edge jointing and the mysterious vanishing camber

[adrian]

I started trying to edge joint 4 ft long boards. The wood is quartersawn American cherry. The last time I tried to do this I gave up and used an electric router with a guided bit. But armed with the methods described by Charlesworth in his dvd using a cambered blade and taking stop shavings, I tried again. And as is usual, things happen that I don't understand.

I started on the first board using a bevel up LV jointer with a cambered blade, and got reasonably close, but my straightedge would rock a bit, and I couldn't seem to get out that last bit of convexity. Mysterious.

Now the LV jointer had been in use for a while. I had used it when face jointing this same timber. I switched to a Clifton #7 that I thought had a fresh blade, also cambered. That seemed to work much better, and I breezed through several boards, using the LV jointer first and then switching to the clifton as I got close and taking stop shavings to remove convexity in length and shifting the plane left or right to correct problems in the width.

Now I worked on a board that had a knot in the middle of the edge, which may not have been the nicest thing for the edge, but as far as I could tell, the Clifton was still going stong, producing fine wispy shavings.

Then I started work on a board that just wants to stay convex. It has a big bump in it and the plane won't cut the bump away. Now my planes do not have perfectly flat soles, but they nevertheless worked OK on the first half-dozen pieces I jointed. I would take stop shavings and eventually the edge would reach a point where the straight edge wouldn't rock. And the crucial test: when I lay two of these pieces together the edges meet with no rocking or gap evident. But in this case, the plane refused to cut, even though the wood had a bump. What is going on? It's the same plane that was working before. I didn't adjust anything. I suppose there's the blade.

The plane was taking nice shavings, but I thought maybe I should sharpen. One oddity was that the plane was squeaking in a certain place on the wood. I inspected the blade and it looked, from the glint on the edge, like the edge had probably deteriorated in the center region where all the work was happening. I held the blade up to a straight edge.

Hmmm. That's funny. The blade seems to be hollow in the center. I have a lot of difficulty looking at blades against straight edges and drawing clear conclusions, but it appears that the blade is making two contact points with the straight edge. I'm pretty sure I didn't sharpen a hollow into the blade! Is this an expected pattern of wear, that I would remove the entire camber in use? The bevel angle was about 28 degrees. The camber was of the size Charlesworth recommends, as I recall, though it was a while back that I established it, and I'm not sure what the blade actually looked like when I started work. I tried to re-establish the camber, but I'm not sure I've succeeded, and in any case I have not recreated the earlier delightful situation where the process was working.

I took the blade out of the LV jointer and looked at it with a straight edge. It appears to be still cambered, though it's possible that the curve is too flat in the center. (Because this is a bevel up plane the camber is much larger.) This plane was working harder than the clifton, because I had it set for a thicker shaving, and I was using it on some rougher edges. Why didn't this plane work? (I suppose the sole could be at fault: maybe it's not flat in a way that's somehow more harmful.)

 

[woodbloke]

Stop shavings will produce a very slightly concave board, so that when two boards are offered up, there will be a slight gap in the middle (traditionally the thickness of a fag paper) After a certain amount of shavings though, the jointer won't cut so the board ought to be done, ie concave, assuming of course that it's square as well - Rob

 

[adrian]

Right. I understand the principle. But it doesn't always work. When I was face jointing I would often encounter situations where there was a clear hump and yet the plane wouldn't cut. My impression is that they usually had something to do with twist, but really, I'm not sure. (The other place to lay the blame is on defects in the sole of the plane.)

In this case, again, the work is humped by a significant amount, like 0.01" or maybe more, but the plane won't cut so I can't make it better. I think the hump is actually somewhat local. If I recall right, half the board was concave (straight edge didn't rock) and the other half had the hump. (There's the odd squeaking on the wood as the plane doesn't cut; I don't think I've ever experienced that before.) If I get to the point where my straight edge doesn't rock then that means I'm slightly concave and I declare victory. As I understand it, with a 4 ft long board I might be able to plane a significant hollow into the work with my jointer, so I'm not trying to proceed until the plane won't cut. Once I'm getting close I start testing with the straight edge. (One other thing was that I decided that my 4 ft straight edge is unreliable, so until I manage to get a replacement, I've been using my 2 ft straight edge at several spots and then doing a final check with the other 4 ft boards I've already worked. I don't see a gap, but I assume a small gap is present.)

 

[woodbloke]

The straight edge as you say may be suspect. If in doubt cut off a bit of mdf at least 300mm longer than your boards and use that as a reference. The machined edge from a bit of mdf is usually pretty straight and ought to help to see where the problem lies - Rob

 

[Paul Chapman]

Hi Adrian,

Provided the boards aren't too thick, you could always try the other method of jointing them, and that is to plane them together. I always favour this method where possible and think it's far more accurate and faster.

Clamp the boards together, face side to face side, plane them straight in the length - but do stop shavings as you are now, so as to ensure that you don't plane them into a curve. Use a straight blade. When you put the edges together any deviation from 90 degrees in one board will be cancelled out in the other, even if it varies along the length of the boards.

This doesn't answer your original question about the hump and the camber but, overall, I find it a better way to joint boards and you might find it sorts out a few of the problems you are having.

Cheers :wink:

Paul

 

[adrian]

There are a few reasons why I don't blame the straight edge. One reason is that I have already jointed 6 boards and they match up with each other. They do not match up with the board currently on the bench. (They rock against each other, so they meet in a convex way.)

My 4 ft straight edge is a cheap one (US$30) made of phenolic. I noticed that it told me the board was convex if I put one edge down and it said concave if I put the other edge down. So as I noted, I'm not using this one any more.

My 2 ft straight edge is a Starrett that I think is reliable. This is the straight edge I'm using to probe for bumps until I do the final test with one of my previously jointed boards. It's a much better reference than the phenolic one. (It cost more than twice as much for half the length). And I haven't noticed any strange behavior coming from it.

 

[Racers]

Hi,

Sounds like you need to resharpen your blade.

The method I use it to match the boards together, plane one check for square and make sure it looks flat and set it to one side, plane the next one and offer it up to the first and check it matches up to the first one adjust, then plane the other edge of the second board then do the third adjust etc the only time I used a straight edge was to check they where faces where flat to each other.

I had all 6 boards for this stacked on edge on my bench without glue.

Pete

 

[adrian]

Provided the boards aren't too thick...

Clamp the boards together, face side to face side, plane them straight in the length - but do stop shavings as you are now, so as to ensure that you don't plane them into a curve. Use a straight blade.

After all that effort trying to camber now I need to go back to straight?

Thanks for the suggestion. What thickness would you consider this suitable for? My material is 0.85 inches and 1.1 inches in thickness and I had the impression that people mainly like this method for thinner like half inch stock.

I have another reason for not wanting to do it that way which is that I haven't figured out how the boards are going to go together, so I don't know which edges will the be ones finally glued together. I assume that doing it this way you not only have to put together the same boards but at exactly the same spot, so you can't shift things around at all. I wanted to get the boards to their final width to make sure my panels will come out wide enough before making the final decision about how to arrange the wood.

One last thing I was wondering about. Presumably if you do this then the angle of the edge to the face might vary over the length. It seems like this would confound my ability to tell if the surface was straight.

 

[dunbarhamlin]

Might be worth checking your straight edge anyway.
Use a piece of MDF (preferably whitewashed - easier to see) score down the length of the straight edge, flip it and score down the same edge over the same line, with end points common to your original line. If its out of true, the resulting line will thicken/split into two. (Rotating instead of flipping could give a false true if the edge is sinusoidal)
(Obviously, also need to check the straight edge edge isn't in wind for this to be worthwhile)
(I've had a (Axminster) 'straight' edge where only one (bevelled) edge was good, and the square edge was significantly out.)

But for the same issue having manifested when edge planing (unless these are squarish in cross section) the stubborn high spot on the face could have been due a low spot on the oposite side (unwound shavings good for support in this case)

 

[Paul Chapman]

Assuming you are using a #6 or #7 plane with a 2 3/8" blade you should be OK with boards up to 1" thick. If you need to remove a lot of material to get them straight you could plane most of it off the boards individually, then finish them off together.

People often criticise this method because they say that if you have to separate the boards and then put them together again you will have difficulties. However, I just square a line around the boards, say an inch in from each end, and it's easy to line them up again. You will plane the line off the edge but simply join it up again.

The method only really works if you sort the boards into order first - so if you don't want to do that, it clearly won't work for you.

It shouldn't cause you any problems in terms of testing the straightness of the boards. You will find that, using your fingers as a fence, you will be able to plane the boards to virtually 90 degrees anyway. To test for straightness, just put one of the boards on top of the other. The problem you are currently having with straight edges only really arises because you are planing the boards separately. If you do them together, you don't really need a straight edge because when you put the boards together, it's obvious where any discrepancy is.

Hope this helps.

Cheers :wink:

Paul

 

[adrian]

Might be worth checking your straight edge anyway.
Use a piece of MDF (preferably whitewashed - easier to see) score down the length of the straight edge, flip it and score down the same edge over the same line, with end points common to your original line.

So the idea is to draw one line and then flip the straight edge over and draw the same line the other way? This could fail if the straight edge is symmetrically out of true, like if it had a hollow or bump right in the middle.

But for the same issue having manifested when edge planing (unless these are squarish in cross section) the stubborn high spot on the face could have been due a low spot on the oposite side (unwound shavings good for support in this case)

I'm not sure what you mean here. Are you saying that when face planing the problem could have been due to the board flexing? (I shim the boards all around to prevent this when face planing.)

 

[MooreToolsPlease]

If the edge has a significant bump in the middle, and you are starting at the end then if it is big enough the plane wont cut.
for boards with a bump in the middle, start in the middle and work your way out.
Thats how I have always tackled this problem, although I use an electron burning planer.
Having said that, the bevel up jointer I received today seems to do a pretty good job after a clifton#7 for those special edge joints :wink:

 

[dunbarhamlin]

Sorry - not very clear. As you said, turning it end for end could falsely indicate all was well if had matching high and hollow in each half.
Mark the line against the straight edge, turn it over like a page of a book and mark again against the same edge, so the same end of the straight edge is towards you both times, and you've marked first to the left, and then to the right of the straight edge. This gives a mirror image of the original line - provided the edge isn't also twisted (in wind) which would mean the two arrisses of the tested edge wouldn''t match.
(Same technique as testing a square)

Yep - did mean flexing (and remedial shimming)

 

[adrian]

It shouldn't cause you any problems in terms of testing the straightness of the boards. You will find that, using your fingers as a fence, you will be able to plane the boards to virtually 90 degrees anyway. To test for straightness, just put one of the boards on top of the other. The problem you are currently having with straight edges only really arises because you are planing the boards separately. If you do them together, you don't really need a straight edge because when you put the boards together, it's obvious where any discrepancy is.

If I have to take the boards out of the vise to test for straight and then put them back in and get them aligned together again each time that seems terrible.

For some reason, my problem seems to have been interpreted as a measurement problem. I really do not think that I have a measurement problem. I did 6 boards with my current measurement process using my 2 ft straight edge to test my progress and a final check with one of the other already jointed boards.

To repeat, I do not have a problem with straight edges. I have a problem with planes: the planes won't cut off the bump (in length) that I know is there because my straight edge and my previously jointed boards all tell me I have a bump. Knowing where the discrepancy is doesn't help a whole lot if I can't somehow exploit that information to remove the discrepancy. (Actually I find it rather difficult to decide where, exactly, a bump is located. It arguably may not But that doesn't matter if the plane will take off high spots. I don't need to know. I just need to know that there is a bump.)

But I find myself in the situation where the plane won't cut any more. That is supposed to mean that I have a hollow. In fact, on a 4 ft board it should be a perhaps somewhat large hollow. But that's not the case. I have a bump. Really. It's the wood that's got the bump, not the straight edge.

Presumably this same situation could arise with two boards together.

One observation: if the straight edge has the bump it has to have exactly the same bump on both edges, because I've flipped it over and gotten the same result. If the straight edge has the bump I should have an enormous hollow between two of my boards that match the straight edge. Nope.

 

[adrian]

If the edge has a significant bump in the middle, and you are starting at the end then if it is big enough the plane wont cut.
for boards with a bump in the middle, start in the middle and work your way out.

I couldn't find any place where the plane would cut along the edge.

So it's possible that I can take stop shavings and the plane won't cut when there's a bump?

 

[Paul Chapman]

I have a bump. Really.

If you really have a bump and the plane won't cut, then it sounds like there is a temporary problem with the plane. Is the blade sharp? Do you have the cap iron set too close to the edge of the blade? Has the cap iron slipped down because it wasn't tightened up properly, thereby stopping the plane cutting?

Can't think of anything else :?

Cheers :wink:

Paul

 

[MooreToolsPlease]

If you are pushing down on the back of the plane then its possible that the edge of the blade wll never even touch the wood.
How big is the bump in the middle of the board?
If it were me, then I would start in the middle and making sure that the plane is close to level slowly work from the middle out, as soona s you have established a flat it will be easy to rectify

 

[PaulO]

Stop shavings end to end will only produce a concave edge if the work piece is the right size. If the piece is longer than three to four times the length of your plane you have to use a different method, checking and working on smaller sections with a straight edge.

You talk about rocking the straight edge to test for bumps. The method that I use, and was taught by David Charlesworth, is to rest the straight edge on the surface to be tested. Support (but don't grip firmly) one end of the straight edge. Now move the other end of the straight edge in the plane of the surface. As you will do this you will see that the straight edge will pivot about the highest spot. If you have achieved a concave surface it will pivot about the end you are supporting. You can check the amount of concavity by inserting shims under the straight edge until it pivots about the shim.

WRT the camber on the blade, it is highly unlikely that you have removed the camber through use. What is more likely is that your sharpening technique has focussed too much pressure on the centre strokes. I test the camber by sighting my blade against a 1" thick piece of ebony that I have planed flat. With that contrast it is easy to see the camber. David C uses a piece of phenolic. What you obviously don't want to check your camber against is a metal surface, as you will destroy your sharp edge.

So I think your problems are a combination of
1) Trying to straighten (concave) too long a piece using stop shavings
2) Possibly a problem with the sole flatness
3) Possibly trying to take too fine a shaving for the size of bump you have
4) Problems with your sharpening technique (maybe)

In edge jointing the camber of the blade should help you square the edge, using the drifting method. But also gives you a slightly concave edge across the width, which in combination with a slight concavity along the length should give a nice tight edge joint.

All of these topics are covered in David Charlesworth's excellent tool tuning course, which I can highly recommend. I have just booked myself in for another week with him next year, and I notice that he has dropped his prices.

 

[adrian]

I have a bump. Really.

If you really have a bump and the plane won't cut, then it sounds like there is a temporary problem with the plane. Is the blade sharp? Do you have the cap iron set too close to the edge of the blade? Has the cap iron slipped down because it wasn't tightened up properly, thereby stopping the plane cutting?

Well, I didn't really think I had a dull blade, because I hadn't noticed any drop in cutting performance. But as I noted in my first message, I inspected the blade and found that it seems to be reverse cambered, with a hollow in the center. (I say "seems to" because I always have uncertainty about what the blade edge looks like when I put it up to a straight edge: light always shines through everywhere.) This is particularly strange since I sharpened the blade to have a camber. I sharpened the blade and tried to restore the camber. I'm not sure I got an adequate camber. But afterwords I reassembled the plane and it still wouldn't take off the bump.

The cap iron is not in the way. The plane cuts. It just won't cut what I want it to cut.

Does the squeaking signify anything?

 

[adrian]

Stop shavings end to end will only produce a concave edge if the work piece is the right size. If the piece is longer than three to four times the length of your plane you have to use a different method, checking and working on smaller sections with a straight edge.

My plane is 21 inches, I think. The workpiece is 48 inches. I'm below your limit, at less than 2.5 times plane length. I'm curious about the recommended approach for longer, though.

You talk about rocking the straight edge to test for bumps. The method that I use, and was taught by David Charlesworth, is to rest the straight edge on the surface to be tested. Support (but don't grip firmly) one end of the straight edge. Now move the other end of the straight edge in the plane of the surface. As you will do this you will see that the straight edge will pivot about the highest spot. If you have achieved a concave surface it will pivot about the end you are supporting. You can check the amount of concavity by inserting shims under the straight edge until it pivots about the shim.

I've done this too, but rocking gives me some information about how big of a bump whereas spinning the straight edge only tells me that there is a bump. I do both, actually. I spin first and then rock.

I have Charlesworth's DVD on jointing where he shows the use of the stop shaving method, and also this approach to finding bumps.

WRT the camber on the blade, it is highly unlikely that you have removed the camber through use. What is more likely is that your sharpening technique has focussed too much pressure on the centre strokes.

So you're thinking that I may have had a camber a while back, but I removed it by sharpening with too much focus on the center? (This is a possibility, though I don't think I could have gotten a concave edge by sharpening. I suppose I can go check my stones right after I flatten them and see that they really are flat and not crowned. I'm not sure I want to spin my straight edge on my stones, though.)

I test the camber by sighting my blade against a 1" thick piece of ebony that I have planed flat. With that contrast it is easy to see the camber. David C uses a piece of phenolic. What you obviously don't want to check your camber against is a metal surface, as you will destroy your sharp edge.

Is phenolic flat? (My phenolic straight edge seems to be a little dubious.) I thought ebony was a kind of unstable wood, prone to warping and such. Do you find that your reference stays flat?

So I think your problems are a combination of
1) Trying to straighten (concave) too long a piece using stop shavings
2) Possibly a problem with the sole flatness
3) Possibly trying to take too fine a shaving for the size of bump you have
4) Problems with your sharpening technique (maybe)

So what about #3. If the shaving is set too small can I get the situation where the plane won't cut? (In other words, it's not just that I'll have to take extra passes?) The shavings were about 0.002" or a bit under. I was using other planes for thicker shavings, and I found that if I got a great deal thicker then I had to rely on momentum to keep the plane moving, and hence I lost control.

All of these topics are covered in David Charlesworth's excellent tool tuning course, which I can highly recommend. I have just booked myself in for another week with him next year, and I notice that he has dropped his prices.

For various reasons, it's impractical for me to take a course right now, so I have to make do with the DVDs. I notice that a new one, 188 minutes long, is coming out.