Mitre gauge position for a blade that tilts to the right

[rogerwilco]

I use an Incra Mitre Express on my RHS tilting blade. Since my blade tilts to the right, it's correct to have the sled on the left hand side. See pic. I'm also left-handed, so this suits me fine.
IMG_0755.jpeg

Now here's the problem:
I get sloppy cuts, and I've figured out it's because I'm not using an outboard support bracket (like this). When cutting large/heavy/long stock, there's a little bit of flex which produces inaccurate and inconsistent cuts. The support bracket will fix this, but the spacing of my mitre slots means that the bracket will clash with socket head screws on the fence that allow for the fence to be extended. You can see that in the pic.

One solution would be to get the larger fence extrusion from Incra. However, that's problematic because it will hang off the sled on the LHS (which is already overhanging the table). Yes, I could trim the fence to fit. But this seems like a compromise. When I need to cut long pieces, the fence and workpiece will be hanging in free air. It also jams up my walk way between benches - it's a small shop.

Another option could be to put the sled on the RHS of the blade. Now the blade tilts into the fence, but that seems easily solved by chopping the corner off at 45 degrees, and using a sacrificial fence when I'm cutting at 90. This seems a bit wrong... but I'm not sure it's really a problem, is it? Maybe there's a chance the blade could clash with flip stop? Maybe it complicates cross cutting very small parts?

Neither option is perfect. Whaddaya reckon?

I guess I could just get the longer fence, and experiment both ways... Loosing a few cms off a fence that big (685-1245mm extension) is no problem.

[rogerwilco]

I think I will put it on the RHS. Really, 95% of cuts are gonna be with blade at 90 degrees. It makes more sense to have the fence well supported by the cast iron table, and make use of the full length without having stuff overhanging. On those rare occasions that I want to do a beveled cross-cut, I can do that on the LHS with a regular mitre gauge. If I buy the longer incra fence for RHS use, I can attach the existing fence to the stock mitre gauge that came with the saw.

Done, and done.

[BobL]

Have you checked for slides slop and more importantly if they run parallel to the blade.
If there is even sub 1/10th of a mm slop or inaccuracies in the slides this causes havoc with sloppy and non-square cuts.
It's not enough to simply square the mitre to the blade because if the slides are not moving parallel it throw everything out.

I wrestled with mine for couple of years before I worked this out.

[rogerwilco]

Yeah, I’ve fiddled around with it quite a bit. And spent a long time checking and setting everything when I first got it. The slots are parallel to the blade. I’ve measured with a dial indicator, and satisfied it’s bang on. The rest of the mitre express assembly is dialled in too. Both the piece that runs in the mitre slot, and the piece that accepts the gauge have been adjusted and are practically zero play. It’s definitely flex in the fence. Most noticeable with heavy big stuff. I’m cutting 75mm thick workbench legs at the moment, and they’re coming off the saw with quite a bit of variability. If it were a slot alignment thing, and all else was equal, I’d expect the error to be the same on all cuts. That’s why I’m pretty sure this can be fixed with the support bracket.

[BobL]

Yeah, I’ve fiddled around with it quite a bit. And spent a long time checking and setting everything when I first got it. The slots are parallel to the blade. I’ve measured with a dial indicator, and satisfied it’s bang on. The rest of the mitre express assembly is dialled in too. Both the piece that runs in the mitre slot, and the piece that accepts the gauge have been adjusted and are practically zero play. It’s definitely flex in the fence. Most noticeable with heavy big stuff. I’m cutting 75mm thick workbench legs at the moment, and they’re coming off the saw with quite a bit of variability. If it were a slot alignment thing, and all else was equal, I’d expect the error to be the same on all cuts. That’s why I’m pretty sure this can be fixed with the support bracket.

Slot alignment should (but not always) give consistent cuts but fence flex or mitre slop will produce all sorts of variability
Slop and flex are a right royal PITA.
After farting around with my cheap contractor TS for ages I went and bought a decent SCMS.
The SCMC is still not perfect but much better than my TS and the longer the pieces the better the SCMS is.
If I want something large cut square I now take it around to my mates workshop where he has a 14" Altendorf Sliding panel Saw.

[rogerwilco]

I know the hassles you’re describing pretty well!
I’m quite confident the support bracket is gonna solve it for me, or at least, bring it to acceptable tolerance. I have an okay SCMS (hitachi), but I’m in the process of eliminating it. It’s a bench hog, sprays dust everywhere, and is only accurate enough for rough cuts. I’ve been rough cutting with a jigsaw lately, and find that I almost never use SCMS. I will probably keep it… but store it on a shelf and only get it out when I’ve got a lot of parts to cut to rough size.

I do have access to an Altendorf slider at Melbourne Guild of Fine Woodworking. But I can’t be running off there every time I need a fairly accurate crosscut!

[BobL]

I know the hassles you’re describing pretty well!
I’m quite confident the support bracket is gonna solve it for me, or at least, bring it to acceptable tolerance.

If it reduces fence flex then it should definitely help.

I have an okay SCMS (hitachi), but I’m in the process of eliminating it. It’s a bench hog, sprays dust everywhere, and is only accurate enough for rough cuts. I’ve been rough cutting with a jigsaw lately, and find that I almost never use SCMS. I will probably keep it… but store it on a shelf and only get it out when I’ve got a lot of parts to cut to rough size.

Your SCMS problems sound familiar but I think I have most of them covered although I realize that these approaches may not suit everyone.
The 12" Bosch glide saw I have has no rails so it takes up less space than regular SCMS. However it does weigh 32 kg!, so putting it on and off a shelf (I don't have any spare bench space) is basically a no-no. So I've got mine on a pump up trolley under the left hand wing of my TS so getting it out and useable is quick and easy. For a couple of cuts my old shop vac picks up most of the chips and I have a large ventilator fan directly above the TS/SCMS that easily takes care of the fine dust. If I'm doing lots of cuts I wheel it outside. It does 104 x 341 mm cuts and they're definitely not "rough", for example it cuts standard picture frame mitres consistently and close enough for my liking.

I do have access to an Altendorf slider at Melbourne Guild of Fine Woodworking. But I can’t be running off there every time I need a fairly accurate crosscut!

My mate's saw is 15 minutes away but I only use it when breaking up large sheets. I've done him quite a few favours so even if he is using it he usually prepared to stop and let me cut up my stuff.

[rogerwilco]

I installed the bigger fence, and attached the support bracket. It didn't solve the problem. The whole thing has been moved to RHS of blade, and I've gone right through the calibration.

Checked:
- blade is square to table
- mitre slots on table are parallel to blade
- mitre gauge fence is square to base
- mitre gauge fence is square to blade
- no slop/play in the mitre express

But no dice. I'm getting similar error.

Here's the alignment of mitre gauge fence to blade.
IMG_0762.jpeg

And here's the error after the cut.
IMG_0761.jpeg

! Obviously something is askew. Some deflection maybe.
I realise I'm asking a lot from a sled that only rides in one slot, on one side of the blade, and cutting big pieces.

I guess if I want precision for such cuts, maybe I gotta use a slider or a Kapex...

Or just accept the error.

[Camelot]

You should not be having such a large error on a small piece of timber, is that wooden try square actually square, often they are not. You need a roofing square or a sheet with a known 90 degree angle (again check the roofing square is true by doing a mirror image square tolerance test) raise your blade fully up and put a straight edge on it (which is at least as long as your table) making sure it misses the teeth, this will increase the reference line of your saw blade in relation to the slot in your table, measure to see if they are parallel. Then also check with the roofing square if the front edge of the table is at 90 degrees to your blade (not that important if your slot is guiding your sledge, but nice to know). If both these are true then the problem is with the timber which is running in the slot, the best way of fixing this without error is to make it over length from your sledge, this will allow you to make sure it is tight up to one side of your slot (the same side) before fixing your sledge to it.

[rogerwilco]

Is that wooden try square actually square, often they are not.

Yes. Confirmed with a trusted engineer's square, and the mirror image test.

If both these are true then the problem is with the timber which is running in the slot

I think this is the most likely cause. Incra Mitre Express has an aluminium extrusion that runs in the mitre slot. It has 4 or 5 expansion points that are adjusted with small cap head machine screws. The mechanism is okay. You can adjust each screw 'til it grabs or catches, and then back it off for snug by free-moving fit. This would seem to me to be the weakest link in the setup. I'll fiddle around with it some more.

[BobL]

A picture of a typical piece of wood being cut might be diagnostic.

Also a description of how you measure the parallel-ness of the mitre slots to the blade might give us a clue.

[barri]

Checked:
- blade is square to table
- mitre slots on table are parallel to blade
- mitre gauge fence is square to base
- mitre gauge fence is square to blade
- no slop/play in the mitre express

The one check you haven't made, is the fence 90 deg to the slots? Just because a fence is 90m deg to the blade doesn't mean that the fence is 90 deg to the slots unless the blade is 100% parallel to the slots and very few are. The blade is short but the slots are long. Is the blade 100% flat? Referencing off the blade doesn't necessarily indicate true 90 deg.
This video explains it better than I can.
Squaring a Miter Gauge Correctly - YouTube

The fence is moving via the slots so it makes sense to me that the fence should be 90 deg to that movement. I also have a mitre express and many other sleds and I always check that the fence is 90 deg to the slots by using an accurate square and a dial indicator.
I have always followed and swear by this method ...... Table Saw Sled Alignment - YouTube. In fact I no longer use the 5 cut method. This is quick, easy and accurate and there are no cuts needed other than one final verification cut.

[rogerwilco]

Don't laugh, everyone. The error was actually me.

To test my calibration changes, I was trimming the end off a 90x75mm leg. The edge that references the fence had a small bit of snipe from the thickness planer. After the cut, I was pushing my square into the corner, not realising that it was referencing this tiny bit of snipe, and not the actual edge of the leg. Eventually, I woke up to myself, and realised that the error was different if I measured in different positions across the thickness of the piece. Flipping the square to the other side showed the cut was square to that edge. So I flipped the piece and use the other edge to reference the fence, ran a few more test cuts, paid attention to my measurement technique, (lateral pressure on the butt of the square against the edge of the piece).

100% deadnuts accurate.

A little embarrassing. But also a relief.

I had completely tore the gauge down, recalibrated everything, and tested it with and without the sled. It was the same.
Then I tested on a small piece, and it was perfectly square (because there was no snipe messing up my measurement).
Based on that, I'd ruled out the sled or mitre gauge being the issue. For a moment, I thought there might be an issue with blade deflection under a fairly deep cut (75mm Vic Ash).
Glad I figured out it out before I started pulling the machine apart!

[BobL]

I'm pleased to hear you solved it.

I found the photo of the item on my mitre slide that gave me most grief. I just have the stock mitre slide that came with the contractor saw some 17 years ago and it had developed a sloppy slide which i repaired by adding the usual asymmetrical rotary adjusters like you see on most upper end mitre slides but I still had fence movement.

Eventually I tracked the problem to slop in the main screw shoulder on the screw that holds the mitre half circle onto the slide. I forget the exact difference between the size of the hole and screw shoulder, I think it was about 0.15mm but that was enough to make a significant difference at the end of the fence. I ended up using am M6 socket screw that had a fat head so had enough metal on to able to turn a tight fitting shoulder on the MW lathe.

Nitreslidescrew.jpg