imperfect circles

[Erik F]

This is the latest of my tests to see what's up. I used new down spiral bit to eliminate the "fuzz" I was getting with the up cut spiral. Instead of using the stock collet, I used the Precise Bits 1/4" collet and cleaned the all the seating surfaces, the collet and nut...just to eliminate as much variable as possible. This file was set up at only 0.050" depth of cut, simply to take all the load off the machine to see what happened. I numbered all the 1" blocks in the order they were cut.

[ger21]

I think it's a mechanical issue.
Every row and column seems to be consistent to within about .002-.003, but the columns and rows vary in size.

Anything other than a mechanical issue, would not be so consistent.

[Erik F]

Thanks Ger21. I read somewhere about pulse width? Which is something that is adjustable in Estlcam control software. Other than pulling the drives all apart I'm at a loss.

[FineLineAuto]

Thanks Ger21. I read somewhere about pulse width? Which is something that is adjustable in Estlcam control software. Other than pulling the drives all apart I'm at a loss.

We can determine if it’s a mechanical issue with a dial indicator. Can you try putting a dial indicator on both the X and Y axes to validate the measurements you’re seeing? This would allow us to isolate the machines movement from additional parameters like spindle runout, material hold down limitations, errors in CAM programming, etc. When you do the test, can you run at a low speed so we can eliminate pulse generation issues?

Also, can you post the DIP switch configuration on the drivers you’re using?

[Erik F]

I was able to dig out the old dial...it is nothing special was a cheap one I got off Amazon. I don't really have a very good means of mounting it, but I will figure something out. Will see if I can put something together tonight. Also will take a pic of the dip switch set up. I am pretty sure I am running half current for idle. Also last night while trying to figure a few things out, I noticed that the wires before the connection and the connector to the steppers was hot to, very hot...this was on all the steppers. Nate I am using the recycled wires you sent my way with the 4 way plugs on them...not sure if this is a sign of anything that could be giving me issue.

[ynnek]

Might I add a critique? The shoulder bolt going through the ball bearing and then through a needle bearing, may work short term, maybe not in the long term. The roller bearings are hardened over 60RC, but the shoulder bolt is soft. Even with lubrication, the pinions will wear quite fast. I have experienced this on industrial equipment. The shoulder bolt should be replaced with a piece of hardened, precision linear rod. The center of these are soft & can be drilled. Just my 2 cents. Carry on.

[Erik F]

https://photos.app.goo.gl/muXkNJEnJVgk2XsC6

I was able to get a dial touched off the router mount. I was able to get pretty repeatable results running from say 0 to 10" and then back to 0. In the video however...I am pushing and pulling on the gantry and you can see the dial move quite a bit.

[difalkner]

Two tips, Erik. Get a tripod and shoot your videos landscape (widescreen). Or get someone to hold the camera/phone while you're pushing on the spindle mount.

I haven't tested ours to see if or how much it moves when I press on the mount. If I have time I might do that and see what movement exists.

David

[Erik F]

Hi David,
Thanks for posting. I have watched your videos and it seems you are not having any issue with accuracy or repeatability. What would you say your machine is hold accuracy wise? If you have seen my previous posts you know I'm seeing some cuts +\- .015-.020 off.

[difalkner]

I created a file very similar to yours and have a board on the CNC ready to cut and see how it compares. But I haven't had time to get to it... ugh! I'm hoping in the next few days I can do that and report back.

David

[Erik F]

I created a file very similar to yours and have a board on the CNC ready to cut and see how it compares. But I haven't had time to get to it... ugh! I'm hoping in the next few days I can do that and report back.

David

Wow! Thank you. Your feedback will be very helpful (as usual).

[ger21]

Put the indicator on the back plate, or on the spindle body. Without the spindle installed, that clamp will flex a lot more than when the spindle is clamped into it.

[Erik F]

Good point Ger21....I didn't really think about that. I will try again tonight if I get a chance.

[Erik F]

https://photos.app.goo.gl/qBVK9WBCQUv3WB4D8
This was the other night just running from zero to ten inches then back to zero

[Erik F]

https://photos.app.goo.gl/RrvdhG7bkLn8umUU8

This is from tonight. The video in the link show going from zero to various distances and back to zero. You can see the gauge is within about half a thousands each time...I'm pleased with that. I'm sure my cheap gauge and my farmer clamping system don't add to the accuracy of this set up. The problem is when any pressure is applied to the gantry the gauge is moving. With small pressure +/-0.005 and with a little bit of force +/- about .015-.02. Ran more 1" squares as well and I see the same issues. Some were off by 0.025 to 0.030 inches. Well almost 1am back to work in a few hours.

[FineLineAuto]

https://photos.app.goo.gl/RrvdhG7bkLn8umUU8

This is from tonight. The video in the link show going from zero to various distances and back to zero. You can see the gauge is within about half a thousands each time...I'm pleased with that. I'm sure my cheap gauge and my farmer clamping system don't add to the accuracy of this set up. The problem is when any pressure is applied to the gantry the gauge is moving. With small pressure +/-0.005 and with a little bit of force +/- about .015-.02. Ran more 1" squares as well and I see the same issues. Some were off by 0.025 to 0.030 inches. Well almost 1am back to work in a few hours.

So I think we can be certain is not linear mechanical error. With the problem described it seems like the idle steppers are not fully locking in place. Have you tried setting the idle current setting to full on the 542 drivers?

[ger21]

.02" is a lot. Something must be loose somewhere. Not sure I buy the idea of the steppers moving.
Try moving the indicator to different parts of the machine, to see if you can narrow down exactly which parts are moving?

I never thought you were losing steps, as the error you're seeing is very consistent.

[difalkner]

I'm tied up most of today but hoping to get back into the shop mid afternoon. I'll try to replicate what you've done for comparison if I can get that far.

Much better video, btw!

David

[Erik F]

https://photos.app.goo.gl/VXu3E77k5Ss2P1gDA

https://photos.app.goo.gl/CrFf9JdcNANz2giGA

One link is to two videos and the other is to one. I was able to get away for 10 min before work today and used the GoPro.

I have the drivers set at half current but I can change that to full easy enough. I have my opinion on the videos but what to see what you guys think before I posion the well.

Busted out the GoPro with tripod for you David ...I think you can really see where the issues are coming from though.

[difalkner]

Wow! That's a lot of movement. Is that the pinion wobbling around the shaft (bolt)? I just went out to see if ours moves and it doesn't budge at all. In the second and third videos I can't tell what all is moving, just that it's a lot. It's almost as though the teeth are worn and there's play in the R&P. At first I thought the whole machine was just rocking on the stand but now I'm not sure what all is moving but I can see it's a lot.

Yes, tripod and GoPro is much better.

David