How to Tram Avid CNC Router

[@giuliobuccini208]

Nice video. 1. Put a plate of glass under those indicators. (Use simple glass, not the tempered one.) 2. Check for squareness of the portal vs the side rails before starting this procedure.

[@GT40Nut]

First I'd like to say thanks for this video it's been very helpful. There is certainly a chicken and egg thing going on when trying to do this adjustment. I just put together the same machine you have. I figured out how to get rid of the chicken the egg and the spoilboard. I put a long piece 15" of .5" drill rod in the spindle and put a dial indicator against the rod, first in the Y axes direction and then in the X while running the spindle up and down to check for out of square . I figured out how to eliminate the runout and was able to use the adjustment nut to get the Y axes so for 10" of travel I was just under a thou of tilt. This is where you really helped. I'm currently getting .029" of Nod over that same 10" but now I know how to fix it, thanks. To bad I can't post a picture.

[@RobertCowanDIY]

Glad it helped!

[@jazzdad52]

Excellent video, you obviously have a lot of skill in CNC work. I had never done this procedure, but last year noticed that my flat cuts looked like clapboard siding on the house. I appreciate the quality of Starrett tools, but I had a different idea to square the spindle, instead of their square. I epoxied a dull end mill into the center of a small flat board, mounted it in my lathe, and turned it round, and square to the end mill. I then drilled a hole near the perimeter of the disk, and mounted 1 gage in it. When I install it in my spindle, I can rotate it 180 deg in any plane, and compare both sides. Not as convenient, but next time I use it, I don’t need to check if jig went out of square, as the t type could. Even if the disk isn’t perfectly perpendicular to the mill, it will read accurately in all positions.

[@RobertCowanDIY]

Yes, this totally works as well! As long as it's rigid and doesn't move when you sweep it, this is perfectly acceptable as well. I just got the spindle square cheap off eBay many years ago.

[@philbuettner1564]

Thanks for sharing. I would suggest a glass or granite slab to zero to though. As you mentioned the mdf has some give to it.

[@RobertCowanDIY]

Oh, I do have a granite surface plate, that was used off-camera to calibrate the tool.

[@jontaylor723]

Very nice Robert! I also have one of these tables and have spent much time trying to tram it. I never really got it to where I wanted it but I got it to the best that it can do with the way I installed it. I think it could be better if I spent more time on the rack and pinion install. My rail seems to dip on the right side .006 about 30 inches into the y, Aluminum in a vice is perfect but facing aluminum on the table isn’t great. It’s fine with a small EM but it’d be faster if using a facer. PAY CLOSE ATTENTION TO INITIAL INSTALL!! BTW i installed 3 extra cross beams and 1/2” T5 aluminum plate on the table with 1”hdpe spoil board. Very sturdy table top. Overall a good machine!!

[@tgammoncad]

Thank you for the video! I have an Avid Benchtop Pro and really want to accurately tram and square the machine for aluminum milling.

[@ptoddcnc]

Thanks much for going through this. This video was very helpful !!

[@StuffWithKirby]

What about tramming the x axis bar before the spindle/z axis. If the x axis is slight angle, wouldnt the z axis then slide forward or back while moving up and down? I also thought when using tram tool you dont try to get both dials read same. You zero dials and then rotated it 180 and verify both are still fairly zeroed. Using the edge tool on my shapeoko was quite the brain burner trying to understand it all the first few times.

[@makethepart]

Great video Robert!

[@MrCrankyface]

Wouldn't this method/holder also assume the dial holder is perfectly manufactured and both dials are perfectly seated in the holders? Let's say one holder sits just a few thou higher than the other, possibly as a result of the shaft having a tiny bend in it, wouldn't this give false readings for example or am I missing something? I really want to implement something similar on my mill so I can stop having to rotate it back and forth constantly to get the head straight ... :) Love the rubber band solution by the way, such a simple way of holding the parallels in place!

[@rsilvers129]

Yes, he didn't do that critical step. Here is how you do it: kzbin.info/www/bejne/qoCWhWCqbZaeaqM

[@crashkg]

I'm curious how you got your tram tool to be aligned in the first place. I usually put a block on the surface and pivot both indicators to that block to zero them out. When I go back to the surface the zeros on both indicators are measuring the same distance even thought the needle will not necessarily be in the same place.

[@RobertCowanDIY]

I used a granite surface plate, which is the standard way to calibrate something like this. It comes with instructions on how to set it initially.

[@KIMKim-ql8zx]

What a clear video and edit. this is surprisingly good. should I consider a MDF from the factory as "flat" surface? I mean at least in terms of flatness required for wood

[@RobertCowanDIY]

Even if the MDF is perfectly flat, the table below it might not be. It's always a good idea to face down your bed/table, that will guarantee it is flat.

[@benheely9508]

When you flexed the frame, how hard were you pushing on that thing? I'm considering buying one of these for aluminum milling and .01" of flex from just a hand touch is a bit concerning. Or I guess the question I should be asking is: how much of a tolerance does it hold in your aluminum parts despite the flex?

[@RobertCowanDIY]

I was putting my whole weight into it. The FRAME flexes just a little bit, maybe a couple thou, but the MDF is pretty flexy of course. If you cut everything using an adaptive tool path, and then come back and do a light cut to sneak up on your final dimension, it works fine. I've had no real issues (yet). BUT, finish could be improved, don't expect a mirror finish on parts. They are functional, but not always pretty.

[@arthurenright4146]

Thanks for the detailed video Robert! I was curious what model Starrett spindle square you are using. The like above doesn't go to a particular product any longer. The 649-1 looks similar to what you use in the video, but it's up to around $200 and I believe you mentioned a price point ~$100 in the video. Thanks again!

[@rsilvers129]

It's not a Starrett. It's an Spi.

[@rho35100]

Thanks for the links!!

[@EdwardThomasCigar]

One thing to ask/add...would you not want to zero each dial off the same reference point?

[@RobertCowanDIY]

I use a granite surface plate to zero everything off beforehand.

[@EdwardThomasCigar]

@@RobertCowanDIY is the idea being the granite sits on top of any irregularities from any out of tram surface ridges?

[@fjbjparts]

iku tuku ndek ndi lek? aku pengen koyok gotean mu lek.

[@EleanorPeterson]

Precision is all well and good, but what I'd like to see is a pair of Robert Cowan-style safety glasses fitted over the machine's googly eyes. And you could mount a full 'beard' in front of the cutter to act as a chip shield...

[@iceman1982one]

No shawes ?

[@RobertCowanDIY]

?

[@rsilvers129]

You forgot to zero out the two dials to each other. It can't work without that. See video on how to use this tool: kzbin.info/www/bejne/qoCWhWCqbZaeaqM

[@RobertCowanDIY]

I mention somewhere in there that I zeroed them out on a granite surface plate.

[@Festivejelly]

You made a big assumption here that your table is flat. You first should have run a dial indicator over a known flat surface such as a surface plate, then shimmed the plate so it shows as flat on each corner, then do the tramming of that surface. Too much room for error in how you have done it especially on an MDF board.

[@RobertCowanDIY]

I don't disagree! The whole surface of the MDF was surfaced flat with the mill, so I'm reasonably confident it's flat. And since the 'reference surface' was machined flat with the router, the tolerance of the bed is only as good as what the machine can achieve.

[@Festivejelly]

Thats a fair assumption but if you surfaced the bed when the head was out of tram what could happen is when measuring you could be measuring from a valley to a high spot. Your point is valid about what tolerances the machine can operate at for sure, but when machining aluminium you want to be as sure as you can be that the reference surface that you started with was indeed flat and level to the gantry. @@RobertCowanDIY

[@RobertCowanDIY]

@@FestivejellyYeah, fair.

[@Festivejelly]

@@RobertCowanDIY Just to come back to this, I saw a chap using a gauge block to do the tramming over rather than doing it directly on the bed. This way it would average out any peaks and valleys on the waste board. These days accurate gauge blocks are quite cheap to buy so thats probably the way to go doing this.

[@acerjuglans383]

​@@FestivejellyThis is my concern too, whenever I see a tramming video. Just because the spoilboard is freshly surfaced doesn't mean anything. If the spindle is out of whack to begin with, you still have nothing to measure to. There needs to be some way where a spindle can be trammed to a laser level line. That's the only way this is going to really work.

[@BillyBob-si2db]

Did you tram your gantry to your table? Seriously, you're tramming to a spoil board? An exercise in futility.

[@RobertCowanDIY]

The table gets surfaced, so that essentially trams the gantry to the table. Then you tram the head to the spoil board. I'm not inventing anything, this is how it's done.