I’ve read a lot about spindle tramming on this forum and the time consuming process to get XYZ dialed in properly. I just built my CNC this last month and was thinking about this issue.
I retired at 52 from R & D mechanical engineering, 8 years ago. I was trying to think about the things I’ve designed, manufactured and built with my own hands in the past. How can I make something challenging for people and make it simple, accurate and spend less time doing it?
I’ve probably designed thousands of jigs in the last 30+ years. Just like building this CNC, some instructions tell you to hand tight the bolts/screws loosely, go onto the next step, and then fully tighten later. Why? It has to do with fitting-makeup of tolerances between parts and it would make assembling less difficult.
Here’s my conceptual idea. I think I’m going to have this built and test it on my CNC.
This large disc (shown here) would be manufactured with a high tolerance for FLATNESS and PERPENDICULARITY to the 1/2" long steel shaft using GD&T.
HOW I WOULD USE THIS:
Insert 1/2" rod into spindle all the way until it bottoms out and tighten the collet nut.
. Lower the spindle until the bottom of the large disc is about 1/4" off the spoilboard.
Loosen the collet nut and let the disc fall onto the spoilboard.
Retighten the collet nut. (the disc may now become where it’s not exactly laying flat 360.)
Loosen all 4 bolts on the spindle tramming plate until the plate lays flat or mostly flat.
Retighten all bolts, leaving the eccentric cam nut for last. Rotate the cam nut until is cannot move, and then lock it down with the bolt. Rotating the cam nut CW or CCW does not matter because we are not using the cam nut to make any adjustments. We essentially removed that process through the use of the setup disc. We are only using it to keep the set position.
If there is any gap between the disc and the spoilboard, it will be due to NOD. If so, shim the back of the spindle according to AVID CNC instructions until the gap is removed.
Remove the disc and raise the spindle a couple inches.
Re-insert the setup disc but DO NOT bottom it out in the spindle and lower it to the spoilboard until it barely touches.
Inspect with feeler gauges to verify no gaps.
Run the tramming procedure to check for flatness and no grooves, as per AVID CNC instructions. No further adjustments should be needed.
Suggestions or comments? Does anyone think this would be work?
Would you consider buying such a setup tool as this to make tramming a thing of the past?
I’m not expert, but a couple of possible concerns.
I don’t think you want to be perpendicular to the spoilboard. I think you want to be perpendicular to the plane defined by the axis the spindle moves on, then surface to spoilboard to be parallel with that. So there should be some shimming of a known flat surface before using a jig like this.
I can picture a jig like this being helpful with the endless shimming process, but would it really be rigid enough? Seems like it would flex under the weight of the spindle and you would still need to do lots of adjusting.
If you try any experiments I’d love to hear how they work out, I am still quite new to cnc usage. Also maybe if you surface the spoilboard with a small bit it will be close enough.
You either compensate the spoilboard to be level (as close as possible) with the machine as it travels X & Y on an imaginary datum plane…or do…
You compensate the spindle adjustments to be parallel to spoilboard that is hard mounted to the frame.
Which way is better? Is either way wrong? What is the pro’s/con’s of either?
If #1, then you’ll probably be skimming more spoilboard material off one side or the other.
If #2, it will be the same as #1 but a lot less, in my opinion.
The image only illustrates the idea, not the actual material being used or any ribs that might be added to the top surface (if any) to the disc. This is the least of my worries.
I hope this gets people to start thinking about this process and why we approach this problem from on point of view or the other.
Maybe a good illustration of GD&T (Geometric Dimension & Tolerance) using 3 axis of perpendicular and parallelism might be a good idea?
How close is close when it comes to precision of the machine for what you need it to do?
The traditional way of tramming is to use a jig similar to this on a glass plate that has been accurately leveled.
Your method might seem easier but when tightening the bolts it will not take much to knock it out of alignment. The traditional method measures the condition of the spindle after all the bolts are tightened. If more alignment is needed you have to go through an iterative process until you get it square using the dial indicators. I would expect the same would be true for your proposed process but it would be much harder to see the error and correct it. Of course if you are willing to accept that initial error to be good enough then by all means this would be easier and faster.
A beefer jig that some how holds the spindle firmly in place while the spindle bolts are tightened might also work but I don’t think you can just count on the weight of the spindle to do it.
I’m well aware about the traditional method. I don’t have this dual dial indicator, but I had a single digital indicator that I used with a jig I created and I got it pretty close. However, when I tram in the Y-axis, the right side is a little deeper than the left side. I still have a little adjusting to do before I feel comfortable creating my spoilboard.
My indicator is set 9" from the spindle. I used it for my drill press to set the table level years ago and reused it for this purpose. My test board was .06" off on the right side but I got it down to .016" in a span of 18", so far. Not sure how close it should be but I think between .005-.010 for my purpose might suffice.
Just for a sanity check, I used my digital cube on the vertical side of the spindle tramming plate and even though is shows 0 deg, obviously the spindle itself is off.
It doesn’t have to be any beefier to hold the spindle. The tensile strength of the rod itself is more than enough to hold and balance the spindle, even if it wasn’t attached to anything.
