Replies

1. 
   eddiefromAustralia | Jan 20, 2004 07:50am | #1

   Rod,

   Does the problem come from the jointer?  (was the jointed face uneven or out of square to start with)

   Has the board moved after planing (does the jointed face cup more after the material is removed from the opposite side of the board?)

   Is the thickness of the component uniform across it's width? (are the knives of the planer parallel to the support table)

   All a planer does is make things parallel - where the wood goes after it's relieved of the pressure from the pressure bars is anyone's guess.

   Just a few thoughts to get you looking for the problem, if it comes from the planer.

   Cheers,

   eddie

2. 
   JeffKurtz | Jan 20, 2004 07:54pm | #2

   Not sure if I'm reading you right, but are you edge jointing before you plane?

   Jeff

3. 
   johnhardy | Jan 20, 2004 09:09pm | #3

   I wouldn't measure parallelism this way. Get a pair of calipers and measure the thickness on one edge and compare the thickness to the other edge.

   Measuring squareness is meaningless ... you might not have jointed a true right angle, your square might be off, etc.

   John

4. 
   rslatm | Jan 20, 2004 11:22pm | #4

   What kind and size planer you talking about?  I usually check for parallelism by indicating off the table to the head. You can have a knife crooked which could give a taper.
1. 
   lostarrow | Jan 21, 2004 12:44am | #5

   What johnhardy said. Looks like your planer is not cutting parallel. I was having the same problem a couple of weeks ago. I was machining oak for raised panels and I noticed that it was taking more effort to push one long edge through under the hold down than the other edge. These were 12 1/2" wide panels and when I measured the thickness it was a strong 1/32 thicker  from one edge to the other. If you lookup post #11939.5 from JohnW you'll find a fix for the problem. It worked for me.Be not afraid of going slowly.  Be afraid only of standing still.   chinese proverb

5. 
   JohnWW | Jan 21, 2004 02:59am | #6

   Rod,

   The usual order of operations for squaring up stock is to flatten one face of a board on the jointer, then plane the board to an even thickness with a planer.  At this point, if you lay a straight edge across the faces of the board, both faces should be perfectly flat and the stock perfectly even in thickness across the full width of the board.  If there is a gap under the straightedge it should only be a few thousandths of an inch.  If the thickness is uneven it should also vary by only a few thousandths.  The best tool for judging thickness accurately is a dial caliper.

   Once you have both faces flat, one edge is straightened and squared up on the jointer.  After you have one good edge, the board is ripped to width on a tablesaw and finally the sawn edge is also cleaned up on the jointer.   

   Following that order of stock preparation, if the tools are sharp and well tuned, and the wood isn't moving from stress relief, you shouldn't see any gap between the plane blade and the stock when you check for square. 

   As mentioned earlier, check that the board coming out of the planer is even in thickness, it is a common source of trouble that is often overlooked.

   Before you start checking the stock and the machines, the first thing to test is the square you are using.  I have found from long experience that a lot of squares, probably 25% or more, aren't square at all.  If your square is off, you'll go crazy trying to find a fault with the machines that isn't there.

   Hope this helps, John W.

    

1. 
   rod | Jan 21, 2004 07:21am | #7

   All,

   Thanks for your suggestions.  I am still investigating.  If I still have trouble I will call Tech support, but I wanted to know what was acceptable first.   The knife head is parallel to the bed within a couple of thousandths.  I had to adjust the chip breaker because it wasn't within the manual's tolerance.  I also had to make adjustments to the infeed roller and the outfeed roller.  All these adjustments were minor.  I adjusted the rollers in the bed to be up .002 which made a long strait edge just touch the rollers and the leading edges and trailing edges of the bed.  I guess the bed is slightly convex.  I did use a dial caliber on the boards and the edges were a uniform width.  I used three squares two small machinests squares and a small Starrett.  When I clamped the boards as if to glue them edge to edge there was a noticeable small convex vee in the bottom.  I had ripped the boards on a bandsaw to get 4 boards out of a larger one rather than 3 so the final step in the preparation was to plane the band sawn edge to final width taking off about a 1/16 th.  These are parts for the legs of a bench.  I can use hand planes to get these parts right, but when I go to rip the parts for the top, I would like the planer to give a little better results.  All my earlier projets were with smaller stock, so I don't know if this is a recent problem or one I had from the start.  I certainly never noticed it before.  By the way the knives were all within .004 of each other and across their width.  I'll let you know what I had to do.

   Rod

1. 
   VeniciaL | Jan 21, 2004 07:19pm | #9

   Rod,

   You should be able to adjust the parallelism of the planer cut to better than a few thousanths inch. Using a caliper is helpful for routine checking which needs to be done occasionally.

   If a few thousanths is OK, then you're done. But that's enough to result in an easily felt ridge from one board to the next when two are glued up edge to edge. They should be the same thickness to such a "tactile" test.

   The folowing is an easy routine that takes less time to do then describe, once it becomes part of your adjustment process. It usually does not have to be repeated until you change knives.

   To adjust the final amount, joint the faces of two, 2" wide test boards. Run them through the planer (with the flat face down, of course). One board should be run through as far to the right as possible, the other as far left as possible.

   Joint the left edge of the right side board and the right edge of the left side board. You must be sure your jointer is set to give you perfectly square edges to a jointed face. (There is an easy technique that will test for this to at least twice the accuracy of a machinist's square. Let me know if you want a description of that). Hold the test piece faces firmly against the jointer fence. You MUST use the original jointed face (NOT the planed face) as the reference to joint the edges.

   If the edges are not square, the boards will gap in the next step and give a false test.

   Lay the test pieces down on a flat, true surface (the jointer table is good) in the direction and same left-right orientation they went through the planer, reference faces down. Bring the jointed edges tightly together. Any non-parallelism of the planer will be felt as a sharp edge, right at the joint as you run a finger over the planed surfaces, side to side. Adjust the planer until further tests bring the alignment of the board surfaces to your liking. At that point, a caliper test will probably be within the smallest increment that can be read.

   VL

1. 
   rod | Jan 22, 2004 05:58am | #10

   Venicia,

      Please send me the test for squareness that beats a machinists square.  I'm sure that others are interested too.

      I went through the maintanence list in the planer manuel and made the adjustments to the best I could in the time I spent.  It has paid off in better parallelism but it is not perfect.  Perhaps the tests you suggest would result in an indication of how to better place the knives.  I've only done this once and as I recall it took a lot of time.   My planer is a 15 inch Grizzly.  I used the Planer-Pal knife setting jigs to help with the process.  It seems the goal with these jigs is to set the knives all at the same height above the cutter head, where as if I understand your goal is to set the knives so that they are the same distance above the reference surface (which I called the bed earlier). 

   Is the process to set the knives roughly with the jigs tightening only the end bolts then rotate the cutter head and measure the height above the bed rotate back up and make minor adjustments to one end or the other and repeat the measurment underneath and adjustment until you are satisfied.  Tighten the remaining bolts and move on to the next knife making sure one end of the next knife is at the same height as the same end on the previous knife?

      I reset my jointer knives to make sure that the problem didn't lie there.  I jointed a face on a 6 inch wide 1.5 inch thick board (ash) and squared up an edge to that face (using a machinest square against light to test) then I took a light pass through the planer.  The same square was held against the squared edge and a 0.007 feeler gage would slide under the end of the blade of the opposite edge.  A dial caliper showed about 1/128 inch difference in thickness at the two edges.

      I spent two afternoons fiddling with the planer and one with the jointer.  I want to work on my bench,  but I would like to get better results and I will try your method of setting knives later on.

      I want to thank everyone who responded.  It is really helpful to a lone woodworker to be able to have conversations about how to fix a problem.

   Rod

1. 
   VeniciaL | Jan 22, 2004 06:53am | #11

   Rod,

   Adjusting the planer for blade parallelism with the bed is a different process than setting the knives to the cutterhead. First the knives must be set as you have done.

   The cutter head raises and lowers on the lifting screw mechanisms on the left and right sides of the machine. The two screws are synchronized by a chain drive and sprockets either at the top or the bottom of the machine. To adjust parallelism, the non-drive lifting screw has to be loosened from the chain and turned independently of the lifting screw on the drive side. This raises or lowers the end of the cutter head on the non-drive side while the drive end stays put.

   I don't know how your machine accomplishes this. On machines that allow you to loosen a central locking screw/nut that secures the sprocket to the lifting screw, very fine adjustments can be made and the locking screw reset. On machines that have the sprocket permanently fixed to the lifting screw, the adjustment can only be made by loosening the chain, removing it from the sprocket and turning the sprocket through an angle equivalent to the tooth pitch. This is still a very fine vertical movement of the end of the cutterhead.

   Whatever your machine's method, you should be able to get much better results than .007" deviation in 6" as you report. The two board test method I described in my first post will give you better measurement of the parallelism than using a square on the edge of a board.

   I'll describe the "square test" in the next message.

   VL

2. 
   rod | Jan 24, 2004 05:55am | #13

   Venicia,

      I tried your two board test and the planer passed perfectly!  Still, I was having wood planed with the planed face not parallel to the jointed face.  As I said before it appeared that the bed was convex front to back.   Using a long straight edge I set the roller height on the left and right to be just under the straight edge ( oriented front to back).  The rollers were about .002 above the bed on the left and about .010 on the right.  I inserted a 3/4 mdf as a false bed.  It is supported by the leading edge of the bed, the trailing edge and the two rollers.  Now the cutter head was .010 high on the right side so I loosened some bolts and adjusted the right side of the bed down .010.  A test cut showed parallel!  If you hadn't encourged me, I might have been satisfied with less.  Thanks,

   Rod

3. 
   VeniciaL | Jan 24, 2004 07:57am | #14

   Rod,

   I'm glad you have improved your machine's performance. I have to admit that I now can't follow just what was wrong. I don't undertand what was happening regarding the edge being out of square - that was unclear from the first.

   But if the machine is now performing, that's good.

   VL

4. 
   VeniciaL | Jan 22, 2004 08:23am | #12

   Rod,

   Here is the squareness test. It is a variation or extension of a commonly known technique that uses an object to mirror itself, making any deviation twice as large, therefore twice as easy to see and to correct. In the case of a cutting tool (table saw, RAS, jointer, the deviation can be tripled or quadrupled if desired.

   To test a tri-square, hold the handle against the straight edge of a sheet of plywood and carefully knife a line on the sheet against the square's blade. Keep the handle against the edge, but flip it so it is pointing in the opposite direction. The blade will only line up with the knife line if the handle and blade are minutely close to or exactly 90 degrees. If their joining angle is more than or less than 90, that error will be doubled by the blade and knife line.

   To set a saw, prepare two sheets of 1/4" plywood about 12" x 12". Cut straight edges on each, or rely on the factory straight edge (it's usually perfect). Hold the sheets together, one on top of the other, the reference edges against the RAS fence or the back fence of a TS cross cut sled. Cut the adjacent edges of the sandwich simultaneously. Flip the top piece so that it lies on the saw bed or sled bed, its top surface down, but keep the reference edges of the two boards against the fence. Bring the newly cut edges together. They will either meet perfectly along their length (while the reference edges are held tight to the fence) or one end or the other of the cut will gap. Adjust the saw accordingly until the gap disappears.

   This is far more accurate than using a machinist's square against the cut board.

   Or continue to cut the original bottom board, using the first cut edge as a new reference edge, then the second cut edge as another reference, and cut the last edge (the fourth edge, counting the reference edge). Hold that last cut edge against the fence and bring the two boards together. Any deviation from a square cut will be multiplied by 4.

   On a jointer, joint one face each of two 14 inch-long 4 x 4s. Use those faces as references against the fence to joint a second face on each board. Put the reference faces down on the jointer bed and bring the freshly jointed faces together (there is only one orientation in which this can happen - one of the boards must be turned end for end). Any deviation from squareness of the fence will be revealed by an angular gap between the faces. The angle will be twice the fence error. Any jointer (once squared) that can't hold up to this test while its fence is moved side to side (then locked back down) is not an accurate macine.

   The jointer method can also be increased to indicate 4X the fence error using the logic from the saw example.

   VL

6. 
   cowtown | Jan 21, 2004 09:42am | #8

   me, I'm not sure I understand,,,,

   is "planer" being used instead of jointer? Kinda sounds like a possibility to me.

   But if it ain't .008 is only two thicknesses of a piece of paper over 4", and while a planer should be able to be a little more parallel than that, maybe the answer lies in blade adjustment or blade wear

   Or maybe the heat of the hand holding the micrometer affected the reading. Who knows.

   Why, if you feed all the stock into the gizmo in the middle, that's where the wear on the blades takes place, and if you should have rollers, why even a small degree of sap/ pitch buildup could account for such a small amount of discrepancy.

   Heck, even a nick in a jointer blade could produce that much deviance when you feed the stock throgh a planer, and any planer user must have encountered a "chip" stuck on or around a bed roller,

   Eric

   in Cowtown

   .