Okay I just went back to the shop and ran some measurements, Iāll do some futher test cuts later today with interim measurements (measure after the clerance path)
I made measurements of three different boards, one was cut with a router table using a 1/4" clearance and the matchfit bit using the matchfit jig to set heigh, the other two were cut with the altmill, one the test board, one the spoilboard, both clearanced and then cut with the matchfit bit. Mesaurements are the width at the very top (narrowest width) and the depth meaured with the caliper sitting on top of the board and putting the depth gauge in
Router table board: width is 0.344" and depth is 0.3675" the matchfit bit is a fairly snug fit in this board, the bottom of the dovetail (widest bit) is around 0.46"
Test board: width is 0.3675" and depth is 0.3675" and the matchfit bit isnāt as snug in this
Wasteboard: width is 0.3690" and depth is 0.376" and the matchfit bit isnāt as snug as the router table, the bottom of the dovetail is 0.4925 much wider
Iāll run a test cut with first a single pass of the 1/4" bit and the double pass to depth with it and see if the width of those cuts change, then Iāll run the matchfit again, I canāt figure out how I might be getting wobble in the matchfit bit in the spindle itās a 1/2" shaft bit and was snugly in the spindle. I feel the problem has to be with the matchfit bit as the 1/4 couldnāt make the dovetail part at allā¦
How can the matchfit bit be an issue? I am assuming you use the same bit in the router table as well as the Altmill.
About the only thing left is that you have runout in the spindle or collet. On the assumption that you are using a 1/4" collet for the 1/4" bit, if the resulting slot is 1/4" then the spindle doesnāt have excessive runout. I would then take some straight bit with a 1/2" shaft, mount it in the 1/2" collet and run a test pass and measure it.
Make sure the 1/2" collet is clean on the inside. Donāt use tape to hold down your test pieces as there could be some sideways shifting.
I just finished up the tests then came back here and read your message. Got some interesting data, the relief cut was supposed to be 0.375 from the top of the material, measuring with my calipers itās 0.388 so the Z Axis is a touch off (the tool pocket cut was supposed to be 0.38 and itās at 0.4075) , the slot was 0.26" for a 0.25" bit so not bad
The I stuck the 1/2" Collet in and put in the matchfit dovetail bit, all seemed good, snugged it up and then spun the spindle up and boom, I can see a wobble in it. I pulled out the collet and cleaned it up, same, then I took a look at hte collet nut and it had a clump of oil in dust in it which I cleaned out, then I spun up the matchfit and it spun nice and straight, ran the sample dovetail slot, and boom itās all like it outta be, mystery solved!
As a follow-up on this, I was kind of bummed that this happened after all my test cuts, but at least I learned to always check that the bit is spinning straight by spooling it up. I went back and forth and decided I really did not want to waste this spoilboard by making a new one, but it was annoying that the matchfit pieces didnāt lock in. So one sleepless night it occured to me to just 3d print slightly larger matchfit pieces that would work with this dove tail, I only needed like 6 of them and then when this waste board is done I Can just recycle them and go back to the normal size ones. I tried it out and it works great!