Improve the Accuracy of Your Miters

[dusty]

When I bought my Incra V120 I sorta quit worrying about miter accuracy. The V120 does a very good job at this task with all of its preset angles.

But guess what. It might be that it can be even more accurate. It may need to be "fine tuned". I might have known this if I had read all of the instructions but, well .....

I happened upon this YouTube video and it got me to looking. Sure enough, there is a way to "fine tune" a V120. And, sure enough, it is mentioned in the instruction sheet.

I have not tried this yet but I will.

[mountainbreeze]

Looks like a good way to test a miter gauge. He said the error was 1/6 th of the resulting gap. Wouldn't that be 1/12 th?

[Ed in Tampa]

Interesting but as I see it this way there is actually 12 places to muff the cut.

For the error he ended up with all it would take is one board not to have been pushed tightly against the mitre fence, or slightly off the stop block.

When you think about it that amount of error could be nothing more than he didn't use the same pressure to hold the board against the mitre fence. Or the same force to push it against the stop block.

I think I like the one where you use a board and cut it, turn it, cut it turn it and until you get the whole way around and then cut a slice and measure it for parallel. That way there is only 5 places to muff the cut.

[dusty]

Looks like a good way to test a miter gauge. He said the error was 1/6 th of the resulting gap. Wouldn't that be 1/12 th?

Hmmm. Maybe. I gotta think on this.

[dusty]

Interesting but as I see it this way there is actually 12 places to muff the cut.

For the error he ended up with all it would take is one board not to have been pushed tightly against the mitre fence, or slightly off the stop block.

When you think about it that amount of error could be nothing more than he didn't use the same pressure to hold the board against the mitre fence. Or the same force to push it against the stop block.

I think I like the one where you use a board and cut it, turn it, cut it turn it and until you get the whole way around and then cut a slice and measure it for parallel. That way there is only 5 places to muff the cut.

Isn't that really the same principle? The error accumulates with each cut.

[JPG]

I gotta agree with Ed on this! My reaction was the same! I am really concerned that the 'error' may be due to variation in the lengths. The pointy end is not conducive to repeatability.

And yes 1/12 is correct.

I still like the cut and stand on end method to check for 90. Yes the side edges must be parallel. Only one cut to screw up!:D

Or, cut two and eliminate the need for parallel. Set on end and check gap between edges set against miter gauge when cut. Longer the better(need to be straight:rolleyes:).

Notice there is always 'something' that can contaminate the results.

[Ed in Tampa]

Isn't that really the same principle? The error accumulates with each cut.

Yes but not as many chances for operator error. The more cuts the operator has to make the more chance he has that he will introduce a error rather than the mitre fence.

[dusty]

Yes but not as many chances for operator error. The more cuts the operator has to make the more chance he has that he will introduce a error rather than the mitre fence.

I live in that world but not when I am attempting to perform or proof an alignment.

[JPG]

I live in that world but not when I am attempting to perform or proof an alignment.

Same world! Error is still possible though diligence will minimize the occurrences(assuming the 'operator' has control over them;)).

Using points as a reference 'surface' is IMHO very error prone.

[forrestb]

I am sure that Incra provided WWGOA all the help to produce this method & video for checking their miter gauge.

But before anyone runs off believing that George has really accurate advice, remember he is the same person who told us how to verify that your bandsaw table was perpendicular to the blade.

Caution is advised. (Well, do so with my advice also. Verify, verify, verify is my advice)

Forrest