]...
> Jeremy Brown wrote:
>
> > with the good face down (playing side down). I held the strips together
> > just with pressure from my hands, but I was also careful to press each
> > strip down against the flat surface. I did that to ensure there would be
> > no height differences between strips (takes care of #1). Overall, less
>
> Yup, that's what killed me I think, after doing little for the last 24
hours
> but mull this over in my head. Even though I didn't crank way down on the
> clamp, the clamping pressure slid everything out of whack. Could have
> maybe averted some of that with a few well-placed dowels.
>
> Given my ability with hand planes, I need to keep this realistic, and the
#2
> attempt is so far off that there are places where I have to lower the
> entire surface evenly by 3/16" or more to compensate for problems. No
> chance of my getting that right. I could do various sanding options, but
> that wouldn't fix the fact that the entire board is skewed like this
> slightly / /, and I'd have to knock it apart and try to re-glue it no
> matter what.
>
> So I'm going to chuck it in the scrap pile and figure out something to do
> with the wood later. These squares are about the right size for train
> wheels. I can also cut them into little blanks for little detail turnings
> for steam domes, smokestacks, etc. Nothing gets wasted at Chateau Silvan;
> not even f-ups.
>
> I can only afford to do this one more time, so I'm going to do my
ridiculous
> tinker toy idea. I think it will work. Buttloads of dowels, very little
> glue, no clamping.

the chances you'll get every one of those hundreds of dowels in exactly the
same spot are pretty slim.

you could just rip the board into blocks again with your sled and a stop
block, and try again. the squares would just be a little smaller.

"Silvan" <[email protected]> wrote in message
news:[email protected]...
> Charles Spitzer wrote:
>
> > you could just rip the board into blocks again with your sled and a stop
> > block, and try again. the squares would just be a little smaller.
>
> Too small to be useful unless I can get my lathe on Christmas Eve and
> successfully turn a small chess set before dawn. :) These are already at
> the smallest size that are properly suited for the smallest, cheapest
> store-bought pieces I can find.
>
> Plus I'd have to use a taper jig to rip exactly on the glue lines I
> think... :)

a taper jig is 2 straight pieces of board and a hinge. a technique i use
when cutting tiles on a 10" tile saw is to use a metal straightedge
alongside the blade (when it is off). that will show you one side of the
kerf. you can move the to-be-cut piece around to get it exactly on the line
without having to measure or know the angle.

> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

The way I'd do it would be to keep it in strips. Glue up alternating
light/dark strips say 2" wide. Then when they dry cut into 2" strips
across the colors. Turn each strip 180 degrees to make a checkerboard
pattern and reglue those strips back together.

On Sun, 16 Nov 2003 21:38:42 -0500, Silvan
<[email protected]> wrote:

>I thought I'd run this idea by everyone. My made-from-strips board-making
>problems are largely with alignment issues in all three axes during
>glue-up. I took every possible precaution to cut and clamp these pieces
>well, but I have low-end equipment, and I just don't think I can do any
>snipped....

Wood Butcher wrote:

> 1. If you have a crappy saw like I did, set the fence once and don't
> touch it. Rip your slats, glue them together, and then crosscut without
> moving the fence. This way you maximize the probability of the
> squares coming out actually square.

What he said. :) I tried very hard to avoid having to move it, but I
couldn't true up a rough edge after the first glue-up without my crosscut
sled. I pretty much can't trust any type of crosscut unless it's
stabilized by riding on both outside table edges and both miter slots
simultaneously.

> 2. When you have the slats ripped, align them so all the pieces have
> the same uphill grain direction so that if (when) you need to plane
> the finished glue-up you're not planing against the grain on any piece.

Figuring out which way is which is going to be a big part of my problem. I
need to do more homework here to make sure I figure it out in a timely
enough fashion. Some of these boards have swirly grain, so it's harder to
pick one direction for the entire length and stick to it. I may have to
make choices at the individual square level.

Or avoid planing it as a practical concesesion to the reality that I'm not
very good with hand planes yet, and I don't have terribly good planes to
use either.

Third time's the charm, right? :)

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

In article <[email protected]>,
[email protected] says...
> Frank McVey wrote:
>
> > IME, conventional methods of doing things usually become convention
> > because they work.
>
> Yeah, well... :)
>
> > I think a belt sander might be a bit brutal for what you want to do. I'd
> > try a scraper and, if you don't have any joy, a ROS. Keep it moving.
>
> Too much wood to remove for a ROS. It needs a lot of help. I figure after
> I plane it against the grain with my crap planes, I will need to remove
> another 1/16" to get the tearout. I got *bad* tearout on the maple planing
> it the wrong way under much more ideal circumstances than what I will face
> on this lumpy, off-kilter board.

Maybe you could find someone with a drum sander? Don't sell a good
ROS short, I've planed down panels with some pretty bad mis-alignments
(back in my early days of woodworking mind you ;-) ). Just start with a
fairly course grit. It may be painful, but will probably be less
painful than having to deal with severe tearout.

Charles Spitzer wrote:

> you could just rip the board into blocks again with your sled and a stop
> block, and try again. the squares would just be a little smaller.

Too small to be useful unless I can get my lathe on Christmas Eve and
successfully turn a small chess set before dawn. :) These are already at
the smallest size that are properly suited for the smallest, cheapest
store-bought pieces I can find.

Plus I'd have to use a taper jig to rip exactly on the glue lines I
think... :)

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

Larry C in Auburn, WA wrote:

> squares. To level the board you could take it to any cabinet shop and
> they
> can use their thickness sander to level the top. Shouldn't cost you much
> at all and it sounds like just a couple of passes will level it.

Hmmm... Sounds appealing, but I'm not aware of any cabinet shops
hereabouts.

Is that one of those gigantic things with belts 36" wide? They have one at
the factory in Asheville. Maybe I can find a way to get down there this
week.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

I thought I'd run this idea by everyone. My made-from-strips board-making
problems are largely with alignment issues in all three axes during
glue-up. I took every possible precaution to cut and clamp these pieces
well, but I have low-end equipment, and I just don't think I can do any
better. I might salvage the most recent one, but I'd like to be
uncompromising on this project, so I'm thinking hard about taking another
fresh start.

So what I just tried out is the original tinker toy idea I was tossing
around back when I asked for my initial advice on this. After doing three
pieces, it _seems_ like it might work, but I thought I'd put it out for
people to point out my obvious stupidity before I start to commit any more
valuable wood to this endeavor.

1) clamp a stop block to my crosscut sled *securely*, maybe even screwing it
down to be sure it never moves.

2) joint the edge of a board so that it's perfectly perpendicular to the
face, and perfectly straight

2a) test and mark grain direction
2b) mark which face is up

3) cut off a small strip to get a straight edge

4) push to the stop, crosscut into strips

5) take the strips, rotate 90 degrees and push into the stop, effectively
ripping like a crosscut

6) take the resulting squares to the DP, to a jig built to *just* hold a
square

7) use the DP to drill four holes into every piece, keeping the same face
out

8) stick dowels into all the little holes, and assemble like tinker toys,
then surround with a frame

*IF* I get all the holes drilled perfectly, this will solve the alignment
problems. If the two woods are of a slightly different thickness (and they
will be), one face should still come out very close to being in the same
plane. I could even drill the end-grain holes all the way through, and
have long dowels that run through the entire piece.

Problems are:

* that's a hell of a lot of glue work to do in 15 minutes
* if the holes don't come out perfect, everything will be out of whack
* clamping from two different directions simultaneously while keeping
everything *square*

Seems with all the parts interlocking, I could get away with light glue.
Maybe just a few dots, which would speed up the spreading. Thoughts?

Other than the obvious "wow, that's a lot of work, why don't you just cut
strips and follow conventional wisdom" type responses, does this sound
plausible?

I already thought of doing alternating strips, then cutting into checkered
strips and then running dowels through those. The problem is that my
checkered strips are made of segments that aren't perfectly parallel to
each other, making for an irregular distance from the fence or doweling
jig. I could try planing the board after the first glue-up, but the last
time I tried that, I wound up with a chucker. I'm hoping to minimize the
amount of opportunity I have to screw up the board surface with my poor
planing skill, and I want to do light planing only.

With so many interlocks, could I get by without clamping at all? I've got
stuff I glued together without clamping that has lasted for years, and this
doesn't have to take any weight. It *will* be the lid of a hinged box, if
that makes any difference. Hinges attached to the surrounding frame. Not
having to clamp would simplfy matters. Seems like this thing might hold
together fairly well with no glue at all. Especially if I put two dowels
into every side of every square.

64 squares * 8 dowels = 512 alignment pegs minus ((8 * 4) * 2 = 64) for edge
pieces = 448 dowels. Probably want to buy ready-made dowels for that, but
with 448 dowels holding everything together, the glue wouldn't have to be
very strong.

OK, I'm babbling now. How *else* could I ensure proper alignment in all
three axes when careful cutting and careful clamping aren't doing it? The
bottom line is that the pieces are coming off the saw parllel to each
other, but not perpendicular to the earth. They're ever so slightly
slanted //// so they squeeze into funny shapes under pressure. When I
tried jointing them with a plane, I just made every strip a different
width, which resulted in a total ruin. I have to glue them straight off
the saw, or not do this at all, so I have to live with and compensate for
the minute slant. I can't get my TS to behave any better than this. This
is all a Skil 3400 is good for.

Blah blah blah... Sorry. Thinking as much to myself as asking questions.
Writing helps me think.

I've about convinced myself to try it anyway, FWIW. What the hell. I
already have $40 worth of future turning blanks. May as well make this the
most obscenely expensive simple chess box the world has ever seen and go in
for another $20 or so.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

I would definitely just trim the edge to straighten the pieces out. I don't
think 1/16" will be noticeable with or without a piece sitting on the edge
squares. To level the board you could take it to any cabinet shop and they
can use their thickness sander to level the top. Shouldn't cost you much at
all and it sounds like just a couple of passes will level it.
--
Larry C in Auburn, WA

"Silvan" <[email protected]> wrote in message
news:[email protected]...
> I'm making a chess board out of multiple strips of wood. Conventional
> wisdom says use nine strips of each color, rip into strips, slide, then
cut
> off the extra squares on the ends.
>
> I just figured out the hard way why conventional wisdom says that. I
tried
> to save material, and thought I was being clever. I cut them a little
> long, glued them up, then cut them into checkered strips of the proper
> width, and rotated every other one 180 degrees. I have a nice looking
> board with a lot of potential, but now I have to get myself around a
couple
> of thorny problems:
>
> 1) the squares aren't all flat and level by a long shot, so the board
needs
> a good deal of planing... having the grain all going the same way sure
> would have been handy... I think I'm probably doomed to face horrible
> tearout, though I haven't yet tried my luck.
>
> 2) the edges after the second glue-up are a little off. I whacked them
> against a straight edge before clamping, but there were minute variations,
> and when I put a piece of frame against those two edges, the little gaps
> stick out like a sore thumb.
>
> For #1 I'm thinking maybe plane it anyway, get it flatish, then remove the
> tearout by renting a portable belt sander. Any better ideas?
>
> I have a contemporary #4 that's my de facto scrub because it's better for
> ugly work than smoothing. I have a 1960-era #5 that can take a pretty
thin
> shaving, but I don't think it's up to tackling this grain either. Could I
> do this with a scraper? Am I likely to get a scraper to work well enough
> to do this job with no previous experience using them? Or maybe try that
> router surface flattening jig?
>
> My router is a piece of crap that will *not* hold a depth setting, so that
> complicates matters. It might be the best way to go, the more I think
> about it. I've proven to myself that while I think planes are fun to use,
> I really do absolutely *suck* at using them for anything more complicated
> than truing the edge of a board or planing a very narrow face. If it's
> wider than my plane iron, I'm going to fuck it up *horribly*. That's just
> a fact right now.
>
> Dealing with #2 is quite thorny. Owing to minute variations between the
two
> times I set my rip fence, even checking it myriad times and using both a
> combination square and a spacer block to ensure the same setting, the
> "squares" are all 1/64" wider than they are long. The extra size adds up
> to 1/8" of extra length on the board, and it's in the direction opposite
> the rough edges. This means if I trim the edges to even them out the
board
> will become even more rectangular than it already is.
>
> I could trim, say, 1/16" all the way around the board, I guess. It might
> not be too obvious. Another idea is to attempt to put a tiny overhang on
> the frame members so that they cover 1/16" of the board and hide the gaps.
> This again takes space away from the perimeter squares. Of more concern,
> I'm afraid of executing it poorly. I'm envisioning a small, thin
overhang,
> and it would be exceedingly easy to break it off with whatever I use to
> craft it, whether it be the table saw, router, or chisels. Removing 1/16"
> from the board might be safer. Or maybe 1/32". I haven't actually
> measured to see just how small an amount I can remove. I could tolerate a
> tiny gap here, but the current gap is much too obvious. Planing all that
> endgrain is out because I don't have a block plane. The way they line up,
> it would be hard to plane anyway. It's like the edges of the boards
> weren't square going in, so there's a little triangular divot at every
> joint.
>
> Well, anyway, there it is. I'm on my second attempt as it is, and I have
> four days tied up in this. I want to use this one if I can come up with a
> way to save it. If not, I can cut it back up and make more turning blanks
> out of it I guess. At least I took notes on what to avoid next time.
>
> Learning is expensive, and time consuming.
>
> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

On Sun, 16 Nov 2003 19:53:47 -0500, Silvan
<[email protected]> wrote:

>T. wrote:
>
>> Sun, Nov 16, 2003, 4:30pm [email protected] (Silvan) says:
>> <snip> Learning is expensive, and time consuming.
>>
>> Often worth the cost.
>
>I hope so. I'm trying to do it right this time. I just finished writing
>out three pages of notes so I can look back and avoid doing stupid things
>next time. :)

ya anywhere near wake forest? i got a decent 4 x 24 belt sander you
could use. you buy the belts. skeez

[email protected] wrote:

> ya anywhere near wake forest? i got a decent 4 x 24 belt sander you
> could use. you buy the belts. skeez

I might be going to Raleigh later this week. We'll see what I come up with
between then and now. I appreciate the offer!

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

I think you are making this much more difficult than necessary. I would
never attempt to do a 3 axis glue up like you are talking about.
See my response to your other thread for 2 suggestions.
Here's a few more.
1. See Rogowski's article on gluing up tabletops in FWW #166 Dec'03.
What you will be doing is building a miniature tabletop. Adapt his method
of using the jointer to your tablesaw. Substitute your tabletop for the
jointers fence and your saw blade for the jointers knives.
2. Use cauls during the glue up to keep the top as aligned and flat as
possible.
3. When dry repeat steps 1&2 for the crosscuts.

Art

"Silvan" <[email protected]> wrote in message
news:[email protected]...
> I thought I'd run this idea by everyone. My made-from-strips board-making
> problems are largely with alignment issues in all three axes during
> glue-up. I took every possible precaution to cut and clamp these pieces
> well, but I have low-end equipment, and I just don't think I can do any
> better. I might salvage the most recent one, but I'd like to be
> uncompromising on this project, so I'm thinking hard about taking another
> fresh start.
>
> So what I just tried out is the original tinker toy idea I was tossing
> around back when I asked for my initial advice on this. After doing three
> pieces, it _seems_ like it might work, but I thought I'd put it out for
> people to point out my obvious stupidity before I start to commit any more
> valuable wood to this endeavor.
>
> 1) clamp a stop block to my crosscut sled *securely*, maybe even screwing it
> down to be sure it never moves.
>
> 2) joint the edge of a board so that it's perfectly perpendicular to the
> face, and perfectly straight
>
> 2a) test and mark grain direction
> 2b) mark which face is up
>
> 3) cut off a small strip to get a straight edge
>
> 4) push to the stop, crosscut into strips
>
> 5) take the strips, rotate 90 degrees and push into the stop, effectively
> ripping like a crosscut
>
> 6) take the resulting squares to the DP, to a jig built to *just* hold a
> square
>
> 7) use the DP to drill four holes into every piece, keeping the same face
> out
>
> 8) stick dowels into all the little holes, and assemble like tinker toys,
> then surround with a frame
>
> *IF* I get all the holes drilled perfectly, this will solve the alignment
> problems. If the two woods are of a slightly different thickness (and they
> will be), one face should still come out very close to being in the same
> plane. I could even drill the end-grain holes all the way through, and
> have long dowels that run through the entire piece.
>
> Problems are:
>
> * that's a hell of a lot of glue work to do in 15 minutes
> * if the holes don't come out perfect, everything will be out of whack
> * clamping from two different directions simultaneously while keeping
> everything *square*
>
> Seems with all the parts interlocking, I could get away with light glue.
> Maybe just a few dots, which would speed up the spreading. Thoughts?
>
> Other than the obvious "wow, that's a lot of work, why don't you just cut
> strips and follow conventional wisdom" type responses, does this sound
> plausible?
>
> I already thought of doing alternating strips, then cutting into checkered
> strips and then running dowels through those. The problem is that my
> checkered strips are made of segments that aren't perfectly parallel to
> each other, making for an irregular distance from the fence or doweling
> jig. I could try planing the board after the first glue-up, but the last
> time I tried that, I wound up with a chucker. I'm hoping to minimize the
> amount of opportunity I have to screw up the board surface with my poor
> planing skill, and I want to do light planing only.
>
> With so many interlocks, could I get by without clamping at all? I've got
> stuff I glued together without clamping that has lasted for years, and this
> doesn't have to take any weight. It *will* be the lid of a hinged box, if
> that makes any difference. Hinges attached to the surrounding frame. Not
> having to clamp would simplfy matters. Seems like this thing might hold
> together fairly well with no glue at all. Especially if I put two dowels
> into every side of every square.
>
> 64 squares * 8 dowels = 512 alignment pegs minus ((8 * 4) * 2 = 64) for edge
> pieces = 448 dowels. Probably want to buy ready-made dowels for that, but
> with 448 dowels holding everything together, the glue wouldn't have to be
> very strong.
>
> OK, I'm babbling now. How *else* could I ensure proper alignment in all
> three axes when careful cutting and careful clamping aren't doing it? The
> bottom line is that the pieces are coming off the saw parllel to each
> other, but not perpendicular to the earth. They're ever so slightly
> slanted //// so they squeeze into funny shapes under pressure. When I
> tried jointing them with a plane, I just made every strip a different
> width, which resulted in a total ruin. I have to glue them straight off
> the saw, or not do this at all, so I have to live with and compensate for
> the minute slant. I can't get my TS to behave any better than this. This
> is all a Skil 3400 is good for.
>
> Blah blah blah... Sorry. Thinking as much to myself as asking questions.
> Writing helps me think.
>
> I've about convinced myself to try it anyway, FWIW. What the hell. I
> already have $40 worth of future turning blanks. May as well make this the
> most obscenely expensive simple chess box the world has ever seen and go in
> for another $20 or so.
>
> --
> Michael McIntyre ---- Silvan <[email protected]>
> Linux fanatic, and certified Geek; registered Linux user #243621
> http://www.geocities.com/Paris/Rue/5407/
>

T. wrote:

> Sun, Nov 16, 2003, 4:30pm [email protected] (Silvan) says:
> <snip> Learning is expensive, and time consuming.
>
> Often worth the cost.

I hope so. I'm trying to do it right this time. I just finished writing
out three pages of notes so I can look back and avoid doing stupid things
next time. :)

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/

Mark & Juanita wrote:

> Maybe you could find someone with a drum sander? Don't sell a good
> ROS short, I've planed down panels with some pretty bad mis-alignments
> (back in my early days of woodworking mind you ;-) ). Just start with a
> fairly course grit. It may be painful, but will probably be less
> painful than having to deal with severe tearout.

Did I mention I don't *have* a good ROS? :)

Maybe I could rent one though.

--
Michael McIntyre ---- Silvan <[email protected]>
Linux fanatic, and certified Geek; registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/