drawer slide adjustment

"Tom Watson" wrote:

> A recent thread on using tape measures in the shop got me to
> thinking
> about this.
>
> I went on Starrett's website and found that the most that they will
> write up a Certificate of Accuracy for on a tape is +/- 1/32".
<snip>

Still remember from a surveying course that steel tape "stretch" had
to be accounted for when measuring with one.

Long since forgot how to do it, just remember it could be a problem.

Lew

"Mike Marlow" wrote:

> Maybe. Depends. I can answer your question this way - when laying
> out for long runs like say a foundation, or a deck, or the likes
> where this kind of measurement is common, 5/16 is not close enough.
> Dead on is close enough - recognizing that dead on has a certain
> error factor built into it by the tolerance of the tape itself.
--------------------------------------------

An old trick to minimize measuring instrument errors is to layout left
to right, then turn instrument around and layout right to left, then
split the difference between the two marks for final location
(measurement).

Errors in the measuring instrument are self cancelling as well as
human errors.

Lew

<[email protected]> wrote
There is no such thing. There is always a tolerance. It's not
always
clear to me what the tolerance ought to be.
<snip>

Your analitis is showing.

$750 for materials?

Must be a small project.

Lew

Pat Barber <[email protected]> wrote in
news:[email protected]:

> Holy Flying Firetrucks don't do it for me.....

I find the mental image somewhat amusing. Holy Flying Firetrucks, Batman!

Puckdropper
--

[email protected] wrote:

> There is no such thing. There is always a tolerance. It's not
> always
> clear to me what the tolerance ought to be.
>
>> and often "close
>> enough" will work
>
> Of course "close enough" will work. You going to tell me you cut
> accurate to the nanometer? And your corners are square to a
> millionth of a degree? I doubt it. There's *always* a tolerance,
> which is to say, there's some measure of "close enough".
>
>> if you can shim "components" into square during final
>> assembly of the cabinet. However, once you start shimming you're wasting
>> time that could better be spent doing something else, and there is NEVER
>> a guarantee of satisfactory results.
>
> Well, if the alternative is to start over it's hard to see how
> shimming
> is "wasting time". If I have to spend, I dunno, 500-1000 hours
> making
> the project again vs. 1 hour shimming? I suppose the second time
> it
> wouldn't take so long, since many of those hours were learning hours,
> but I'd learn less the second time around, and it wouldn't be
> interesting
> to make exactly the same project again.
>
>> 'Measuring diagonals' will tell you whether the four sides of a
>> "component" (drawer, door, drawer front, or the casework) are square to
>> each other; using an 'accurate square' of any kind will also tell you if
>> the corners are square (keep in mind that without proper preparation,
>> the stock between square corners could still be twisted, bent, warped,
>> or vary in thickness, and still cause problems).
>
> Of course I can measure diagonals, though this won't find twist.

Let me put it to you gently, Bubba ... I spent some time outlining what
an _experienced_ cabinet maker does to insure that he does NOT have the
problems you are experiencing, now you want to fucking argue with me?

Kiss my ass, shithead!

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

"Mike Marlow" <[email protected]> wrote in message
news:[email protected]...
>
> <[email protected]> wrote in message
> news:[email protected]...
> On Dec 3, 9:59 am, "Mike Marlow" <[email protected]> wrote:
>>
>> > By the time I would get to the point where I'm
>> > supposed to check if the assembly is square I've already been fighting
>> > against the glue and I'm at the end of my open time and probably tired
>> > and drained from the stress of it all.
>>
>> Try things like cutting a stick to the predetermined length for the
>> diagonal, and either tacking it or screwing it in place to hold the
>> carcass
>> square as you're gluing it up. Once you have done your measurements of
>> the
>> diagonals, you know the exact length that they must be, and it's quck and
>> easy to throw a brace in. No more measuring - just throw it in.
>
> This is an intriguing idea. I assume you have to assemble your
> project dry, square it up and then you can fit your diagonal brace to
> that.
>
> **************************************************************************************
>
> Correct - except don't attach the brace until you begin the glue up.
> Assemble dry, check square, determine your required brace length, and
> disassemble for glue. You should make it a practice to assemble dry
> before commiting to the glue up. Better to discover the aw-shits there
> than after the glue dries.
>

Sorry Adrian - I should have said the above differently. It's not necessary
to cut the brace to the exact length. I usually grab a piece of scrap that
is longer than the diagonal I need, and place it on the diagonal when I get
the piece squared up. Then I just mark where the edges of the pieces should
lie. I don't bother cutting to exact length. It's quick to get everything
lined back up and the brace placed properly, and attached. One diagonal
brace will hold a box (drawer, etc.) square. If you're working in stages as
you glue up, you can square up, and brace all in one step if your project is
not too big - so the glue doesn't get ahead of you.

--

-Mike-
[email protected]

On Dec 2, 11:41=A0am, "Mike Marlow" <[email protected]>
wrote:
> <[email protected]> wrote in message
>
> news:[email protected]...
> On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:
>
> > I'm not arguing. =A0I agree, making things square is important. =A0All
> > along I've made making things square a goal. =A0For the most part, I di=
d
> > the things you listed. =A0There was nothing on your list that made me
> > think, "If only I'd done that my project would be better." =A0 But even
> > if there was...that part of my project is done. =A0It's history.
>
> That's good. =A0It can only be concluded then, that what you need to do i=
s to
> perfect what you are already doing. =A0Increase your precision, take your=
time
> before going on in order to make sure each step is complete and ready for
> the next step, etc. =A0Generally, those projects that we really don't lik=
e
> when they are complete fall into two categories. =A0The first is projects=
that
> someone else dreamed up and were clearly bad ideas to start with... =A0An=
d the
> second (more seriously), are projects that we just plain old rushed our w=
ay
> through.

I'll try to go slower next time: I'll see if I can take 15 years on
my next project. :)

But more seriously, I prefer to be slow are careful and I think I
mostly do work that way. But being rushed by the glue is one problem
that I didn't see a way around until I was forced to really look for
different glue. (There is no way I ever could have put the case
together in the 15 minutes or so that PVA allows. I enlisted my
wife's help in glue application and assembly and it still took 45
minutes to get everything together.) So when I glued up the drawers I
was always rushed. By the time I would get to the point where I'm
supposed to check if the assembly is square I've already been fighting
against the glue and I'm at the end of my open time and probably tired
and drained from the stress of it all. I'd much prefer it if I could
come back tomorrow and double check the assembly for square. But I
haven't found any 48 hour glues. At least the next time I do
something like this I'll have an hour instead of 15 minutes---that
should cut down on the rush factor. (The next project is a table
which seems like it should be a lot easier to put together than 27"
long dovetails.)

>
> > If an experienced cabinet maker would burn my project and start
> > over....well, that really isn't useful information for me. If that's
> > the case, then I will end up with a project that doesn't look like it
> > was made by an experienced cabinet maker. That's OK, because I'm not
> > an experienced cabinet maker. I consider this my first real furniture
> > project. I want to know how to make the best of what I have. Being
> > told "you're screwed" is not helpful. I want to look forward, not
> > backward.
>
> Ok - then take the input you receive, and try to figure out what you are
> doing that is not really in keeping with that advice, rather than getting
> defensive.

You're right. I did respond in a defensive way and I can see this now
that it was not what I should have done.

<[email protected]> wrote in message
news:[email protected]...


> But more seriously, I prefer to be slow are careful and I think I
> mostly do work that way. But being rushed by the glue is one problem
> that I didn't see a way around until I was forced to really look for
> different glue. (There is no way I ever could have put the case
> together in the 15 minutes or so that PVA allows. I enlisted my
> wife's help in glue application and assembly and it still took 45
> minutes to get everything together.) So when I glued up the drawers I
> was always rushed.

That's the nature of glue up. One is always under the gun.

> By the time I would get to the point where I'm
> supposed to check if the assembly is square I've already been fighting
> against the glue and I'm at the end of my open time and probably tired
> and drained from the stress of it all.

Try things like cutting a stick to the predetermined length for the
diagonal, and either tacking it or screwing it in place to hold the carcass
square as you're gluing it up. Once you have done your measurements of the
diagonals, you know the exact length that they must be, and it's quck and
easy to throw a brace in. No more measuring - just throw it in.

> I'd much prefer it if I could
> come back tomorrow and double check the assembly for square. But I
> haven't found any 48 hour glues. At least the next time I do
> something like this I'll have an hour instead of 15 minutes---that
> should cut down on the rush factor. (The next project is a table
> which seems like it should be a lot easier to put together than 27"
> long dovetails.)

Actually Adrian, when I originally spoke of taking one's time I was
referring more to the other stages of a project such as taking the time to
make exact measurements, do proper stage preparation such as sanding and
cleaning, etc. Those are the areas where we usually rush and in one way or
another, they lead to either errors compounding errors or they lead to some
other form of displeasure when the finish goes on. Glue up is always a
fight against the clock. Things like jigs and braces and the like can help
you when the clock is ticking, but you will still be moving briskly.

It helps to swear a bit in this part of the project...


> You're right. I did respond in a defensive way and I can see this now
> that it was not what I should have done.

Wait - apologies and admissions are not acceptable practice here at the
wreck. When faced with things like this the appropriate type of response
includes words that begin with "f" and end with "k" (not firetruck), as well
as some sort of ad hominem attack. In general, references to either a
political party that you don't like, or a religeous affiliation you consider
to be beneath a worm's belly, make good starting points.

--

-Mike-
[email protected]

Holy Flying Firetrucks don't do it for me.....

Mike Marlow wrote:
>
> Wait - apologies and admissions are not acceptable practice here at the
> wreck. When faced with things like this the appropriate type of response
> includes words that begin with "f" and end with "k" (not firetruck), as well
> as some sort of ad hominem attack.

On Wed, 25 Nov 2009 06:18:22 -0800 (PST), "[email protected]"
<[email protected]> wrote:

>I'm installing drawers using Accuride 4034 slides, a type where part
>of the rail rides on top of, rather than nesting within, the rail that
>is secured to the case. I carefully laid out for the slides and
>installed them with the intent to get the drawers 1/4" recessed into
>the case. But each drawer came out a little different. They aren't
>recessed by the same amount, and one drawer front was twisted. (The
>top left is recessed a 1/8 and the bottom left recessed 1/4, or
>something like that.)
>
>I figured it should be easy enough to make some small adjustments to
>get the drawers to line up a little better. So for the drawer whose
>front wasn't lined up I changed the angle of its slide mounting to the
>case. But it seems that when I make this sort of change, I get
>unpredictable results. This change cause the drawer to twist in its
>opening so that the drawer front wouldn't even fit. The left side was
>too far forward so I slid its side back...but this had no effect at
>all.
>
>Does anybody have any tips on how to adjust drawer slides?


...looks to me like you're into a bunch of PITA work, pal. If your
cabinet *and* drawers are built square and true, then it's a simple
matter of a jig or two and you are in business. If you're out of
square, and it seems that indeed you are, you are in for a bunch of
fussing and fudging. Strip everything down to the carcass and start
there...

cg

<[email protected]> wrote in message
news:[email protected]...
On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:


> I'm not arguing. I agree, making things square is important. All
> along I've made making things square a goal. For the most part, I did
> the things you listed. There was nothing on your list that made me
> think, "If only I'd done that my project would be better." But even
> if there was...that part of my project is done. It's history.

That's good. It can only be concluded then, that what you need to do is to
perfect what you are already doing. Increase your precision, take your time
before going on in order to make sure each step is complete and ready for
the next step, etc. Generally, those projects that we really don't like
when they are complete fall into two categories. The first is projects that
someone else dreamed up and were clearly bad ideas to start with... And the
second (more seriously), are projects that we just plain old rushed our way
through.

> If an experienced cabinet maker would burn my project and start
> over....well, that really isn't useful information for me. If that's
> the case, then I will end up with a project that doesn't look like it
> was made by an experienced cabinet maker. That's OK, because I'm not
> an experienced cabinet maker. I consider this my first real furniture
> project. I want to know how to make the best of what I have. Being
> told "you're screwed" is not helpful. I want to look forward, not
> backward.

Ok - then take the input you receive, and try to figure out what you are
doing that is not really in keeping with that advice, rather than getting
defensive. Cabinet making in its most basic form is not all that complex or
complicated. There are a few basic principles to it all. You've had those
explained to you. After that - there's no silver bullet. Put those to work
with an increasing focus on refining them. More accurate measurements.
Finish things before moving on the the next stage. Get all those sanding
scratches out before gluing up, etc.

> I tried to measure diagonals last night. Since nobody has yet told me
> what the tolerance is I don't know how well I ought to try to
> measure. I'll admit that I'm the type to analyze things a great
> deal. But this issue of tolerance is fundamental. How square does it
> need to be? If I don't know that, how do I know if the experienced
> cabinet maker would burn the project or not? I don't even know the
> right measurement instruments to use. Calipers? Tape measure? The
> experienced cabinet maker knows the appropriate tolerance and doesn't
> really think about it.

Why should anyone have to tell you the tolerances? You were told to measure
the diagonals. It only stands to reason that you want them to be the same.
Make them the same. You can indeed over analyze this kind of thing and only
end up with analysis paralysis. If you are making a case against a very
seasoned and accomplished woodworker who offered you good advice, based on
the fact that you take things too far, then you're not going to get a lot of
sympathy here or anywhere else. Your questions above really do nothing more
than show a person who is looking to make the pot stink, more than a guy who
appreciates good advice and tries to act on it.

> In any case, taking inside measurements on the cabinet back I
> estimated the difference in the diagonals at around 0.02" in 40". (I
> did this by using a bar gauge and inserting feelers at the end to
> measure the gap.) Seems to me I can hardly ask for better.

Quite impressive. You are right - you could not ask for any better. That
is well beyond the degree of accuracy necessary for a material like wood.

> However,
> an 8" square on the cabinet front drawer cavities shows deviations of
> about 0.015" from square. This might be because I cut the dividers a
> bit too long or the joint not quite deep enough. (The case is made
> from four panels, dovetailed together with dividers inserted by
> sliding dovetails.)

Or your square is out of perfection by that amount.

> In the case of the drawers, I found the error in the diagonals to be
> nearly 1/8" in the worse case out of a diagonal length of 27". The
> other two drawers it was about 1/16" and the third was around 0.02".
> I estimated that a 1/8" error could twist the drawer front by about
> 1/16", so that would seem to explain part of my problem. I can think
> of two possible fixes: shim the drawer slides crooked or plane the
> drawer front crooked.

That 1/8" error is a lot, for a normal size drawer. Go ahead and shim it.
Next time you'll pay more attention, or slow down, or whatever is necessary,
based on what you discovered on this project. That's the way it goes.
Swingman showed you some pictures of some really nice work. What he didn't
tell you are the aw-shits in the project. Bet on it - while he may not have
the same number of, or the same type of aw-shits in his project, he could
certainly tell you about a couple that are meaningful to him.

--

-Mike-
[email protected]

On Dec 3, 9:59=A0am, "Mike Marlow" <[email protected]> wrote:
>
> > By the time I would get to the point where I'm
> > supposed to check if the assembly is square I've already been fighting
> > against the glue and I'm at the end of my open time and probably tired
> > and drained from the stress of it all.
>
> Try things like cutting a stick to the predetermined length for the
> diagonal, and either tacking it or screwing it in place to hold the carca=
ss
> square as you're gluing it up. =A0Once you have done your measurements of=
the
> diagonals, you know the exact length that they must be, and it's quck and
> easy to throw a brace in. =A0No more measuring - just throw it in.

This is an intriguing idea. I assume you have to assemble your
project dry, square it up and then you can fit your diagonal brace to
that.

"-MIKE-" <[email protected]> wrote in message
news:[email protected]...

>
> I agree you strive for dead-on. My point is I'm not loosing sleep if, at
> the end of a project, I find I'm within a tolerance that is probably
> close to the seasonal expansion/contraction of the wood I'm using. :-)
>
>

Sorry Mike - it's possible I read your post too fast and misunderstood your
intent.


--

-Mike-
[email protected]

"-MIKE-" <[email protected]> wrote in message
news:[email protected]...
> Mike Marlow wrote:
>> "-MIKE-" <[email protected]> wrote in message
>> news:[email protected]...
>>> Tom Watson wrote:
>>>> I went on Starrett's website and found that the most that they will
>>>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>>>> also say on that website that whatever tool you use to measure with
>>>> should be capable of measuring to 1/10 of what your tolerance is. So,
>>>> if their best tape is only capable of +/- 1/32", then my tolerances
>>>> can't be any tighter than 5/16", which seems a tad generous to me for
>>>> cabinet work.
>>>>
>>> If you had a wall 25 feet long (length of tape measure) that had to be
>>> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>>>
>>>
>>
>> Maybe. Depends. I can answer your question this way - when laying out
>> for long runs like say a foundation, or a deck, or the likes where this
>> kind of measurement is common, 5/16 is not close enough. Dead on is
>> close enough - recognizing that dead on has a certain error factor built
>> into it by the tolerance of the tape itself.
>>
>
> If I'm building 10 cabinets to line a 25' wall, 5/16 gives me 10/32.
> 1/32 fudge room on the width of each cabinet isn't really unreasonable, is
> it?
>

In my mind yes it does seem unreasonable. Others might well disagree and
I'll defer to experience on this. I look at it this way - error compounds
over time/distance. Therefore, strive for dead on and let the error of the
environment ( the wall) be the minimal error you have to contend with. The
more tolerance you allow, the more make up you have to introduce. Sooner or
later it shows. My rule of thumb - you can't hide it as it grows. 10/32 on
each cabinet does not seem like much, but in reality you don't have that
small amount to deal with. One cabinet will be dead on and now you're
dealing with 20/32 on the next cabinet. Or worse. Error compounds error.

As for whether 1/32 on each cabinet is reasonable - well maybe. My point is
only to strive for dead nut. You will by default, miss that expectation.
Don't assume a fudge factor going in. Assume dead on going in. When you
find that you're off by 1/32 or 1/8 or whatever, you'll find that you're
dealing with a lot less make up than if you anticipate error going in.

--

-Mike-
[email protected]

On Dec 1, 4:01=A0pm, Swingman <[email protected]> wrote:
>
> Let me put it to you gently, Bubba ... I spent some time outlining what
> an _experienced_ cabinet maker does to insure that he does NOT have the
> problems you are experiencing, now you want to ... argue with me?

I'm not arguing. I agree, making things square is important. All
along I've made making things square a goal. For the most part, I did
the things you listed. There was nothing on your list that made me
think, "If only I'd done that my project would be better." But even
if there was...that part of my project is done. It's history.

If an experienced cabinet maker would burn my project and start
over....well, that really isn't useful information for me. If that's
the case, then I will end up with a project that doesn't look like it
was made by an experienced cabinet maker. That's OK, because I'm not
an experienced cabinet maker. I consider this my first real furniture
project. I want to know how to make the best of what I have. Being
told "you're screwed" is not helpful. I want to look forward, not
backward.

I tried to measure diagonals last night. Since nobody has yet told me
what the tolerance is I don't know how well I ought to try to
measure. I'll admit that I'm the type to analyze things a great
deal. But this issue of tolerance is fundamental. How square does it
need to be? If I don't know that, how do I know if the experienced
cabinet maker would burn the project or not? I don't even know the
right measurement instruments to use. Calipers? Tape measure? The
experienced cabinet maker knows the appropriate tolerance and doesn't
really think about it.

In any case, taking inside measurements on the cabinet back I
estimated the difference in the diagonals at around 0.02" in 40". (I
did this by using a bar gauge and inserting feelers at the end to
measure the gap.) Seems to me I can hardly ask for better. However,
an 8" square on the cabinet front drawer cavities shows deviations of
about 0.015" from square. This might be because I cut the dividers a
bit too long or the joint not quite deep enough. (The case is made
from four panels, dovetailed together with dividers inserted by
sliding dovetails.)

In the case of the drawers, I found the error in the diagonals to be
nearly 1/8" in the worse case out of a diagonal length of 27". The
other two drawers it was about 1/16" and the third was around 0.02".
I estimated that a 1/8" error could twist the drawer front by about
1/16", so that would seem to explain part of my problem. I can think
of two possible fixes: shim the drawer slides crooked or plane the
drawer front crooked.

"Tom Watson" <[email protected]> wrote in message
news:[email protected]...
> Treat squareness as an absolute. You will fail anyway but in trying
> for the absolute you will come as close as possible - by definition.
>
> Treat the difference between the diagonals as an absolute. You will
> fail anyway but in trying for the absolute you will come as close as
> possible - by definition.
>
> Remember the following:
>
> Measuring Stuff Is Impossible
>

<...snip a bunch of the best stuff I've ever seen from Tom Watson...>

Thomas - that was a great post. Not the best, but "approaching" the best.
Can one ever really be the best? Is there really a dark side of the moon?

--

-Mike-
[email protected]

"Swingman" <[email protected]> wrote in message
news:[email protected]...
> [email protected] wrote:
>
> Which proves that your sudden interest in "square" has come after the
> fact, and after lack thereof has bitten you in the butt. :(
>
> This is the way you learn ... and we've all done it.
>
<...snip a bunch of other similar Swingman comments...>

Swing - you're getting awful mellow. You getting laid again?

--

-Mike-
[email protected]

On Nov 27, 12:48=A0am, Charlie Groh <[email protected]> wrote:
> On Wed, 25 Nov 2009 06:18:22 -0800 (PST), "[email protected]"
>
>
>
> <[email protected]> wrote:
> >I'm installing drawers using Accuride 4034 slides, a type where part
> >of the rail rides on top of, rather than nesting within, the rail that
> >is secured to the case. =A0I carefully laid out for the slides and
> >installed them with the intent to get the drawers 1/4" recessed into
> >the case. =A0But each drawer came out a little different. =A0They aren't
> >recessed by the same amount, and one drawer front was twisted. =A0(The
> >top left is recessed a 1/8 and the bottom left recessed 1/4, or
> >something like that.)
>
> >I figured it should be easy enough to make some small adjustments to
> >get the drawers to line up a little better. =A0So for the drawer whose
> >front wasn't lined up I changed the angle of its slide mounting to the
> >case. =A0But it seems that when I make this sort of change, I get
> >unpredictable results. =A0This change cause the drawer to twist in its
> >opening so that the drawer front wouldn't even fit. =A0The left side was
> >too far forward so I slid its side back...but this had no effect at
> >all.
>
> >Does anybody have any tips on how to adjust drawer slides?
>
> ...looks to me like you're into a bunch of PITA work, pal. =A0If your
> cabinet *and* drawers are built square and true, then it's a simple
> matter of a jig or two and you are in business. =A0If you're out of
> square, and it seems that indeed you are, you are in for a bunch of
> fussing and fudging. =A0Strip everything down to the carcass and start
> there...

I don't understand your advice. What does it mean to "strip
everything
down to the carcass and start there?" I mean, there's only the
drawers
and the carcasse. How square do things need to be and what sort of
fixes might I entertain if I determine that something isn't square
enough?

(The only idea I had so far was to plane the drawer front corner down
so
that the drawer sits evenly in the case. This would make the drawer
front thickness non-uniform.)

Lew Hodgett wrote:
> "Tom Watson" wrote:
>
>> A recent thread on using tape measures in the shop got me to
>> thinking
>> about this.
>>
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32".
> <snip>
>
> Still remember from a surveying course that steel tape "stretch" had
> to be accounted for when measuring with one.

Artillery rule for getting a round on target: "one mil of angle subtends
an arc of one meter at one thousand meters".

A 105 howitzer shell has an effective kill radius of 50 meters from
point of burst.

Making the difference between an "arc" and a "chord", in the above rule,
a moot point. ;)

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

[email protected] wrote:
> On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:
>

> I tried to measure diagonals last night.

Which proves that your sudden interest in "square" has come after the
fact, and after lack thereof has bitten you in the butt. :(

This is the way you learn ... and we've all done it.

> Since nobody has yet told me
> what the tolerance is I don't know how well I ought to try to
> measure. I'll admit that I'm the type to analyze things a great
> deal. But this issue of tolerance is fundamental. How square does it
> need to be?

As square as you can possibly get it!! :)

Forget "appropriate tolerance" and take whatever remedial action, _
during assembly_, that is necessary to make both diagonals READ THE SAME!

If your diagonals both read the same with a tape measure, or story
stick, or a piece of string, and you've taken due care in using whatever
device you used for measuring the diagonals, then it is most likely as
square as it needs to be for a woodworking project.

And the ultimate test of that is whether the components fit together as
intended.

If I don't know that, how do I know if the experienced
> cabinet maker would burn the project or not? I don't even know the
> right measurement instruments to use. Calipers? Tape measure? The
> experienced cabinet maker knows the appropriate tolerance and doesn't
> really think about it.

There is no "appropriate tolerance", because there in no finite reference.

1/16th may be fine for 48" cabinet as long as everything else is less,
but if a component of that same cabient is out 1/16 in the opposite
direction, then some remedial action may have to be taken to make the
components fit together as intended.

See first above ...


> In the case of the drawers, I found the error in the diagonals to be
> nearly 1/8" in the worse case out of a diagonal length of 27".

That much error, because diagonals are obviously NOT the same when
measured, would have been sufficient cause for me to take remedial
action on the component during assembly.

(this can generally be corrected, DURING GLUEUP, by clamping the longer
diagonal back into compliance with a clamp along that diagonal)

(... about the only place I would have accepted 1/8" would be in
squaring the foundation for a house ... 1/4" and we would be considering
some remedial carpentry.)

Once again, the goal is to make both diagonals READ THE SAME, forget
about "appropriate tolerances".

Now, and as you've already found, after the fact remedial action can be
difficult, if not impossible.

Without actually seeing your project, it is almost impossible to advise
you on a course of action. If shimming does not work, try other methods
like planning down the offending parts, within reason. If these types of
remedial actions ruin the look, or function, or intended fit, of the
project, than you obviously have to take more drastic measures, like
redoing the offending parts causing the problem ... once again, consider
it a learning experience.

Can you take the offending components apart? This can sometimes be
accomplished depending upon the glue used, so that you re-glue back to
square ... a heat gun or hair dryer on the joint often works, and, if
you have to redo the component as a last resort, it is always worth a try.

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

-MIKE- wrote:
> Tom Watson wrote:
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>> also say on that website that whatever tool you use to measure with
>> should be capable of measuring to 1/10 of what your tolerance is. So,
>> if their best tape is only capable of +/- 1/32", then my tolerances
>> can't be any tighter than 5/16", which seems a tad generous to me for
>> cabinet work.
>>
>
> If you had a wall 25 feet long (length of tape measure) that had to be
> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?

Different kettle of fish than "square" .. but in your example, that
depends upon the width of the _available_ trim piece!! :)

IOW, your trim piece that hides the gap better be wider than 5/16" ...

<g>


--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

Tom Watson wrote:
>
> Zeno may have been the first framing carpenter, although I am not
> entirely sure about that - nor anything else, it seems.
>
I know not but that I know not.
mahaloklos,
"soc"

In article <[email protected]>,
Mike Marlow <[email protected]> wrote:
>
>"jo4hn" <[email protected]> wrote in message
>news:[email protected]...
>> Tom Watson wrote:
>>>
>>> Zeno may have been the first framing carpenter, although I am not
>>> entirely sure about that - nor anything else, it seems.
>>>
>> I know not but that I know not.
>> mahaloklos,
>> "soc"
>
>I think, therefore I am...

"I think I am, therefore, I am, I think!"

In article <[email protected]>,
Mike Marlow <[email protected]> wrote:
>
> .... Is there really a dark side of the moon?


No, the moon doesn't have a dark side -- now, it's "evil twin", on the other
hand .......

[email protected] wrote:


> and the carcasse. How square do things need to be and what sort of
> fixes might I entertain if I determine that something isn't square
> enough?


Basic premise: Cabinet "components" (generally speaking the casework,
the drawers and the doors) are all made up of individual "parts".

Perfectly square is the goal for all these components, and often "close
enough" will work if you can shim "components" into square during final
assembly of the cabinet. However, once you start shimming you're wasting
time that could better be spent doing something else, and there is NEVER
a guarantee of satisfactory results.

'Measuring diagonals' will tell you whether the four sides of a
"component" (drawer, door, drawer front, or the casework) are square to
each other; using an 'accurate square' of any kind will also tell you if
the corners are square (keep in mind that without proper preparation,
the stock between square corners could still be twisted, bent, warped,
or vary in thickness, and still cause problems).

The first step in solving your problem is to answer the questions:

Did you pay particular attention to the elements of "square" when you
built both your cabinet and drawers - IOW, ALL individual component
"parts" are indeed the specified project thickness, width, and length?

Did you use properly milled, straight, stock, of the equal thickness,
for each part?

Did you "batch cut" these parts before assembly? (more below)

Did you take steps during component assembly and glue-up to insure a
square results? (measuring diagonals, proper clamping techniques to
preclude warping by too much pressure, etc?)

The pursuit of "square" is the holy grail of cabinet making ... if you
did none of the above, you may well need to start over again as you can
spend hours attempting to shim the drawer slides and non-square
components, in all planes, and still not have a satisfactory end result.

One simple method/practice which will take you a long way to insuring
that your basic components (drawers, doors, casework) end up square is
to "batch cut" ALL "parts" of like dimension for these components.

"Batch cutting" parts is the practice of using the EXACT SAME machine
setup to cut ALL like project parts BEFORE changing machine settings
(move the table saw fence, move the planer table, etc).

AAMOF, this practice can't be stressed enough and will take you a long
way toward alleviating the problem you are currently experiencing.

Examples of this:

Cut ALL your "parts" (drawer sides, rails and stiles, casework sides,
etc.) of like WIDTH in the ENTIRE project, BEFORE you move your table
saw fence from that WIDTH setting.

Cut ALL your "parts" of like LENGTH in the ENTIRE project, BEFORE you
move your table saw fence from that LENGTH setting

Thickness, to project specs, ALL stock with the SAME final setting on
your planer, BEFORE you change that setting.

Etc, ad infinitum ...

This one simple practice (which does require some organization, planning
and thought) will insure that ALL project components parts, that have
identical dimensions, in thickness, width, and length, are indeed
identical and have not been subjected to errors introduced when moving
fences, machine tables/settings, etc..

(There are other things, like when using face frame cabinets, build your
face frames first, taking the time and necessary steps to insure they
are square, then assemble your cabinet sides on top of the already
"known square" face frames).

Paying particular attention to "square" with steps like the above when
building the three basic components of a "cabinet" (the casework, the
drawers and the doors) will save countless hours of trying to fit
non-square components during final assembly.

Multiply that by the number of cabinets in the average shop built
kitchen and the importance of pursuing the holy grail of "square"
becomes paramount.

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

[email protected] wrote:


> and the carcasse. How square do things need to be and what sort of
> fixes might I entertain if I determine that something isn't square
> enough?


Basic premise: Cabinet "components" (generally speaking the casework,
the drawers and the doors) are all made up of individual "parts".

Perfectly square is the goal for all these components, and often "close
enough" will work if you can shim "components" into square during final
assembly of the cabinet. However, once you start shimming you're wasting
time that could better be spent doing something else, and there is NEVER
a guarantee of satisfactory results.

'Measuring diagonals' will tell you whether the four sides of a
"component" (drawer, door, drawer front, or the casework) are square to
each other; using an 'accurate square' of any kind will also tell you if
the corners are square (keep in mind that without proper preparation,
the stock between square corners could still be twisted, bent, warped,
or vary in thickness, and still cause problems).

The first step in solving your problem is to answer the questions:

Did you pay particular attention to the elements of "square" when you
built both your cabinet and drawers - IOW, ALL individual component
"parts" are indeed the specified project thickness, width, and length?

Did you use properly milled, straight, stock, of the equal thickness,
for each part?

Did you "batch cut" these parts before assembly? (more below)

Did you take steps during component assembly and glue-up to insure a
square results? (measuring diagonals, proper clamping techniques to
preclude warping by too much pressure, etc?)

The pursuit of "square" is the holy grail of cabinet making ... if you
did none of the above, you may well need to start over again as you can
spend hours attempting to shim the drawer slides and non-square
components, in all planes, and still not have a satisfactory end result.

One simple method/practice which will take you a long way to insuring
that your basic components (drawers, doors, casework) end up square is
to "batch cut" ALL "parts" of like dimension for these components.

"Batch cutting" parts is the practice of using the EXACT SAME machine
setup to cut ALL like project parts BEFORE changing machine settings
(move the table saw fence, move the planer table, etc).

AAMOF, this practice can't be stressed enough and will take you a long
way toward alleviating the problem you are currently experiencing.

Examples of this:

Cut ALL your "parts" (drawer sides, rails and stiles, casework sides,
etc.) of like WIDTH in the ENTIRE project, BEFORE you move your table
saw fence from that WIDTH setting.

Cut ALL your "parts" of like LENGTH in the ENTIRE project, BEFORE you
move your table saw fence from that LENGTH setting

Thickness, to project specs, ALL stock with the SAME final setting on
your planer, BEFORE you change that setting.

Etc, ad infinitum ...

This one simple practice (which does require some organization, planning
and thought) will insure that ALL project components parts, that have
identical dimensions, in thickness, width, and length, are indeed
identical and have not been subjected to errors introduced when moving
fences, machine tables/settings, etc..

(There are other things, like when using face frame cabinets, build your
face frames first, taking the time and necessary steps to insure they
are square, then assemble your cabinet sides on top of the already
"known square" face frames).

Paying particular attention to "square" with steps like the above when
building the three basic components of a "cabinet" (the casework, the
drawers and the doors) will save countless hours of trying to fit
non-square components during final assembly.

Multiply that by the number of cabinets in the average shop built
kitchen and the importance of pursuing the holy grail of "square"
becomes paramount.

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

On Dec 3, 9:53=A0am, "charlie" <[email protected]> wrote:
> <[email protected]> wrote in message
>
> news:[email protected]...
> On Dec 2, 11:41 am, "Mike Marlow" <[email protected]>
> wrote:
>
>
>
> > <[email protected]> wrote in message
>
> >news:[email protected]...
> > On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:
>
> > > I'm not arguing. I agree, making things square is important. All
> > > along I've made making things square a goal. For the most part, I did
> > > the things you listed. There was nothing on your list that made me
> > > think, "If only I'd done that my project would be better." But even
> > > if there was...that part of my project is done. It's history.
>
> > That's good. It can only be concluded then, that what you need to do is=
to
> > perfect what you are already doing. Increase your precision, take your
> > time
> > before going on in order to make sure each step is complete and ready f=
or
> > the next step, etc. Generally, those projects that we really don't like
> > when they are complete fall into two categories. The first is projects
> > that
> > someone else dreamed up and were clearly bad ideas to start with... And
> > the
> > second (more seriously), are projects that we just plain old rushed our
> > way
> > through.
>
> I'll try to go slower next time: =A0I'll see if I can take 15 years on
> my next project. =A0:)
>
> But more seriously, I prefer to be slow are careful and I think I
> mostly do work that way. =A0But being rushed by the glue is one problem
> that I didn't see a way around until I was forced to really look for
> different glue. =A0 =A0(There is no way I ever could have put the case
> together in the 15 minutes or so that PVA allows. =A0I enlisted my
> wife's help in glue application and assembly and it still took 45
> minutes to get everything together.) =A0So when I glued up the drawers I
> was always rushed. =A0By the time I would get to the point where I'm
> supposed to check if the assembly is square I've already been fighting
> against the glue and I'm at the end of my open time and probably tired
> and drained from the stress of it all. =A0I'd much prefer it if I could
> come back tomorrow and double check the assembly for square. =A0But I
> haven't found any 48 hour glues. =A0At least the next time I do
> something like this I'll have an hour instead of 15 minutes---that
> should cut down on the rush factor. =A0(The next project is a table
> which seems like it should be a lot easier to put together than 27"
> long dovetails.)
>
> =3D=3D=3D=3D=3D=3D=3D=3D
>
> try epoxies. the open time can vary from seconds to weeks for different
> products. the one i use for glass takes approx 1 week to set up.
>
> http://www.hisglassworks.com/cart/cart.php?m=3Dproduct_detail&p=3D94
>

Do you have any idea what the working time of this stuff is? They say
it takes a week to cure. I have an epoxy that takes 2 weeks to cure.
But the working time is only 80 minutes.

> of course, it would be hard to undo it if you did make a mistake.

I haven't had much luck with undoing "reversible" glues. I
experimented with hide glue. Turns out my shop is too cold for it,
and even though I warmed the Old Brown Glue to 130 degrees I only got
about 5 minutes of working time. But when I tried to reverse the glue
because of large unsightly glue lines I ended up giving up and sawing
the pieces apart.

The fish glue I used is supposedly reversible, like hide glue. I
haven't tried to reverse it. But they claim a 60-90 minute working
time which is as long as the epoxies I'm familiar with.

"-MIKE-" <[email protected]> wrote in message
news:[email protected]...
> Tom Watson wrote:
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>> also say on that website that whatever tool you use to measure with
>> should be capable of measuring to 1/10 of what your tolerance is. So,
>> if their best tape is only capable of +/- 1/32", then my tolerances
>> can't be any tighter than 5/16", which seems a tad generous to me for
>> cabinet work.
>>
>
> If you had a wall 25 feet long (length of tape measure) that had to be
> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>
>

Maybe. Depends. I can answer your question this way - when laying out for
long runs like say a foundation, or a deck, or the likes where this kind of
measurement is common, 5/16 is not close enough. Dead on is close enough -
recognizing that dead on has a certain error factor built into it by the
tolerance of the tape itself.

--

-Mike-
[email protected]

"jo4hn" <[email protected]> wrote in message
news:[email protected]...
> Tom Watson wrote:
>>
>> Zeno may have been the first framing carpenter, although I am not
>> entirely sure about that - nor anything else, it seems.
>>
> I know not but that I know not.
> mahaloklos,
> "soc"

I think, therefore I am...

--

-Mike-
[email protected]

<[email protected]> wrote in message
news:[email protected]...
On Dec 3, 9:59 am, "Mike Marlow" <[email protected]> wrote:
>
> > By the time I would get to the point where I'm
> > supposed to check if the assembly is square I've already been fighting
> > against the glue and I'm at the end of my open time and probably tired
> > and drained from the stress of it all.
>
> Try things like cutting a stick to the predetermined length for the
> diagonal, and either tacking it or screwing it in place to hold the
> carcass
> square as you're gluing it up. Once you have done your measurements of the
> diagonals, you know the exact length that they must be, and it's quck and
> easy to throw a brace in. No more measuring - just throw it in.

This is an intriguing idea. I assume you have to assemble your
project dry, square it up and then you can fit your diagonal brace to
that.

**************************************************************************************

Correct - except don't attach the brace until you begin the glue up.
Assemble dry, check square, determine your required brace length, and
disassemble for glue. You should make it a practice to assemble dry before
commiting to the glue up. Better to discover the aw-shits there than after
the glue dries.

--

-Mike-
[email protected]

On Dec 1, 11:06=A0am, Swingman <[email protected]> wrote:
> [email protected] wrote:
>
> =A0> and the carcasse. =A0 =A0How square do things need to be and what so=
rt of
> =A0> fixes might I entertain if I determine that something isn't square
> =A0> enough?
>
> Basic premise: Cabinet "components" (generally speaking the casework,
> the drawers and the doors) are all made up of individual "parts".
>
> Perfectly square is the goal for all these components,

There is no such thing. There is always a tolerance. It's not
always
clear to me what the tolerance ought to be.

> and often "close
> enough" will work

Of course "close enough" will work. You going to tell me you cut
accurate to the nanometer? And your corners are square to a
millionth of a degree? I doubt it. There's *always* a tolerance,
which is to say, there's some measure of "close enough".

> if you can shim "components" into square during final
> assembly of the cabinet. However, once you start shimming you're wasting
> time that could better be spent doing something else, and there is NEVER
> a guarantee of satisfactory results.

Well, if the alternative is to start over it's hard to see how
shimming
is "wasting time". If I have to spend, I dunno, 500-1000 hours
making
the project again vs. 1 hour shimming? I suppose the second time
it
wouldn't take so long, since many of those hours were learning hours,
but I'd learn less the second time around, and it wouldn't be
interesting
to make exactly the same project again.

> 'Measuring diagonals' will tell you whether the four sides of a
> "component" (drawer, door, drawer front, or the casework) are square to
> each other; using an 'accurate square' of any kind will also tell you if
> the corners are square (keep in mind that without proper preparation,
> the stock between square corners could still be twisted, bent, warped,
> or vary in thickness, and still cause problems).

Of course I can measure diagonals, though this won't find twist.
I think the case is reasonably close. I don't recall how square the
drawers are. It's been a couple years since I finished them. I'm
planning to take a look tonight. But how close is close enough?

>
> The first step in solving your problem is to answer the questions:
>
> Did you pay particular attention to the elements of "square" when you
> built both your cabinet and drawers - IOW, ALL individual component
> "parts" are indeed the specified project thickness, width, and length?

Of course I made my best effort to make all the parts square. But
wood isn't plastic. Parts may not have stayed as straight and square
as
I made them. They may have warped a bit when being planed after
jointing, or while waiting to be joined together.

> Did you use properly milled, straight, stock, of the equal thickness,
> for each part?

Of course I made my best effort to do this.

>
> Did you "batch cut" these parts before assembly? (more below)
>
> Did you take steps during component assembly and glue-up to insure a
> square results? (measuring diagonals, proper clamping techniques to
> preclude warping by too much pressure, etc?)

Of course I tried to keep everything square during glue up, though
some of the glue ups were a little panicked...

When I glued panels together I used a setup I read about in this
very group involving cauls clamped over the panels to keep them
flat. All the other joints in the piece are dovetails.

> The pursuit of "square" is the holy grail of cabinet making ... if you
> did none of the above, you may well need to start over again as you can
> spend hours attempting to shim the drawer slides and non-square
> components, in all planes, and still not have a satisfactory end result.

Well, starting over really isn't going to happen. I've been working
on this
project for about 8 years. (I did other things too...) It's got $750
of wood
in it, or thereabouts. I'm going to get the best end result I can get
with
what I have and move on to a new project.

> One simple method/practice which will take you a long way to insuring
> that your basic components (drawers, doors, casework) end up square is
> to "batch cut" ALL "parts" of like dimension for these components.
>
> "Batch cutting" parts is the practice of using the EXACT SAME machine
> setup to cut ALL like project parts BEFORE changing machine settings
> (move the table saw fence, move the planer table, etc).

Note that I do not have a table saw. I did attempt to cut some parts
together by stacking them (e.g. matching parts from a given drawer)
so ensure that they would come out identical. (Though this doesn't
guarantee square.)

>
> Thickness, to project specs, ALL stock with the SAME final setting on
> your planer, BEFORE you change that setting.

I thicknessed parts at the same time for each drawer, and for the
case.
Though really, if the parts had different thicknesses, the piece
would still be square, as long as the parts were straight. It would
have
made the joinery more of a pain. In fact, the top of the case is
1/4"
thicker than the sides. (That was intentional.)

"Lew Hodgett" <[email protected]> wrote in message
news:[email protected]...
> "Tom Watson" wrote:
>
>> A recent thread on using tape measures in the shop got me to thinking
>> about this.
>>
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32".
> <snip>
>
> Still remember from a surveying course that steel tape "stretch" had to be
> accounted for when measuring with one.
>
> Long since forgot how to do it, just remember it could be a problem.
>

They stretch or shrink with temperature. Like you - great minds thinking
alike and all that, I can't remember the compensating formulas either.
Doesn't matter - we've both proven ourselves to be demi-gods with this
little bit of knowledge. Let the young bulls challenge us, we'll just sit
up here on this hill and let the cows come up to us.

--

-Mike-
[email protected]

> I have a Starrett dial caliper that will measure to 1/1000" - now that
> will let me have tolerances of about 1/100", which is heading in the
> right direction but when I think about it, a piece of newsprint is
> about 4/1000", or 1/250" and I know that my joints are tight enough,
> when they are cut properly, that I can't fit a piece of newspaper into
> them.
>
> And yet, that can't be possible because the best measuring instrument
> that I have in my shop will only allow me to have tolerances of
> 1/100".

If you mark one part of a joint from another, what's the error
associated with that procedure? Isn't this what determines the
accuracy of a joint rather than the ability to measure anything?
(I'm assuming here that you cut the joint by hand.)

I started wondering about tolerances when I started trying to face
joint wood. If I aimed for "perfect", meaning that I couldn't detect
any deviation from my Starrett, then I'd still be jointing. And I
concluded that seeking that level of accuracy doesn't make sense
because of the changeable nature of wood. (I plane the other side and
it bends 0.1" anyway.) So if perfect isn't the goal...then what is?
I picked 0.004", but I don't know if that's the right answer.

-MIKE- wrote:
> dpb wrote:
>> They're quoting an absolute accuracy--you can do much better than that
>> w/ repeatability using the same tape --
>
> It only took me once, to learn to use the same tape throughout a project.

BINGO!!!

Also, a woodworkers delight is finding that his project tape measure
coincides precisely with the tape measure on his table saw fence!!

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

[email protected] wrote:
...
> I tried to measure diagonals last night. Since nobody has yet told me
> what the tolerance is ...

I generally shoot for 1/64"; that'll normally w/ reasonable care get me
at least 1/32" for cabinets, etc., of that size. OTOH, if you're making
miniatures, that could be a terribly large error... :)

...

> In the case of the drawers, I found the error in the diagonals to be
> nearly 1/8" in the worse case out of a diagonal length of 27". The
> other two drawers it was about 1/16" and the third was around 0.02".
> I estimated that a 1/8" error could twist the drawer front by about
> 1/16", so that would seem to explain part of my problem. I can think
> of two possible fixes: shim the drawer slides crooked or plane the
> drawer front crooked.

That's definitely far too much. I'd suggest simply rebuilding the two
worst of the drawers as a far more satisfactory solution in the long
run. To install them cockeyed will be a hack that you'll not be
satisfied with for the long run imo...

--

In news:[email protected],
Tom Watson <[email protected]> dropped this bit of wisdom:
<SNIP>=20
> Zeno may have been the first framing carpenter, although I am not
> entirely sure about that - nor anything else, it seems.
>=20
And now you know why one should never countenance Philosophy. :-)

P D Q

Tom Watson wrote:
> I went on Starrett's website and found that the most that they will
> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
> also say on that website that whatever tool you use to measure with
> should be capable of measuring to 1/10 of what your tolerance is. So,
> if their best tape is only capable of +/- 1/32", then my tolerances
> can't be any tighter than 5/16", which seems a tad generous to me for
> cabinet work.
>

If you had a wall 25 feet long (length of tape measure) that had to be
filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

Mike Marlow wrote:
> "-MIKE-" <[email protected]> wrote in message
> news:[email protected]...
>> Tom Watson wrote:
>>> I went on Starrett's website and found that the most that they will
>>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>>> also say on that website that whatever tool you use to measure with
>>> should be capable of measuring to 1/10 of what your tolerance is. So,
>>> if their best tape is only capable of +/- 1/32", then my tolerances
>>> can't be any tighter than 5/16", which seems a tad generous to me for
>>> cabinet work.
>>>
>> If you had a wall 25 feet long (length of tape measure) that had to be
>> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>>
>>
>
> Maybe. Depends. I can answer your question this way - when laying out for
> long runs like say a foundation, or a deck, or the likes where this kind of
> measurement is common, 5/16 is not close enough. Dead on is close enough -
> recognizing that dead on has a certain error factor built into it by the
> tolerance of the tape itself.
>

If I'm building 10 cabinets to line a 25' wall, 5/16 gives me 10/32.
1/32 fudge room on the width of each cabinet isn't really unreasonable,
is it?


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

Swingman wrote:
> -MIKE- wrote:
>> Tom Watson wrote:
>>> I went on Starrett's website and found that the most that they will
>>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>>> also say on that website that whatever tool you use to measure with
>>> should be capable of measuring to 1/10 of what your tolerance is. So,
>>> if their best tape is only capable of +/- 1/32", then my tolerances
>>> can't be any tighter than 5/16", which seems a tad generous to me for
>>> cabinet work.
>>>
>>
>> If you had a wall 25 feet long (length of tape measure) that had to be
>> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>
> Different kettle of fish than "square" .. but in your example, that
> depends upon the width of the _available_ trim piece!! :)
>
> IOW, your trim piece that hides the gap better be wider than 5/16" ...
>
> <g>
>

I'm guessing one of the end walls will be out of plumb by double that,
anyway. :-)


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

-MIKE- wrote:
> Tom Watson wrote:
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>> also say on that website that whatever tool you use to measure with
>> should be capable of measuring to 1/10 of what your tolerance is. So,
>> if their best tape is only capable of +/- 1/32", then my tolerances
>> can't be any tighter than 5/16", which seems a tad generous to me for
>> cabinet work.
...

They're quoting an absolute accuracy--you can do much better than that
w/ repeatability using the same tape and good technique where you don't
care if the measurement is off relative to the reference NIST but simply
need to reproduce multiple pieces or partition the total length, etc.,
etc., but all w/ the same measuring device most absolute error cancels.

--

Mike Marlow wrote:
>> If I'm building 10 cabinets to line a 25' wall, 5/16 gives me 10/32.
>> 1/32 fudge room on the width of each cabinet isn't really unreasonable, is
>> it?
>>
>
> In my mind yes it does seem unreasonable. Others might well disagree and
> I'll defer to experience on this. I look at it this way - error compounds
> over time/distance. Therefore, strive for dead on and let the error of the
> environment ( the wall) be the minimal error you have to contend with. The
> more tolerance you allow, the more make up you have to introduce. Sooner or
> later it shows. My rule of thumb - you can't hide it as it grows. 10/32 on
> each cabinet does not seem like much, but in reality you don't have that
> small amount to deal with. One cabinet will be dead on and now you're
> dealing with 20/32 on the next cabinet. Or worse. Error compounds error.
>

I'm sorry if I said anything that made you think I believe +/-1/4" on a
30" cabinet is acceptable.


> As for whether 1/32 on each cabinet is reasonable - well maybe. My point is
> only to strive for dead nut. You will by default, miss that expectation.
> Don't assume a fudge factor going in. Assume dead on going in. When you
> find that you're off by 1/32 or 1/8 or whatever, you'll find that you're
> dealing with a lot less make up than if you anticipate error going in.
>

I agree you strive for dead-on. My point is I'm not loosing sleep if, at
the end of a project, I find I'm within a tolerance that is probably
close to the seasonal expansion/contraction of the wood I'm using. :-)


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

dpb wrote:
> They're quoting an absolute accuracy--you can do much better than that
> w/ repeatability using the same tape
> --

It only took me once, to learn to use the same tape throughout a project.


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

Mike Marlow wrote:
> "-MIKE-" <[email protected]> wrote in message
> news:[email protected]...
>
>> I agree you strive for dead-on. My point is I'm not loosing sleep if, at
>> the end of a project, I find I'm within a tolerance that is probably
>> close to the seasonal expansion/contraction of the wood I'm using. :-)
>>
>>
>
> Sorry Mike - it's possible I read your post too fast and misunderstood your
> intent.
>

That's unfortunate.
You see, I strive for dead on accurate when reading usenet posts.
You obviously have much looser tolerances.

:-p

<running for cover>


--

-MIKE-

"Playing is not something I do at night, it's my function in life"
--Elvin Jones (1927-2004)
--
http://mikedrums.com
[email protected]
---remove "DOT" ^^^^ to reply

-MIKE- wrote:
> Tom Watson wrote:
>> I went on Starrett's website and found that the most that they will
>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>> also say on that website that whatever tool you use to measure with
>> should be capable of measuring to 1/10 of what your tolerance is.
>> So, if their best tape is only capable of +/- 1/32", then my
>> tolerances can't be any tighter than 5/16", which seems a tad
>> generous to me for cabinet work.
>>
>
> If you had a wall 25 feet long (length of tape measure) that had to be
> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?

Depends on which way the error goes. If the cabinets end up 5/16 too short
you can fudge it any number of ways. If they end up 5/16 too long then
you've got a royal pain in the butt.

<[email protected]> wrote in message
news:[email protected]...
On Dec 2, 11:41 am, "Mike Marlow" <[email protected]>
wrote:
> <[email protected]> wrote in message
>
> news:[email protected]...
> On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:
>
> > I'm not arguing. I agree, making things square is important. All
> > along I've made making things square a goal. For the most part, I did
> > the things you listed. There was nothing on your list that made me
> > think, "If only I'd done that my project would be better." But even
> > if there was...that part of my project is done. It's history.
>
> That's good. It can only be concluded then, that what you need to do is to
> perfect what you are already doing. Increase your precision, take your
> time
> before going on in order to make sure each step is complete and ready for
> the next step, etc. Generally, those projects that we really don't like
> when they are complete fall into two categories. The first is projects
> that
> someone else dreamed up and were clearly bad ideas to start with... And
> the
> second (more seriously), are projects that we just plain old rushed our
> way
> through.

I'll try to go slower next time: I'll see if I can take 15 years on
my next project. :)

But more seriously, I prefer to be slow are careful and I think I
mostly do work that way. But being rushed by the glue is one problem
that I didn't see a way around until I was forced to really look for
different glue. (There is no way I ever could have put the case
together in the 15 minutes or so that PVA allows. I enlisted my
wife's help in glue application and assembly and it still took 45
minutes to get everything together.) So when I glued up the drawers I
was always rushed. By the time I would get to the point where I'm
supposed to check if the assembly is square I've already been fighting
against the glue and I'm at the end of my open time and probably tired
and drained from the stress of it all. I'd much prefer it if I could
come back tomorrow and double check the assembly for square. But I
haven't found any 48 hour glues. At least the next time I do
something like this I'll have an hour instead of 15 minutes---that
should cut down on the rush factor. (The next project is a table
which seems like it should be a lot easier to put together than 27"
long dovetails.)

========

try epoxies. the open time can vary from seconds to weeks for different
products. the one i use for glass takes approx 1 week to set up.

http://www.hisglassworks.com/cart/cart.php?m=product_detail&p=94

of course, it would be hard to undo it if you did make a mistake.

regards,
charlie
cave creek, az

<[email protected]> wrote in message
news:ab66adf5-87bd-48d9-a1c5-268ccb823e30@m25g2000yqc.googlegroups.com...
On Dec 3, 9:53 am, "charlie" <[email protected]> wrote:
> <[email protected]> wrote in message
>
> news:[email protected]...
> On Dec 2, 11:41 am, "Mike Marlow" <[email protected]>
> wrote:
>
>
>
> > <[email protected]> wrote in message
>
> >news:[email protected]...
> > On Dec 1, 4:01 pm, Swingman <[email protected]> wrote:
>
> > > I'm not arguing. I agree, making things square is important. All
> > > along I've made making things square a goal. For the most part, I did
> > > the things you listed. There was nothing on your list that made me
> > > think, "If only I'd done that my project would be better." But even
> > > if there was...that part of my project is done. It's history.
>
> > That's good. It can only be concluded then, that what you need to do is
> > to
> > perfect what you are already doing. Increase your precision, take your
> > time
> > before going on in order to make sure each step is complete and ready
> > for
> > the next step, etc. Generally, those projects that we really don't like
> > when they are complete fall into two categories. The first is projects
> > that
> > someone else dreamed up and were clearly bad ideas to start with... And
> > the
> > second (more seriously), are projects that we just plain old rushed our
> > way
> > through.
>
> I'll try to go slower next time: I'll see if I can take 15 years on
> my next project. :)
>
> But more seriously, I prefer to be slow are careful and I think I
> mostly do work that way. But being rushed by the glue is one problem
> that I didn't see a way around until I was forced to really look for
> different glue. (There is no way I ever could have put the case
> together in the 15 minutes or so that PVA allows. I enlisted my
> wife's help in glue application and assembly and it still took 45
> minutes to get everything together.) So when I glued up the drawers I
> was always rushed. By the time I would get to the point where I'm
> supposed to check if the assembly is square I've already been fighting
> against the glue and I'm at the end of my open time and probably tired
> and drained from the stress of it all. I'd much prefer it if I could
> come back tomorrow and double check the assembly for square. But I
> haven't found any 48 hour glues. At least the next time I do
> something like this I'll have an hour instead of 15 minutes---that
> should cut down on the rush factor. (The next project is a table
> which seems like it should be a lot easier to put together than 27"
> long dovetails.)
>
> ========
>
> try epoxies. the open time can vary from seconds to weeks for different
> products. the one i use for glass takes approx 1 week to set up.
>
> http://www.hisglassworks.com/cart/cart.php?m=product_detail&p=94
>

Do you have any idea what the working time of this stuff is? They say
it takes a week to cure. I have an epoxy that takes 2 weeks to cure.
But the working time is only 80 minutes.

=============

i've heard of people moving things glued using hxtal for a couple of days at
least. you have to clamp or have it sit still for almost the entire week, as
it can creep until it sets. here's what the faq on it says

"HXTAL sets slowly - at 75°F, it requires about one week to achieve most of
the final bond strength, (see Physical Properties Sheet). However,
ordinarily HXTAL is set sufficiently after 24 hours to hold the two parts
together as long as no stress is applied to the glue joint."

regards,
charlie
http://glassartists.org/ChaniArts

Swingman wrote:
> [email protected] wrote:
>
>> mostly do work that way. But being rushed by the glue is one problem
>> that I didn't see a way around until I was forced to really look for
>> different glue. (There is no way I ever could have put the case
>> together in the 15 minutes or so that PVA allows. I enlisted my
>> wife's help in glue application and assembly and it still took 45
>> minutes to get everything together.) So when I glued up the drawers I
>> was always rushed. By the time I would get to the point where I'm
>> supposed to check if the assembly is square I've already been fighting
>> against the glue and I'm at the end of my open time and probably tired
>> and drained from the stress of it all. I'd much prefer it if I could
>> come back tomorrow and double check the assembly for square. But I
>> haven't found any 48 hour glues. At least the next time I do
>> something like this I'll have an hour instead of 15 minutes---that
>> should cut down on the rush factor. (The next project is a table
>> which seems like it should be a lot easier to put together than 27"
>> long dovetails.)
>
> Simple solution: Don't glue-up the whole piece at once!! Break it into
> two, or more, glue-ups! Rare is the component that you can't break down
> into multiple glue-ups if need be.
>
> CAVEAT: when doing partial glue-ups, it's a good practice (actually
> imperative) to go ahead and clamp up the full assembly, both glued
> joint, and unglued joints, and check for square.
>
> Use some common sense with multiple glue-ups, think it through so that
> you don't glue yourself into a corner; practice the steps first before
> each glue-up, with clamps at ready; and your project will thank you ...
>

I'm coming in real late on this one, but I think it's important to
stress what Swing is saying here, but from a different angle.

My first glueups were as stressed as what Adrian is describing. I
suspect most experience that "time is running out" syndrome, and the
added pressure almost ensures that something will be forgotten.

Because of that, anything that could go wrong needs to be anticipated
long before the glue lays down. Although it's a royal pain in the ass, I
do a lot of dry fits. That's probably to my detriment, as each dry fit
is going to loosen the joints slightly but I'll take that hit if I can
do one dry fit and have it run flawlessly. Including measuring the
diagonals.

Then, when I do my final wet assembly, that one gawdammed thing that
didn't show up in dry fits rears its ugly head. Happens each time, and
if it didn't I'd think something was way off. However, all of the other
smaller problems are dealt with and I can still do it inside the glue's
open time.

The key is to not be surprised by much. That's why the dry fits work for me.

Tanus

On Wed, 25 Nov 2009 06:18:22 -0800 (PST), "[email protected]"
<[email protected]> wrote:

>I'm installing drawers using Accuride 4034 slides, a type where part
>of the rail rides on top of, rather than nesting within, the rail that
>is secured to the case. I carefully laid out for the slides and
>installed them with the intent to get the drawers 1/4" recessed into
>the case. But each drawer came out a little different. They aren't
>recessed by the same amount, and one drawer front was twisted. (The
>top left is recessed a 1/8 and the bottom left recessed 1/4, or
>something like that.)
>
>I figured it should be easy enough to make some small adjustments to
>get the drawers to line up a little better. So for the drawer whose
>front wasn't lined up I changed the angle of its slide mounting to the
>case. But it seems that when I make this sort of change, I get
>unpredictable results. This change cause the drawer to twist in its
>opening so that the drawer front wouldn't even fit. The left side was
>too far forward so I slid its side back...but this had no effect at
>all.
>
>Does anybody have any tips on how to adjust drawer slides?


I'm a little late on this but...... look at the slides very carfully.
there is a tab on the end of the slides at the back of the cabinet.
look at both side by side. Is one bent differently? I ran into this
not too long ago. everything was square and should have fit but one
drawer stuck out on the left and one stuck out on the right. I screwed
around with this crap for 2 days and finally took the slides out and
compared them. sure enough the tab at the back on one side of each set
of slides was bent! bent the offending tabs, put em back in and all
was right with the world. was it because of made in china or damaged
in shipping? dunno but it was a simple thing just not obvious.

skeez

J. Clarke wrote:
> -MIKE- wrote:

>>>
>> If you had a wall 25 feet long (length of tape measure) that had to be
>> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>
> Depends on which way the error goes. If the cabinets end up 5/16 too short
> you can fudge it any number of ways. If they end up 5/16 too long then
> you've got a royal pain in the butt.

This has only happened to me once, when the framers moved a kitchen sink
window rough opening in the stud wall 4" closer to the only
perpendicular wall I had too worry about; then, to compound the problem,
mounted the window in the RO closer to one side than the other.

In order to center the sink base cabinet on that kitchen window
(something that is ABSOLUTELY required in most installations where a
kitchen window is involved over a sink), I ended up having to completely
rebuild the last of four base cabinet in the run to fit the now too
small space.

... amazing how little time a one off, angled, base cabinet takes to
build when it's urgent.

Three years later I still have that right hand, angled base cabinet in
storage, and still looking for a place to install it ... the houses I've
been since have been mirror images in layout. :(

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

Treat squareness as an absolute. You will fail anyway but in trying
for the absolute you will come as close as possible - by definition.

Treat the difference between the diagonals as an absolute. You will
fail anyway but in trying for the absolute you will come as close as
possible - by definition.

Remember the following:

Measuring Stuff Is Impossible

A recent thread on using tape measures in the shop got me to thinking
about this.

I went on Starrett's website and found that the most that they will
write up a Certificate of Accuracy for on a tape is +/- 1/32". They
also say on that website that whatever tool you use to measure with
should be capable of measuring to 1/10 of what your tolerance is. So,
if their best tape is only capable of +/- 1/32", then my tolerances
can't be any tighter than 5/16", which seems a tad generous to me for
cabinet work.

I have some Starrett and Rabone-Chesterman metal rules that will
measure to 1/64", which would allow me to have tolerances of a little
heavier than 1/8". I guess I could use these rules for framing houses
- but they still aren't accurate enough for building cabinets.

I have a Starrett dial caliper that will measure to 1/1000" - now that
will let me have tolerances of about 1/100", which is heading in the
right direction but when I think about it, a piece of newsprint is
about 4/1000", or 1/250" and I know that my joints are tight enough,
when they are cut properly, that I can't fit a piece of newspaper into
them.

And yet, that can't be possible because the best measuring instrument
that I have in my shop will only allow me to have tolerances of
1/100".

It makes you wonder why framing carpenters and masons even bother to
own measuring devices at all and, it has been my suspicion for some
time that many of them don't.

It is gratifying to me that I am capable of doing the impossible but
it makes me a bit squeamish, if you follow me. A man needs to know
where he stands in this world and how can you do that if you can't
measure anything proper like?

When I had my first philosophy course in college we studied this old
boy named Zeno the Eleatic and his paradoxes. Now, Zeno said that you
can never get from one place to another because, first you have to
cover half the distance from A to B, then you have to cover half of
the remaining distance and then half of that remaining distance, and
so on for ever and ever. So, there's no sense in trying to measure
anything because it just ain't gonna work out.

Zeno may have been the first framing carpenter, although I am not
entirely sure about that - nor anything else, it seems.





Regards,

Tom Watson
http://home.comcast.net/~tjwatson1/

[email protected] wrote:


> and the carcasse. How square do things need to be and what sort of
> fixes might I entertain if I determine that something isn't square
> enough?


Basic premise: Cabinet "components" (generally speaking the casework,
the drawers and the doors) are all made up of individual "parts".

Perfectly square is the goal for all these components, and often "close
enough" will work if you can shim "components" into square during final
assembly of the cabinet. However, once you start shimming you're wasting
time that could better be spent doing something else, and there is NEVER
a guarantee of satisfactory results.

'Measuring diagonals' will tell you whether the four sides of a
"component" (drawer, door, drawer front, or the casework) are square to
each other; using an 'accurate square' of any kind will also tell you if
the corners are square (keep in mind that without proper preparation,
the stock between square corners could still be twisted, bent, warped,
or vary in thickness, and still cause problems).

The first step in solving your problem is to answer the questions:

Did you pay particular attention to the elements of "square" when you
built both your cabinet and drawers - IOW, ALL individual component
"parts" are indeed the specified project thickness, width, and length?

Did you use properly milled, straight, stock, of the equal thickness,
for each part?

Did you "batch cut" these parts before assembly? (more below)

Did you take steps during component assembly and glue-up to insure a
square results? (measuring diagonals, proper clamping techniques to
preclude warping by too much pressure, etc?)

The pursuit of "square" is the holy grail of cabinet making ... if you
did none of the above, you may well need to start over again as you can
spend hours attempting to shim the drawer slides and non-square
components, in all planes, and still not have a satisfactory end result.

One simple method/practice which will take you a long way to insuring
that your basic components (drawers, doors, casework) end up square is
to "batch cut" ALL "parts" of like dimension for these components.

"Batch cutting" parts is the practice of using the EXACT SAME machine
setup to cut ALL like project parts BEFORE changing machine settings
(move the table saw fence, move the planer table, etc).

AAMOF, this practice can't be stressed enough and will take you a long
way toward alleviating the problem you are currently experiencing.

Examples of this:

Cut ALL your "parts" (drawer sides, rails and stiles, casework sides,
etc.) of like WIDTH in the ENTIRE project, BEFORE you move your table
saw fence from that WIDTH setting.

Cut ALL your "parts" of like LENGTH in the ENTIRE project, BEFORE you
move your table saw fence from that LENGTH setting

Thickness, to project specs, ALL stock with the SAME final setting on
your planer, BEFORE you change that setting.

Etc, ad infinitum ...

This one simple practice (which does require some organization, planning
and thought) will insure that ALL project components parts, that have
identical dimensions, in thickness, width, and length, are indeed
identical and have not been subjected to errors introduced when moving
fences, machine tables/settings, etc..

(There are other things, like when using face frame cabinets, build your
face frames first, taking the time and necessary steps to insure they
are square, then assemble your cabinet sides on top of the already
"known square" face frames).

Paying particular attention to "square" with steps like the above when
building the three basic components of a "cabinet" (the casework, the
drawers and the doors) will save countless hours of trying to fit
non-square components during final assembly.

Multiply that by the number of cabinets in the average shop built
kitchen and the importance of pursuing the holy grail of "square"
becomes paramount.
--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

[email protected] wrote:

> I started wondering about tolerances when I started trying to face
> joint wood. If I aimed for "perfect", meaning that I couldn't detect
> any deviation from my Starrett, then I'd still be jointing. And I
> concluded that seeking that level of accuracy doesn't make sense
> because of the changeable nature of wood. (I plane the other side and
> it bends 0.1" anyway.) So if perfect isn't the goal...then what is?
> I picked 0.004", but I don't know if that's the right answer.
>

CONSISTENCY in dimension is oftentimes more important than precision in
measuring the dimension.

Recall the "batch cutting" references ...

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

-MIKE- wrote:
> Swingman wrote:
>> -MIKE- wrote:
>>> Tom Watson wrote:
>>>> I went on Starrett's website and found that the most that they will
>>>> write up a Certificate of Accuracy for on a tape is +/- 1/32". They
>>>> also say on that website that whatever tool you use to measure with
>>>> should be capable of measuring to 1/10 of what your tolerance is. So,
>>>> if their best tape is only capable of +/- 1/32", then my tolerances
>>>> can't be any tighter than 5/16", which seems a tad generous to me for
>>>> cabinet work.
>>>>
>>>
>>> If you had a wall 25 feet long (length of tape measure) that had to be
>>> filled with cabinets, wouldn't 5/16" over 25 feet be tight enough?
>>
>> Different kettle of fish than "square" .. but in your example, that
>> depends upon the width of the _available_ trim piece!! :)
>>
>> IOW, your trim piece that hides the gap better be wider than 5/16" ...
>>
>> <g>
>>
>
> I'm guessing one of the end walls will be out of plumb by double that,
> anyway. :-)

And you'd most likely be right!! :)

Now, try building cabinets for a kitchen, months before the house is
even in existence!

There are more things that can complicate that endeavor than you can
imagine, and it is something I have been doing on a routine basis.

Creative solutions also have a way of being another mother of invention.

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)

[email protected] wrote:

> mostly do work that way. But being rushed by the glue is one problem
> that I didn't see a way around until I was forced to really look for
> different glue. (There is no way I ever could have put the case
> together in the 15 minutes or so that PVA allows. I enlisted my
> wife's help in glue application and assembly and it still took 45
> minutes to get everything together.) So when I glued up the drawers I
> was always rushed. By the time I would get to the point where I'm
> supposed to check if the assembly is square I've already been fighting
> against the glue and I'm at the end of my open time and probably tired
> and drained from the stress of it all. I'd much prefer it if I could
> come back tomorrow and double check the assembly for square. But I
> haven't found any 48 hour glues. At least the next time I do
> something like this I'll have an hour instead of 15 minutes---that
> should cut down on the rush factor. (The next project is a table
> which seems like it should be a lot easier to put together than 27"
> long dovetails.)

Simple solution: Don't glue-up the whole piece at once!! Break it into
two, or more, glue-ups! Rare is the component that you can't break down
into multiple glue-ups if need be.

CAVEAT: when doing partial glue-ups, it's a good practice (actually
imperative) to go ahead and clamp up the full assembly, both glued
joint, and unglued joints, and check for square.

Use some common sense with multiple glue-ups, think it through so that
you don't glue yourself into a corner; practice the steps first before
each glue-up, with clamps at ready; and your project will thank you ...

--
www.e-woodshop.net
Last update: 10/22/08
KarlC@ (the obvious)