Thread: How to do accurate miter joints

In my experience mitres tend to open up once cut and exposed to moisture in the air. I see you’re in QLD which wouldn’t help. I haven’t tried it, but cutting the mitres at slightly less than 45 degrees might help.

Build a mitre shooting board out of some scrap and plane them. Alternatively if you mark the cutting lines with a knife you can just plane up to them.

HERE is a thread I started a few weeks ago on shooting panels to size; the principle is the same. I did use a decent low angle jack plane but pretty much any plane will do providing the blade is SHARP.

Its not just getting the angles right. Opposite sides need to be exactly the same length. I made a simple tablesaw sled and have the blade canted to 45*. Stop blocks ensure the length of opposite sides are exactly the same. I have a miter shooting board but using that I can not always get the lengths spot on. Stop blocks may sound like training wheels but with miters I dont mind all the help I can get.
Regards
John

I do my mitres on the disc sander. Mark the mitres with a marking knife on the outsides. Mark the opposite sides together, so they are exactly the same. Set up a fence on the table of the sanding disc (or belt sander) at about 45 and sand two test pieces. Check if they form 90 when you put them together. If they don't, adjust the fence. When the angle is correct, sand your mitres. As you get to the cut a small feather of wood will come away. When that happens, the mitres are correct.
As Jack says, mitres on boards more than about 50mm will inevitably open up. To prevent this, and for strength in any picture frames, I put in slipfeathers (splines).

Thanks for the wisdom Gents.

I went back to the jig and tried again, making two small changes. First was supporting the track saw track mid span as I suspected that my pushing the saw may have been distorting the track a smidge.

The other thing was to revisit my clamping of the timber in the jig. I was using one clamp to hold the timber against the fence of the jig but this introduced a small gap between the timber and the jig base on the side opposite the fence. Second time around I lightly clamped the timber to the jig base, then clamped the timber to the jig fence.

I've then trimmed 1mm or so off each of the 8 miters and now the fit is >99%. Very happy with the fit and I've only lost maybe 3mm off the total width and height.

Lessoned learned here was to measure and check everything 10 times and not to take things for granted, then cut once.

Cheers.

20230922_091405_resized_small.jpg

Originally Posted by orraloon

... Opposite sides need to be exactly the same length. ... Stop blocks may sound like training wheels but with miters I dont mind all the help I can get.
Regards
John

I disagree profoundly with John.

Stop blocks in this application are not training wheels - I regard them as essential. Just watch a professional picture framer at work. They almost always use an expensive guilotine, always with an integral stop block.

There are "stop blocks" that sort of work, and there are decent table saw fence stops that are an integral part of the cutting system that also increase accuracy and efficiency.

Measuring in any system is fraught with potential errors. Fence stops greatly improve measurement & hence cutting accuracy & precision IF they are well designed.

Most wood workers rely upon the measurement of long point to long point, or short to short point, for each side of the picture frame assembly to control cutting & assembly then the diagonals to check.

Another point to consider is what are the "critical dimensions?" Is it the external face or the internal face of the 'frame'? All measurement, setting out, & cutting procedures are then designed to achieve the critical dimension/s.

For a traditional square / rectangular picture frame we require the opposing internal faces (sides) of the frame to be truly parallel and for the diagonals to be equal i.e. the face to face (internal) measurements to be the same to be parallel & square. Accurately marking, and identifying the corners, internal or external, then cutting to those marks has its problems.

Back in the late 1970's & early 1980's my Dad was a truss and wall frame designer / estimator for two truss plants in Cairns, in the days before computerization of cutting lists etc. Each element in the truss was determined by a "center-line" for the chords & webs etc. with offsets (not necessarily equal) from the CL to each face of the stock. The Center-Line length was initially computed with all measurements, cutting lists and jig setup referenced to the CL length. For each truss element "approximate" external dimensions were then calculated by long hand. Being a surveyor with extensive hand held calculator programming experience I wrote a series of programs for Dad using Texas Instruments scientific programmable calculators to reduce calculation effort and eliminate / minimize potential errors.

Using the CL calculation and set-out methodology also overcame errors due to variable stock dimensions (which still had to be within acceptable dimensional tolerances for width & thickness) and ensured the faces of the "cuts" were accurate and that the internal faces of significant elements such as top & bottom chords became secondary "references" for setting up.

So the short of it is use a fence stop that is preferably set up so the long point of the stock is against the fence and that it is accurately cut to the desired complimentary angle to match the "mitre angle" and relieve (chamfer) the internal corners (of the fence stop block) so that any furring of the stock or debris does not affect the stock cut accuracy. The cuts then become mitre face to face cuts not point to point cuts, eliminating a lot of potential for errors. If the stock is dimensioned accurately then everything should fit nicely. Small variations in stock dimensions or small amounts of cut debris trapped between stop, fence & stock can significantly reduce cutting accuracy & precision.

I'm with Alex. The disc sander creates perfect mitres quickly and repeatedly. Everyone has their own preference, but I've never had a failure with this method using a 12" disc sander and a shop-made 45 degree jig. A chute plane is also very good and a well calibrated mitre sled for the table saw also does a good job so as usual it's just a matter of preference.

Best regards,

Brian

Originally Posted by AlexS

I do my mitres on the disc sander. Mark the mitres with a marking knife on the outsides. Mark the opposite sides together, so they are exactly the same. Set up a fence on the table of the sanding disc (or belt sander) at about 45 and sand two test pieces. Check if they form 90 when you put them together. If they don't, adjust the fence. When the angle is correct, sand your mitres. As you get to the cut a small feather of wood will come away. When that happens, the mitres are correct.

Thanks for that advice. I must try it.

I haven't thought of disc sanders as a precision tool until now, but it probably makes sense when sometimes I've cleaned up scribed joints on architrave and skirts with a 5 inch grinder (not usually regarded as a precision tool) and 1mm disc and it can work better than a jigsaw on awkward tiny cuts and angles.

Yes, I know that a coping saw in competent hands can do the same and I've seen it done as well or better and quicker by competent hands, but it's not something I do very often and my hands aren't that competent.

If there's one thing I've learned as a woodworker, it's that there's rarely only one way to do something. If whhat you're doing doesn't work for you (and your equipment) talk to other people and try what they're doing.
If what you're doing works for you, it's the right way.

Interesting read.
Off topic now but...
In reply to Mobyturns
As I wandered through the post I thought "I'd know how I'd tackle that" only to see another surveyor has same thinking.
I'll bet that was very well received.
Just had a builder build an extension and he was taking ages to get angles etc worked out and I quickly knocked them up.
But he persisted his way!

Re trimming final cut.
I don't have disc sander but do same on an old Paulcall belt sander with "micro" adjustable fence.
That gives excellent results.

Originally Posted by Avondale

As I wandered through the post I thought "I'd know how I'd tackle that" only to see another surveyor has same thinking. ...

Yep, I often react the same way.

And sometimes I see three different solutions and think: "I've tried them all ... and next time I'll try a fourth method."

A little late but might help.

A post from 2015, the mitres haven’t moved at all and still tight, no gaps after going through some “trying” weather conditions we have experienced over the last few years.

Turkish Tiles WIP - Finished

I get perfect mitres every time on the tablesaw using a feeler gauge to adjust the angle, like this:

IMG_E0222.jpg IMG_E0223.jpg
IMG_E0224.jpg

As always there's more than one way to skin a cat in woodworking...