I almost had a very discouraging weekend, and I hope somebody can help
explain what happened. I had some 4/4 oak to glue up into a 14 3/4"
wide by 24 3/4" panel. I had previously flattened one edge of each,
then rough cut them to 5 1/2" wide and about 27 " long and flattened
one face of each. I then left each board, stickered, to dry and move
for two weeks. On Sunday, my plan was to dimension it and glue up the
As I expected, my flattened faces were not flat anymore, having moved
a bit on the stickers. The boards had both bowed and crooked. (I
think I have used those terms correctly. I mean that if you laid the
board face down it made an arch, and if you laid it on edge it made an
Before I started, I changed my jointer knives. I used a dial
indicator, zeroed on the outfeed table, to set the knife height. I
couldn't get them to exactly zero, but the total error was on the
order of 2/1000". I thought that was good enough.
Then, I just couldn't flatten the boards. I have flattened bowed
stock before, and there is a rhythm to it. You hold the board condave
side down. As you pass it over the blade, you hear cut-nothing-cut.
On each pass, the "cut" phase gets longer as the flat part lengthens,
and the "nothing" part gets shorter. When you hear "cut" along the
whole length, the board is flat. This time, there were two possible
outcomes. Sometimes, it seemed the bow was unchanged and it was cut
along the entire length. At other times, I somehow created a
concavity (!) that I then had to carefully pare off.
Eventually, I gave up on "flat" (no light visible under my
straightedge) and settled on "kind of flat" (small enough errors that
I can hand plane and/or sand them away). By that time, I was down to
11/16" thick stock at the narrowest point.
I have two questions:
1) Is it possible that somehow in setting my jointer knives, I caused
this? I've never used a dial indicator to do it before, so I suspect
they were much better set than I have ever managed before, rather than
worse. Also, I would understand if improperly set knives meant the
stock wasn't coming out *square*. I don't understand how it could
cause it not to be flat.
2) I have a benchtop jointer. The total length is about 24", with 12"
on the infeed side. I know there is a limit to how long a board you
can flatten with a given jointer. Is that what happened here? Do I
need a new jointer, or at least a longer infeed table? It would make
sense that the table was short enough that it could follow the
curvature of the stock, but I think I have flattened stock as long as
38" before on the same jointer. The legs for the table I'm working on
are from 6/4 red oak, and 30" long, and I had no trouble squaring
I know it is hard to say what happened if you weren't there, but I
hope somebody has some ideas. (The glued-up panel, in case you were
wondering, seems to have turned out fine, so I didn't have to start to
A jointer with a longer bed would have definitely helped for stock of that
length. Another possibility is that your infeed and outfeed tables were not
perfectly aligned to the same plane. If one or both ends of the table were
below the plane of the knives that might explain the induced concavity.
I was troubled by several things that you said.
1. Red oak boards only 6" x 27" moved substantially? Are you buying
good kiln-dried oak? If these were 6 foot long boards, then yes, they
will move a bit, but 2 feet long boards?
2. Re: Using a jointer to flatten a wide board. This is my opinion,
but jointers really are not designed to flatten boards, they are meant
to flatten and straighten board edges for "joining". Look at how they
cut. The knives move in a circular motion as they turn into the stock,
the reaction of the unguided stock is to move up and back (away from
the knives). To compensate for the movement up, you have to put
pressure downward and forward. If you put downward pressure on a 3/4"
thick board, it will flex. If it flexes, you are not flattening it.
You are just performing a thickness planing operation versus a
So how do you flatten a board? For 27" pieces, I would bandsaw or
hand plane one side and then use a thickness planer to finish it up.
For a glue-up panel, I would glue up the panel first, aligning the
edges, and alternating boards that bow up with boards that bow down,
and then either hand-plane or sand with a good belt sander the whole
panel to desired flatness.
Ken McIsaac wrote:
> I almost had a very discouraging weekend, and I hope somebody can help
> explain what happened. I had some 4/4 oak to glue up into a 14 3/4"
> wide by 24 3/4" panel. I had previously flattened one edge of each,
> then rough cut them to 5 1/2" wide and about 27 " long and flattened
> one face of each. I then left each board, stickered, to dry and move
> for two weeks. On Sunday, my plan was to dimension it and glue up the
Let me play the devil's advocate for a moment.
Why finish machine the individual pieces when it is the assembly that
you are interested in fabricating accurately?
Your objective is to get a flat top, 14 3/4" x 24 3/4" x 3/4".
The measurements and conditions of the individual pieces are not
relevant, it is the end result that matters.
Do you have a planer?
If so, have you considered planing your stock to say 7/8", then using
the Swingman's "complementary angle" technique to obtain adjacent
edges for the glue up, using either your jointer or for short pieces,
your table saw?
That technique is slicker than frozen snot on a door knob in January.
Makes the alignment problem go away and produces nice tight joints.
Once you have an oversize glue up, head to the top shop and have them
drum sand the top to final thickness.
BTW, starting with 7/8" thick material means you don't have to be so
fussy about keeping your glue ups perfectly flat, drum sanding levels
out the top to 3/4".
Clean up the edges as the final machining process (My sled is handy
for this operation), and get some finish on the top within 24 hours or
expect it not to stay flat.
Didn't answer your question directly, but maybe gave you some food for
My first jointer was a vintage sears benchtop that someone had mounted
to a stand long ago. I think it was about the length you mention,
maybe just a tad longer, I can't remember.
I found that two feet was about my limit for a good job. No matter how
I tried I couldn't seem to get the hang of jointing boards longer than
that. I got about the same results you did, all the time.
Fortunately I had a Stanley 7 from an estate sale and I got pretty
good with it. Not great but flat enough to run it through the planer,
and that was fine. Built my workbench that way.
But when I bought that 75-inch Griz, it sure did make my life easier.
You might be able to get better at it. I've heard several people say
you can get good reference joints on a long board with a short
jointer, it's just tricky. I tried and tried, and all I did was take
good maple and make a pile of nice little shavings with it. Never got
the hang of it and I consider myself pretty good at finding techniques
for stuff like that. Frustrating, but for me it was a jointer plane
for boards longer than a couple feet.
Well --- before I dive into fixing what may not be broken....
It sounds like you had the boards stickered for about 2 weeks. Humidity is
going up around here ....is it possible that with fresh wood exposed you are
seeing the natural cupping/bowing that occurs when exposed to moisture?
Just a thought. If the machines had worked well in the past what has changed
other than fresh dimenioninf and some sticker time?
Thank you to everybody who replied. There were some very interesting
comments. Rather than replying individually, I thought it would save
time and space to write one reply addressing everything that came up.
That's a neat idea. I had been concentrating on feeding the stock
only in one direction to avoid tearout. I should have been thinking
that I can always clean up tearout later at the thicknesser. First
the damned stock has to be flat.
Lew Hodgett suggested that I was going about things backwards, and
that I should have joined the boards first and flattened the panel.
That is effectively what I wound up doing, although I don't know
anybody with a wide belt sander, so I did the flattening with a
mixture of my handplane and a lot of sanding. Anyways, the problem is
more general than just gluing a table top, right? You need to start
with nearly perfect rectangular prisms or every joint will suck.
Boxes won't be square, drawers will rock on their runners, doors will
not close... If I can't flatten a face or joint the edge at 90 degrees
to that face, it seems like I'm just doomed.
The idea of complementary angles for edge joints is a good one, but
that only helps if the boards are not square, right? It doesn't seem
like it will help if the edges are not parallel to one another.
There were questions raised about my stickering technique. I am
entirely self-taught. I read an article in the Power Tools book,
published by Fine Woodworking about how to dimension stock. It said
you dimension the stock to near width, near thickness and near length,
leaving about 1/8 extra in width and thickness, then sticker it and
wait. My understanding is that wood is naturally under tension, and
that cutting it disrupts the equilibrium and can actually cause
movement. My article says that if you let it sit for a week or so,
then shave off the last 1/8, the stock should remain stable as long as
it is in the shop.
So, that's what I did. The movement wasn't large, probably 1/64", but
that is easily enough to feel with fingers, or see, in a finished
piece, so I wanted to get rid of it.
I don't have a moisture meter, so I have no idea how dry the stock was
at any point in this operation. It has been sitting my my shop since
January, though, which should be long enough to acclimate.
In thinking about this now, I have developed a theory that my problem
was that I was taking too large a cut on each pass. If I had an ideal
jointer, with infintely long infeed table, then the outfeed table just
defines a cutting plane. You put your bowed stock down on the infeed
table and set the infeed table so the cutting plane defined by the
outfeed table is above the curvature. Then, you just feed the stock,
pressing down only on the outfeed side (where the stock is already
flat) and things are guaranteed to be perfect.
On the other hand, with a short infeed table, the trailing end of a
bowed workpiece actually hangs below the edge of the infeed table.
This causes two problems that I can see:
1) The cutting plane defined by the outfeed table is actually too
acute, right at the start. That is to say, you are going to cut more
into the leading edge (and probably more off the trailing edge, but
that's a harder sketch to draw so I'm not sure) than you want to.
2) The workpiece has to climb up the edge of the infeed table as you
feed it. This must cause either flexing of the workpiece, or a
rotation about the leading edge that lifts the partially flattened
part off the outfeed table. Either way, it ain't coming out flat.
These two things together I think explain what was happening. If I
had a 1/64" bow, and I set it for a 1/64" cut, it would take too much
off the ends and nothing off the middle. There's the mysterious
convexity I was getting. On the other hand, if I really pushed down
on it, I could (as many suggested) just flatten the stock against the
cutter and take a smooth 1/64" shaving off the whole board, thus
changing nothing about the curvature.
I think the solution is to just creep up on the cut. Assuming the
stock has constant curvature and is twice as long as the infeed table,
a total curvature of 1/64" looks like an effective curvature of 1/128"
when the edge first hits the cutter. I think I need a cut depth less
than that (around 5/1000") in order to avoid taking too much at the
ends. My el-cheapo jointer is not calibrated in thousandths, but I
have a dial indicator now! In this way, I sitll don't expect "flat",
but I should get "flat enough".
Then again, I recently saw articles about a sled for flattening stock
at the thicknesser and jointing stock with a router. Maybe I should
just put the jointer in the garage and not look at it again.:)
Thanks again for all your help!
You should joint the wood with the concave face/edge down. Why?
Because the wood will have two contact points with the table at all
times. The key to getting a flat and straight reference edge is to
make sure the edge stays firmly against the fence (the jointer's
reference surface). Your idea has some merit but may cause tearout
depending on grain direction. Once you get a flat face and square
edge, you can then go to your tablesaw and thickness planer to square
up the other two surfaces....
Nope. For millennia we have removed the high points to flatten wood.
Makes no difference whether it's a hand or powered tool, getting close to
flat and then taking the best average fit will get much more than two points
referenced, one of which may be off the table (or sole of the plane) while
the other ain't.
I know that you used a dial indicator but it sure sounds like your
outfeed table is a little too high or low or one of the knives is
slightly higher than the others. BTW I always use my jointer to face
plane one side of rough stock and I have never had any problems doing
it. The only problem I ever had was when I replaced my knives the
first time using the factory jig. A magnetic jig and a little help
from the wrec about how to set my outfeed table by sound of all
things(DAGS the group for outfeed table and you should find the thread
on how to do that), and I haven't had a problem since (excluding of
course operator errors like not checking the fence for square and
other such idiocies that I seem to be prone to). My only question is
why you would partially mill the stock and then let it sit for two
weeks. I may be doing it wrong but I always mill and glue the same
day and apply a finish asap. If I do have to let unfinished milled
stock sit I always stack it on a flat surface and weight it down so it
minimizes the chance of warping. Hope that helps. bc
Wood moves as you have readily seen. If you were experiencing problems
with the bowed stock, you could have stopped and checked your jointer
alignments (without making changes at first - adjusting after a full
assessment) then continue to join your wood.
If you joint a piece with the concave face down, and you apply ample
pressure to hold the stock flat to your jointer surface, it will
spring back to concave after the cut. Let's face it, some concavity
can be too much to successfully flatten. Not without reducing the
stock to a thin board. And that board may be tapered to some extent
depending on how you finish up the other side.
Remember that the use of the jointer is to create a reference for all
the other edges to "square to". Concentrate more on keeping your
reference edge against the jointer fence and let the knives to their
stuff. A hint for your that I find helps me is to wax the tables
before I start...
In reference after reference and from my experience I know that you
first joint cancave edges or faces face-down first. While there may be
some occasions that you might opt to do otherwise, the best results
will be achieved using the tried-and-true methods that have stood, to-
date, the test of time.....
It is, however, easier to join either end with the advantage of the space
over the knives to begin the passes than to guess the exact midpoint of the
convexity and then try to plane off equal distance either side of it to
maximize the board width. Additional advantage in some woods is that the
concave side is the sapwood side, so you'll be removing unwanted off-color
wood with things like cherry or walnut, instead of the more desirable
Concave pieces are set up with the end touching the infeed table, and the
concave portion rides clear of the knives. Pass or two to reduce the ends
to where they're close to the center, then through and through. Convex is a
bear, because if you have a tilt, you'll screw up the board width.
Standard texts should cover it if this isn't good enough.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.