What constitutes a 90 / 45 / 22.5 degrees cut ?

alecail

Member
Joined
Feb 10, 2024
Messages
2
Hi,
I'm about to receive a TS 60 KEBQ with a Festool rail.
Obviously, I am buying this kind of saw for its accuracy.

In order to not be disappointed, can you suggest a method to check if the saw actually cuts  at 90 / 45 / 22.5 ?
Right now with my current circular saw (not even a plunge saw / track saw), if I lay a board on the freshly cut edge, I have a deviation of about 3 mm at 400mm, using the distance from a square that lays on the reference surface.
I know there is a lot to unpack here, like... is it really necessary to have an exact 90 degrees, knowing that if the deviation is constant, I could flip the next board to cancel it.. 
This is in the context of furniture building. I don't have a jointer.

 
With a good tri-square (I have a Starrett, but others make good ones too) you can directly measure your cuts to check for accuracy.

For the 22.5” cut, a less direct method is to make two cuts with that angle.  It should match up with the 45 degree angle.

Every shop should have a decent tri-square in my opinion.


You can get a good tri-square for about $100.00, and it will prove useful in your shop.

Starrett and others produce an adjustable protractor for this purpose.  I don’t have one, so I cannot comment.
 
A 3mm deviation over 400 mm would be equivalent to a 0.135mm over 18mm

Are you suggesting that I check for a 0.1mm deviation by putting the square against the edge ?
 
If you want to check the 45 without any measuring, cut a 12” wide strip of a sheet of 3/4” plywood with 45 degree (indicated) on both edges of the 12” width.

Then cut four pieces about 3” wide from that strip. 

Then lay the pieces down on a table with all the 45 degree cuts touching and tape the ends together. 

Finally, fold the pieces up into a square. 

If all the joints are tight on the inside and the outside, your cuts are right on the mark. 

You can check other angles taping them into triangles, pentagons, hexagons, etc.

I don’t remember all the angles for all the geometric shapes.  I knew them back in 1964, but it escapes me now.  But luckily Google comes to the rescue:
https://www.cuemath.com/geometry/interior-angles/
 
This is what [member=74278]Packard[/member] is saying, and it is the most accurate method. You can "measure" angles all day, but if they don't fit right, it doesn't matter.
I do use a digital angle gauge or a tilt box, depending on the situation, but taping up test parts is always best.
As far as "canceling" angles...yes. that does work, even jointer guys do it. You just can't be too far off though or the parts will slide around when glue gets applied. A whole lot of dry-fit clamp-ups don't go as well as hoped because of the glue during final assembly.
 

Attachments

  • miterfold3.jpg
    miterfold3.jpg
    157 KB · Views: 83
Crazyraceguy said:
I do use a digital angle gauge or a tilt box, depending on the situation, but taping up test parts is always best.

I'm constantly surprised at how much confidence people place in something because it's digital. Those digital angle gauges are typically accurate only to +/- 0.2º. That means a hexagon (12 cuts) could have a 2.4º gap - very noticeable. For some reason CMT claims their digital angle gauge accuracy is +/-0.1º, but I don't know what CMT would have done that others haven't done, especially since these are all Chinese manufactured.

Go ahead, take a look at the specs.

You can get a Stabila level that has better accuracy, but only at horizontal and vertical. In between it's much less accurate.
 
smorgasbord said:
Crazyraceguy said:
I do use a digital angle gauge or a tilt box, depending on the situation, but taping up test parts is always best.

I'm constantly surprised at how much confidence people place in something because it's digital. Those digital angle gauges are typically accurate only to +/- 0.2º. That means a hexagon (12 cuts) could have a 2.4º gap - very noticeable. For some reason CMT claims their digital angle gauge accuracy is +/-0.1º, but I don't know what CMT would have done that others haven't done, especially since these are all Chinese manufactured.

Go ahead, take a look at the specs.

You can get a Stabila level that has better accuracy, but only at horizontal and vertical. In between it's much less accurate.

I fall trap to that too, and I know better.

My speedometer has an analogue needle readout, and also a digital readout.  Both are working from the same data from the same sending unit.  But still, in the back of my mind, the digital is more accurate.
 
Packard said:
But still, in the back of my mind, the digital is more accurate.

But not as accurate as the police officer’s radar speed detector which he’s pointing at you  [crying]
 
smorgasbord said:
I'm constantly surprised at how much confidence people place in something because it's digital. Those digital angle gauges are typically accurate only to +/- 0.2º. That means a hexagon (12 cuts) could have a 2.4º gap - very noticeable. For some reason CMT claims their digital angle gauge accuracy is +/-0.1º, but I don't know what CMT would have done that others haven't done, especially since these are all Chinese manufactured.
...
Not sure about CMT, but it is possible to make a very, very accurate angle measuring utensil. The "small" problem is that one actually needs find customers willing to pay for it.

My local maker offers digital angle gauges for $40 (Chinese resell) and $200 (their own chassis with high-end sensors) ... both look very similar on the outside. Guess which one sells.

Then comes the practicality of it all - 0.2° is about 0.02 mm/0.001" (!) max error over the under 100 mm/4" reference surface of the device.
At such a precision, flying piece of dust stuck between it and the blade affects the result. Not to mention the unevenness of the blade, the saw bearing micro-wobble, etc., etc.
 
alecail said:
I'm about to receive a TS 60 KEBQ with a Festool rail.
1. They’re pretty accurate out of the box, or at least my TS 55 KEB was. In any case, everything is adjustable. In particular, you need to make sure that the saw is dialled-in for the track that you are using. Peter Parfitt made a video on making adjustments to the TS 55 — not the same as your saw, but the principles are the same:


2. When making angled cuts, the saw can get tippy on the track and try to fall over. It’s key that the workpiece and track are well clamped and you apply firm pressure on the track side of the saw’s plate to make sure that it stays perfectly flat throughout the cut. Sounds difficult, but it really isn’t. There are 3rd-party 3D printed gizmos to keep the saw on the track, but they cause more problems than they solve, IMO.

 
alecail said:
A 3mm deviation over 400 mm would be equivalent to a 0.135mm over 18mm

Are you suggesting that I check for a 0.1mm deviation by putting the square against the edge ?

Yes.

Holding a known square tool against an edge is how this has been done for several thousand years. No reason to reinvent the wheel. .01mm is about the thickness of a piece of paper. Held up to a light with your eye aligned directly down the line, that is easily seen. If not, you need an optometrist, not another tool.

Now another issue is the resolution you're trying to hold and the tool you're using. Move 100mm down the cut and measure again. I'm reasonably confident you'll get a different result. A track saw will be significantly better, but still barely comparable to either a table saw or jointer. I'm not saying you can't build furniture without a table saw and jointer, but if that's your plan I suggest you get a couple hand planes and learn how to use them to create straight, square and glue-ready lines to follow up your circular saw cuts.

And make a shooting board, too.
 
woodbutcherbower said:
But not as accurate as the police officer’s radar speed detector which he’s pointing at you  [crying]

That statement brings back memories.  [smile]  [off topic]

There was a time when the speedometer drive cable was connected to a replaceable speedometer gear on the tail shaft housing of the transmission and driven by the transmission output shaft. If you changed the original differential ring & pinion ratio or if you even substituted larger/smaller diameter tires, the indicated speedometer speed would be incorrect and you'd have to change the speedometer gear to get the correct reading.

It was such a commonly accepted occurrence that even the judges/courts would allow you some wiggle room if you showed up in court for a minor speeding ticket and claimed that the "studded winter tires" were still on the car. I'm certain you'd not fare as well nowadays with the current encoder technology.  [crying]
 
woodbutcherbower said:
Packard said:
But still, in the back of my mind, the digital is more accurate.

But not as accurate as the police officer’s radar speed detector which he’s pointing at you  [crying]

But beware of the cop with a tuning fork.  Back in the early 1970s I had a Mustang Mach I in bright red.  It looked like it was speeding even while parked.

I got a radar ticket for 55 mph in a 35 mph zone.  I was certain I was not going over 45.

The cop showed me the gun’s readout.  It said “55”.

I told that story to a friend who was a cop and he laughed.  Apparently they calibrate the radar units with a tuning fork that reads out at 55 mph.  I wasn’t speeding; the tuning for k was.
 
Cheese said:
woodbutcherbower said:
But not as accurate as the police officer’s radar speed detector which he’s pointing at you  [crying]

That statement brings back memories.  [smile]  [off topic]

There was a time when the speedometer drive cable was connected to a replaceable speedometer gear on the tail shaft housing of the transmission and driven by the transmission output shaft. If you changed the original differential ring & pinion ratio or if you even substituted larger/smaller diameter tires, the indicated speedometer speed would be incorrect and you'd have to change the speedometer gear to get the correct reading.

It was such a commonly accepted occurrence that even the judges/courts would allow you some wiggle room if you showed up in court for a minor speeding ticket and claimed that the "studded winter tires" were still on the car. I'm certain you'd not fare as well nowadays with the current encoder technology.  [crying]

My buddy, the cop, said on Long Island, NY the cops allowed you 5 mph over the limit plus 2 mph to account for speedometer inaccuracies.

Upstate, where I now live, that number is (unofficially) 10 mph + 2 mph.

I would note that every shop should have a “reference gage”.  That is a known accurate angle gage.  For me that is the Starrett combination square. 

But there are gage blocks that are probably more reliable over the years.

Woodpeckers sells some, also Starrett and a few Japanese manufacturers.  I would not choose a Chinese version (though a made in Taiwan version is probably OK).  These blocks rarely go out of tolerance, though you can damage the faces.  They are ground so smooth that the blocks stick together as though they were magnetized. The Woodpeckers version appears to have a matte finish—that would cripple the “stick together” property that genuine angle blocks have.  A deficiency in my opinion.


And here is a demonstration on how the ultra-flat and ultra-smooth block gages hold together on a vacuum-like property.
 
smorgasbord said:
Crazyraceguy said:
I do use a digital angle gauge or a tilt box, depending on the situation, but taping up test parts is always best.

I'm constantly surprised at how much confidence people place in something because it's digital. Those digital angle gauges are typically accurate only to +/- 0.2º. That means a hexagon (12 cuts) could have a 2.4º gap - very noticeable. For some reason CMT claims their digital angle gauge accuracy is +/-0.1º, but I don't know what CMT would have done that others haven't done, especially since these are all Chinese manufactured.

Go ahead, take a look at the specs.

You can get a Stabila level that has better accuracy, but only at horizontal and vertical. In between it's much less accurate.

That's exactly my point. I don't trust it. I use them as a starting point, since they seem closer than the needle/arrow on the saw itself.

I also use a Bosch laser measuring device for the huge things I build on a regular basis. I didn't trust it at first ether and checked with a regular tape measure, when possible. Often that takes another pair of hands, which I don't always have around.
It's all relative anyway. Someone else measured the space where it goes.....with who knows what
 

Attachments

  • 2013-12-31 15.31.26[747].jpg
    2013-12-31 15.31.26[747].jpg
    136.2 KB · Views: 44
  • Crcl3.jpg
    Crcl3.jpg
    143.8 KB · Views: 43
Packard said:
woodbutcherbower said:
Packard said:
But still, in the back of my mind, the digital is more accurate.

But not as accurate as the police officer’s radar speed detector which he’s pointing at you  [crying]

But beware of the cop with a tuning fork.  Back in the early 1970s I had a Mustang Mach I in bright red.  It looked like it was speeding even while parked.

I got a radar ticket for 55 mph in a 35 mph zone.  I was certain I was not going over 45.

The cop showed me the gun’s readout.  It said “55”.

I told that story to a friend who was a cop and he laughed.  Apparently they calibrate the radar units with a tuning fork that reads out at 55 mph.  I wasn’t speeding; the tuning for k was.

My issue with any of that radar stuff, is that there is no way to definitively prove that the speed on the gun is actually you. All a dishonest cop needs is an actual speeder, a reading of it, and the accusation that he was pointing it at you. Two or three tickets from the same reading? Sure. But the courts always believe the cop. When it comes to a "he said-she said" situation, with no other evidence, the cop always wins. Back in the late 80s, I got a speeding ticket in an out-of-state situation, in the middle of the night. I had my cruise control on at exactly the speed limit of 55. I was pulled over and given a ticket for 72! There is 100% no chance in the world that this was true, but 400 miles from home, on a Friday, they knew I would just pay it by mail and let it go. Fighting a losing battle would have cost more, even if I did "win"

Somewhere, years ago, the thinking was that they had nothing to gain, but you had the potential to lose, so you were more likely to lie. Promotions/awards given for higher numbers and of course the infamous "Brady List" have shown this not to be the case, but it still happens.
 
Crazyraceguy said:
My issue with any of that radar stuff, is that there is no way to definitively prove that the speed on the gun is actually you. All a dishonest cop needs is an actual speeder, a reading of it, and the accusation that he was pointing it at you. Two or three tickets from the same reading? Sure. But the courts always believe the cop. When it comes to a "he said-she said" situation, with no other evidence, the cop always wins.
...
Over here the police must provide a photo from an (independently) calibrated and sealed radar-and-photo appliance. Plus any metrologic uncertainties (usually 3%) are automatically detracted from the measured speed for any legal purposes.

Were the cops not have a photo proof, or had it but not from a calibrated device with a valid calibration mark, the case would be thrown immediately by the magistrate.

The courts do accept "he said she said" for cases that are considered as being possible to "freely assess" by a human. In theory. Like when someone endagers someone etc.
But even there the Cop story must be consistent with other evidence or the accused caught of confabulating (seen as evidence of malice). Otherwise a pure he-said-she-said is automatically thrown out on grounds of "inconclusive" which automatically goes for the dedendant. Even if the magistrate would not throw it out, appeals court will. As a result it rarely happens .. the cops know they would not get any traction.

That said, for such a cautious setup to settle it took a couple centuries of cops being, in effect, the armed arm of a totalitarian government. Be it the monarchy or the commies more lately. It goes a long way to prevent cops abusing their authority .. especially goes for the speedcam boys who better get their stuff right or there will be no bonuses from successful prosecutions for them ..
[smile]
 
The “he said—she said” can easily be defeated by a GPS enabled dash cam.  Indeed, Consumers Reports said such an item was useful for accidents (they did not say it was useful for dealing with cops that lie.)

My GPS camera records the geo-location (longitude/latitude) and speed along with images.  Hard to defeat its data.  And you don’t have to produce it if doing so convicts you.  [big grin]
 
alecail said:
A 3mm deviation over 400 mm would be equivalent to a 0.135mm over 18mm

Are you suggesting that I check for a 0.1mm deviation by putting the square against the edge ?

A 0.1 mm deviation is rather large (.004") so it's readily seen when back lit.

Here's what Jeff & Packard were referring to although you just measure a single board rather than a stack of them. I wanted to know what the variation in cut length was by measuring and cutting each board individually instead of using a hard stop to cut all of them.

At the green arrow location, look for light between the square and the board. It will be difficult to measure any discrepancy but it will be easy to see any discrepancy. 

[attachimg=1]

Here's another example, I was checking the accuracy of the 45º detented position on a Woodpeckers Posi Lock T-Square.

1. Place a bright light source under the area you want to check.

[attachimg=2]

2. Place a known ACCURATE measuring instrument between the angle you wish to measure.

[attachimg=3]

3. Gently align the edges of the measuring instrument and the part you want to measure and look for any leakage of light. If there's a discrepancy it'll be easy to see. [smile]

[attachimg=4]
 

Attachments

  • 9274 marked.JPG
    9274 marked.JPG
    1 MB · Views: 561
  • 10342.JPG
    10342.JPG
    851.9 KB · Views: 587
  • 10336 marked.jpeg
    10336 marked.jpeg
    930.1 KB · Views: 582
  • 10337 marked.jpeg
    10337 marked.jpeg
    1,001.6 KB · Views: 558
[member=61254]mino[/member] that seems like the logical step to avoid this problem. If there is anything like that here, I have never seen/heard of it.

Packard said:
The “he saidshe said” can easily be defeated by a GPS enabled dash cam.  Indeed, Consumers Reports said such an item was useful for accidents (they did not say it was useful for dealing with cops hat lie.)

My GPS camera records the geo-location (longitude/latitude) and speed along with images.  Hard to defeat its data.  And you don’t have to produce it if doing so convicts you.  [big grin]

I think that actually save a lot of people in recent times. The cop starts that accusation, the driver points to the camera, and "suddenly" it's just a warning, and "watch your speed".

That would not have been possible in 1989.  [big grin] He probably only had what would be a very crude version of radar back then too.

Yeah, [member=44099]Cheese[/member] that what I do for my jointer fence, generally with a 1 2 3 block. It's pretty amazing how accurate that can be.
 
Back
Top