Bandsaw Resawing: How To Set Fence So Blade Cuts Exactly At Middle

[John Poole]

On a band saw, I want to resaw some Port Orford cedar decking (5 1/2" x 1 1/2" x whatever length) to basically get two thinner boards of equal thickness from one, e.g. 5 1/2 x 11/16" x whatever length. I want to maximize my yield and therefore minimize the loss on thickness. Ideally, 1 1/2" split into two less 1/2 the size of a saw kerf (theoretical 1/8")should yield 11/16" (3/4" - 1/16") without planing. I'm finding that the method of setting my fence so it placed the blade smack in the middle of the width of the board proved problematic.


282448


I tried to approximate the one half mark, then by hand turning the bandsaw wheel and pushing the work into the blade I create a light cut, almost a scar or sratch, of where the line is. I then flip the wood piece upside down and do the same thing. If my fence is exactly 1/2 the wide of the board away from the center of the blade, the both scratch marks will be in the same place and the pieces should be of equal thickness in an ideal cut. If they are off, then I need to nudge my fence 1/2 the distance between the two scratch marks to resolve the distance between them.

What I found was using this eyeball blade-scratch method, I couldn't accurately nudge the fence so that both cuts were in the same "trench" [saw cut]. When I thought I was close enough, I proceeded with the cut only to learn come planing the pieces that I was off by good deal as one piece planed and the other did not start planing until the 3rd run-through on the planer; basically one piece was several 16ths thicker. I take light cuts on my planer.

Does anyone have a method they use to basically place a fence 1/2 the thickness of a given board from the center of the blade? I like techniques that do not require setting to some ruler, but rather by fit or slight of hand using geometry. Like some other tricks of the trade, I'm guessing there is some manner which does not require squinting and determining if two saw kerfs are perfectly in alignment or not.

This leads into another problem I'm wondering about: how do you determine the center point of a rectangle that is not a square? A combination square with a 45 degree angle accomplishes that task for squares. If I mark the center point of a rectangle, then I could try aligning the blade in front of the center-point mark while holding the piece securely against the fence and then locking the fence once that sweet spot is arrived at.

[Andy Shearon]

This leads into another problem I'm wondering about: how do you determine the center point of a rectangle that is not a square? A combination square with a 45 degree angle accomplishes that task for squares. If I mark the center point of a rectangle, then I could try aligning the blade in front of the center-point mark while holding the piece securely against the fence and then locking the fence once that sweet spot is arrived at.

I might be missing the point of your question, but if you want the center point, draw a diagonal line between the corners, and the intersection is the center.

Andy

[David Wong]

Have you determined that blade drift is not causing the cut to wander? Several 16ths seems like more than a measurement error.

[John Poole]

Yes, but the act drawing on a rectangle is subject to a high degree of variance. For example, the width of the lead, the alignment of the ruler/bar offset from the corner to allow for the marking instrument's width. Drawing on a square surface using a Starrett combination square will be highly accurate because of the 90 degree holder. With a rectangle, you do not have that built-in stabilizer. I recall wrestling with rectangular ends trying to draw lines form corner to corner resigning myself to the fact that center point might be several thousands of an inch off and that it doesn't matter -- I wasn't doing high precision work. There's got to be some trick that doesn't depend on your eyesight and lead thickness that assures making a mark at the precise center.

[John Poole]

I had not considered that, but will do so in further tries. Thank you.

[Harvey Miller]

Do you know the three pin method? Drill three equidistant holes in a piece of wood and push in pins (straight nails). The centre nail should be flush to the jig face. Then angle the wood jig so the two outer pins are tight against the sides of your board and give the centre nail a whack (or you can scratch in a centre line if the nail is just proud). That will mark the mid point of the board.
As for the the bandsaw, I’m no expert on bandsaws but I’d try the same thing with the band saw running and watching the cut line. For me ‘wasting’ a 1/4” of wood is worth it in my time and frustration level.

[Brian W Smith]

A very slight tangent.........

To begin with,it's rare that we use a pce "straight off the bandsaw"....or any "saw" really.Meaning there's almost always a second op.In the rare occasions we may be looking for a particular "finish" on,say a pce of rough cut siding.Our tolerances for this would be also,"rough".It sort of goes with the territory.

Trying to minimize waste....maxing out a yield......or in anyway making the cut somehow "better",is always at the forefront however.So your question has some validity.I'd have a dial calipre and a D.I. at the ready for any BS work,but that wasn't part of your criteria.....so,good luck and keep at it.

[Brian Tymchak]

Not saying this is the best method, and I'm certainly not a pro with a bandsaw, but I've had reasonably good luck resawing boards in half by using a tip I learned here on SMC. Take a square, setting it to as close to 1/2 the width of the board as you can, but not needing to be exact. The closer the better though. Then mark a line down the center of the entire edge of the board using that square. Then, do the same thing from the other side of the board so that you end up with 2 lines close together. The true center of the board lies between these 2 lines. Then as you cut, keep the blade as close to the center of these 2 lines as you can. Instead of using a fence directly, I used a point guide on my fence that allowed me to make minor corrections as the blade wandered a bit. You definitely want to crank up the tension on the blade to minimize the drift. And you have to slow the feed rate down as you resaw to clear the saw dust. that will also help to minimize the blade from wandering in the cut.

[John Coloccia]

There's something that doesn't add up here. You were trying to split a 1 1/2" piece of wood, you did all of this careful setup, and when you were done you were off by nearly a 1/4"?? That's a HUGE amount to be off, even if you just eyeballed the entire setup.

Generally, I just measure and make a little mark on the board. Then, I either grab a pencil and draw a centerline freehand, using my fingers as a guide, or if I'm feeling industrious I grab a marking gauge. No reason to do all this board flipping and stuff like that.

But that 1/4" is just gargantuan on a 1 1/2" piece of wood. Is something else going on? The blade bowing, maybe? Maybe try cutting some test strips, maybe an 1" x 1", and see if you can reliably cut those in half.

[jim gossage]

I built a small hardwood block that rides on my bandsawfence rail and can be locked down just like the fence. You can use this blockand some feeler gauges to exactly dial in your fence. Let’s say you need to move the fence 0.01” to the right: place a 0.01” feeler gauge to the right of the fence head, slide the block up against the feeler gauge, tighten the block on the rail, remove the feeler gauge, then slide the fence head against the block and tighten the fence. You could also use this block to set up your cut in the first place without making the test cuts: measure your blade kerf and wood thickness with a caliper, subtract the kerf from the thickness and divide by two to get your final thickness, slide your fence against the left side of the blade until it just touches the teeth and then lock it down, slide the block against the fence head and tighten it, loosen the fence and use your caliper to move the fence to the left by exactly the final thickness of the wood and lock it. I just tried this and got within 0.005”
282455

[Jim Matthews]

I adapted the Rockler Thin rip jig (http://www.rockler.com/thin-rip-tablesaw-jig?utm_source=google&utm_medium=cpc&utm_term=&utm_content=pla&utm_campaign=PL&gclid=CIrB7NPiy7wCFSbxOgode1UA0g) to describe the same width from the blade on the "free" side of the cut.

I true the face that rides the fence and square the edge that lies on the table.
Those stay against the fence and table, for all cuts in resawing.

After I take a pass, I advance the fence until the next face reaches the Thin Rip Jig wheel.

This all presumes you've got a properly adjusted blade set to track straight.
That's a big presumption, really and the bane of my resawing attempts.

I follow the method shown in the video below.


http://www.rockler.com/thin-rip-tablesaw-jig?utm_source=google&utm_medium=cpc&utm_term=&utm _content=pla&utm_campaign=PL&gclid=CIrB7NPiy7wCFSb xOgode1UA0g

[Jim Finn]

I resaw a lot. Mostly cedar and maple, and some oak and walnut. To find the center I just measure as best I can from each side of the wood and cut between the marks as has been suggested by Brian. I use high tension on my 5/8" carbide blade from Grizzly and a resaw fence. A sharp blade is imperative. I replace my blades as soon as they start to require more pressure to move the wood through the saw. Expensive, true but then so is wood.

[Kent Parker]

Hey neighbor John,

Whereas careful measurements are important to achieve your goal there are additional factors to consider. Drift has already been mentioned to you and are we to assume the planks are uniform in thickness with edges square to the sides ? Cutting a piece free hand in two will give you an understanding of your bandsaws capacity and the direction that the fence should be placed, ignoring the edges of your table as reference. Cedar is considered a softwood however Port Orford cedar is very resiny which contributes to friction during cutting. During cutting the wood chips tend to expand when they leave the “mother ship”. You’ll need a blade that can evacuate the dust quickly to prevent choaking or stalling the blade, so blade hook angle and set need to be considered. The long pieces you’re working with are large levers and if not supported in their length and held securely to a tall fence can easy flex your fence, the blade and any feather board (s) your using. Additional factors could include the HP of your saw and the rate in which you feed your wood and the bearing support mechanism of the blade.

The fence you’ve seen in front of my house is entirely Port Orford cedar. Much of it resawed on a small, old Delta band saw with an underpowered ¾ hp motor. Photos below of my set up. Oh and the beer bottle was just placed in the photo for height comparision...honest..

Cheers,

Kent

282458 282459 282456

[John McClanahan]

So, is that green bottle some kind of lube? :D:D:D


John

[Kyle Iwamoto]

Dumb question #1, did you joint/plane at least the bottom and side that rides the fence?

I use the Kreg resaw fence that is curved, allowing you to adjust for blade drift. You could try that.

OH, and Kent, you should not be drinking beer and putting it on your table. It leaves those rust "O"s that are hard to get out. Use a coaster.:)

[Pat Barry]

I tried to approximate the one half mark, then by hand turning the bandsaw wheel and pushing the work into the blade I create a light cut, almost a scar or sratch, of where the line is. I then flip the wood piece upside down and do the same thing. If my fence is exactly 1/2 the wide of the board away from the center of the blade, the both scratch marks will be in the same place and the pieces should be of equal thickness in an ideal cut.

This is all you need to do (t get things started). If the results weren't satisfactory, the something else was wrong, ie: drift, blade tension, fence misalignment, operator error, etc.

As far as the method of drawing two lines from the corners and finding the intersection, this should be as good as your Starret Gauge, except of course, this method is only accurate for parallelograms and perfect rectangles, not all quadrilaterals. Don't you have the same problem with the Starett anyway (pencil line width, etc)?

[John Coloccia]

So, is that green bottle some kind of lube? :D:D:D


John

I prefer my lube a lot darker than that, but I have to say that picture is extremely irresponsible. You should ALWAYS use a coaster on cast iron.

[John Poole]

Beautiful jig! Guess who's building a jig like that this weekend.

My pieces came out square, I've got all my tolerances and techniques down. The single problem I faced was setting the blade and the exact middle -- this is a problem I've face before and did the old trial and error way and adjusting, then I thought with all the vocal woodworkers on the Internet, I'd tape their wisdom.

[John Poole]

I also posted my question on reddit's forum. Folks interested in this topic may want to see the responses on reddit:

http://www.reddit.com/r/woodworking/comments/1xvmqo/bandsaw_resawing_how_to_set_fence_so_blade_cuts/

[John Poole]

Just another thought, I saw some comments about having "high tension." That suggestion needs to be qualified that it is dependent upon the blade. For example, I'm using Timber Wolf Band Saw Blade 2-3VVARI PC .025" (just put it on for this test). Timber Wolf goes into great detail on how to properly tension their blades, see: http://www.suffolkmachinery.com/six-rules-of-sawing.html. It looks like Timber Wolf's blades are low tension. I'm just mentioning this to underscore the importance of knowing what the manufacturer's tension specifications are for their blades; they should know best as they've undoubtedly tested then to the nth degree.

[John Poole]

Do you know the three pin method? Drill three equidistant holes in a piece of wood and push in pins (straight nails). The centre nail should be flush to the jig face. Then angle the wood jig so the two outer pins are tight against the sides of your board and give the centre nail a whack (or you can scratch in a centre line if the nail is just proud). That will mark the mid point of the board.


Of course: that works. And I don't need to know the midpoint of the long dimension, just having a mark somewhere on the center line of the width dimension is precisely what I am seeking. Make a pin prick, place the face behind the blade and center the prick behind the blade and then bring the fence to the wood. I think this may be the ultimate solution. (I have another I was churning in my mind about the scarring equidistant from the edges and then using that small amount of space between the different lines as the field in which to center the blade line. But the three pin method is much less effort and possibly more precise.)

Thank you.

[Brian Holcombe]

Someone beat me to it, If the board is not jointed on the side that meets the fence that may be where your issue is forming.

[Kent Parker]

John,

Many moons ago I cut out some measurement sticks out of hardwood that allow for quick set up for a variety of purposes. 4" in length, 1/2" in width and one each of 1/8", 1/4" and 3/8". I can stack them as needed or use them by themselves.

Someone had the same idea (I'm sure I thought of it first) and began to market them. I've seem them in brass and aluminum. I keep them in the front of my work apron as they get regular use. I mention them because they allow me repetitive cutting without the need to measure or mark. So with that thought firmly in your apparently very active mind....how about just making a marking tool....say a 1/2" thick piece of hardwood that is 1 1/2" in width (or whatever thickness your working with) and an appropriate length and cut a slit beginning at one end that is a centerline rip that straddles your blade...not the teeth. You'll probably have to cut it by hand so its a snug fit. Place your little marking gauge on the back side of the blade and slide your fence up to the side of gauge to effectively position the fence.

...and for the rest of you yahoos...I wax the top of my band saw table so a cold beer is noooo problem :D

Cheers,

Kent

[Jim Matthews]

Dumb question #1, did you joint/plane at least the bottom and side that rides the fence?

I'm not the only one thinking this, if it's a dumb question, there are at least two idiots in the room.

If you rip off the jointed face, the next face must be rejointed - and so on.
I prefer to keep the same jointed face and squared edge against the fence,
and advance them toward the blade with each pass.

With a properly adjusted fence, all cuts following the first will be parallel.

[John Poole]

After reading the reply about three equidistant points on a diagonal will always determine a point on the center line if the two end points intersect the parallel edges of a rectangle, I envisioned the following tool.

Using 5/16" stainless steel rod and a piece of purple heart, I created a marking device that pinpoints a point on the center line of a rectangular end of a board to be sawn into equal widths. I made my tool so it could accommodate a wide variety of thicknesses by placing holes 1", 2", and 3" from center. Simply tapping out the steel rods with a nail and then pounding them back in makes them snug, yet not permanent. I drilled a 5/16" hole for the rods. The 5/16ths hold in purple heart does not readily accommodate a 5/16th rod, you have to pound it in. I tried using a next-larger drill, but there was too much play. I drilled a 5/64ths hole in the center to accommodate a 1/16" stainless steel rod that has a sharpened point. (Turns out the pin is too loose and has too much play, I'll need to get a thicker rod and possibly widen the hole to accommodate a larger pin.)

Here's the layout, I had to offset from center the rod holes by 5/32" which is the radius of the rod; this makes the end of the rod represent the edge of the wood when firmly seated.

282654

After the drilling:
282653
The tool with rods inserted and a sample piece to be marked:
282656
Marking the sample. I'll take the pin and tap it into the end of the wood:
282655
I then take the diagonal off and the pin remains to serve to highlight where the pinpoint mark will be:
282658
Here's the pin point:
282657
Here's two pin points; as you can see the slop in the center produces varied results.
282659
Here's the wood placed against the fence and being aligned to the back edge of the saw. This exercise is to set the fence in the ideal position. (Yes, I'm getting a new fence, this stock one moves when you secure it.):
282660

(The rust on the table is from a roof leak, not spilled beer.)

[Andy Pratt]

The first thing I would mention is that you have to have a jointed face riding on your fence, and an edge that is jointed at 90 degrees to that riding on your table. You are probably already doing this, but if you aren't no amount of other advice is likely to get you where you want to be. If there is any twist left in the board, for example, you are going to see that in a thickness error after resawing. Also, if you start a cut slightly off the mark as a test, you are going to influence your blade to wander into that "scratch" when you later start the cut in the correct semi-overlapping location. When I do mine I just lock down the fence, hold the board against the blade with the motor off, and note where each side of each type (set) of tooth on the blade would be cutting and adjust until it is equal on each side. You can mark that with a marking knife if you really want total precision for the fence setting.


If your blade is an appropriate style for what you are resawing and is properly tensioned/tracking then I don't think you should be more than 1/16th off when resawing something in half. I rarely get one truly perfect, but if I am within 1/16th thickness for the two resulting parts I feel like I have done as good as necessary for my work (which is nearly always planed after).

[Myk Rian]

...and for the rest of you yahoos...I wax the top of my band saw table so a cold beer is noooo problem
But putting it on iron is a bad habit to get into.

[Rob Damon]

One of the problems with milled up wood these days that you buy is they never seem to be the same thicknes between two boards. For example I needed some plywood spacers for setting the height of drawer slides, so cut up twenty 1" wide strips x the length needed. When I stacked them up into two piles, one pile was 1/16" higher that the other and they all came from the same sheet of plywood.

So the first step would be to make sure the boards are actually all the same exact thickness the entire length (i.e. run them through a planer).

The next step I would cut off the last 4-6" of the board and cut up some small 1"x1"x thickness of board from that cut off. This gives you some trial/error test pieces.

Set the fence to what looks like the center of the small block and cut it in half. Set the two blocks beside each other and see which one is thicker.

Then take a Bessey E-Z style clamp (or equal) and clamp it to the fence rail as a stop block for the tee of the fence 1/32" of an inch in the direction the fence needs to move. This will limit the amount you can move the fence to incremental divisions.

Then cut another small block and check again. It shouldn't take more than a few cuts to dial it in to dead center of the blocks.

Then cut away. Make sure the board is fully support by the fence. Also if the boards are more than 6' long, I would add infeed and outfeed support, so that I am not trying to balance it and push it from the end. By using full infeed/outfeed support, you can push the board by standing right at the bandsaw table and use a push pad to hold the board against the fence at the infeed side of the blade. Also, by being right next to the blade you can see if the blade is starting to drift.

If you are just cutting a few boards is one thing but if you are cutting several dozen, I would take the time to get rid of the bandsaw blade drift. It can be done, but it does take patiences and time. It took me a half a day to get my table/fence/blade dead on parallel and I have now have zero blade drift on my 14" and 18" Bandsaws.

Rob