Need help deciding if Granite Devices will work with my servos

[Al_The_Man]

Have you measured your backlash?
Your motor to load inertia ratio is important lowering the accel/decel improves it.
The ratio is also reduced by the square of any reduction.
Are you saying that you are close to 1:1 gear ratio?
Al.

[JerryFlyGuy]

Have you checked the system for backlash? The different sizes of the 'legs' of the H would lead me to believe there is some lash in there somewhere, either than or something is flexing?

Reducing the accel will cause rounded corners when using Constant Velocity [CV] in Mach3. However, if you don't have the juice to drive the gantry at the speed it needs to then your no better off no matter how high the accel is set.

Can you extrapolate a bit on your 1:1 gearing. Are we talking rack and pinion, ballscrew? If it's R&P then your issue is not enough gearing.. if it's a ballscrew then the jury is still out until one know's what the lead is.

Fwiw
J

[gavztheouch]

Hi Jerry, its rack and pinion prob close to 3:1, both axis are quite easy to move with my hands.

I had a feeling the rounded corners where something to do with the CV setting, I reckon the next thing to do is to turn that off and try again.

Regarding the backlash, if that where the problem why would it only occur now and not last week, is there a test for backlash built into mach 3?

[gavztheouch]

Just found a lose grub screw on the y-axis that was producing a massive amount of backlash, its now tightened and the movement on both axis feels very tight. The loose grub screw backlash was only present when the y-axis was held tight while I manually moved the motor so it is very unlikely it effected the tests as there was not enough friction to cause the pulley to slip when using the pen, and also the problem seems to be with the heavy X axis not the Y. I can't physically feel any backlash in any of the axis, certianly not enough to give the visible results in the photos.

[Al_The_Man]

Still not sure what your travel/motor rev is, if from what you say it seems like 2"/Motor rev, which for that weight is quite considerable, For a gantry like that, I usually calculate my maximum motor rpm when going with the highest ratio possible that will give me give the desired max travel speed ("/Min) rate I need.
Al.

[JerryFlyGuy]

How large is your pinion? I'd guess that 3:1 is pretty small. I've got a heavy router [gantry weigh's ~ 1 ton] w/ two motors slave driving it. They are the MCG3304 motors.. 250oz-in cont w/ a 800oz peak [IIRC] and it will drive the gantry at ~75ipm w/ an accell of 22in/s/s. However if I push things harder than that is just doesn't get up to speed. I can also run into corner rounding issues. The major difference is that I'm using 10:1 reducers and 1.5" pinion w/ those servo's.

If that helps as a reference.. my best guess is that you don't have enough reduction on the servo's, however I don't know what kind of spec's you've got on your motors.

Backlash is an easy test. I'm not sure which axis your having issue w/ but for now I'll assume it's the X. You can do this test on any axis you want [to check lash].

Using the MDI jog out away from the home switch at least 1".

Zero the axis.
G1 x-0.050 F25[assuming inches on you machine.. use 2mm if metric]
G1 x0
Re-zero the axis.

Now the test:
Put a dial indicator on the machine such that it's measuring travel on the X axis. You'd prefer to have the indicator partially depressed so that it will register movement +/- from the x=0 position. Once you've got this set up and have noted the indicated # do the following.

G1 x 0.050"
G1 x0

Now read the Dial indicator and whatever deviation it has from the original reading is your backlash.

Some notes in regards to back lash readings.

You want to make your first move in the same direction as your going to take your reading from. Ie; an x+ move, then zero then another x+ followed by the X- move back to zero.
The reason for this is to take up any lash in the system. Then when you zero your DRO you know there is no lash that will effect that first +0.050" move. This allows you to see the difference when you do the x- move.

Hope that makes sense..

J

[gavztheouch]

Had another go at tuning the motors today and noticed some values I had plummed into the GDtool where wildly off ie... I had a value of 1 for the P position gain on the big x axis So I reread Teros manual and tried again. Now its working great, perfect letters curves and all, and quickly too, a good speed for routering out which is going to be the next step.