Rigid Tapping Setup

[RJS100]

I just completed upgrading my spindle motor to an AC servo motor and thought it was time to setup the rigid tapping parameters for the first time. I wired an Omron 2000 count E6B2-C encoder some time ago, but never used it. I wanted to start by comparing the 0-10v analog output to the spindle speed.

The servo motor has a speed analog instruction gain setting of 1rpm/V. Factory default is 300. Using the manual spindle control, I turned the spindle speed to max, and the encoder reported 5320 rpm. Since there is a 1.8:1 pulley ratio between motor and spindle, that means the motor was actually spinning at 5320/1.8 or 2955 rpms. I adjusted the factory default analog instruction gain from 300 to 295, to reflect the fact that the servo's max rpm is actually 2955, not 3000. I also noted that the min rpm of the spindle was 115 rpm.

I also temporarily disconnected the acorn analog output from the servo drive and re-ran the spindlebenchtest program to be sure acorn output was calibrated properly.

I changed wizard spindle settings to reflect the min and max rpm found above, but now when I manually adjust the spindle speed, the minimum spindle speed indicates around 200 instead of 115.

Also, there is a discrepancy between instructed and actual spindle speeds. Using MDI commands, I issued Sxxx followed by M03 and recorded the speeds:

S500 ---> Around 600 rpm
S1000 ---> 1120 rpm
S2000 ---> 2180 rpm

Lastly, the readings from the encoder bounce around a bit and are hard to read a steady number. Is this normal?

I have attached a recent report file. Not sure how to correct these issues.

Thanks... Richard

[cnckeith]

rigid tapping requirement is the encoder has to be at 1:1 with the Spindle

[RJS100]

Yes... The encoder is 1:1 with the spindle. Sorry I forgot to mention this. The spindle is splined, so I 3D printed the same size timing gear that is on the encoder, and made one with a splined hub that fits over the spindle. I am reading the exact position as the spindle 1:1. See attached. I was not sure how a 3D printed timing gear would hold up... but so far so good.

[RJS100]

I experimented with this a bit further and made the following changes:

1. I changed the minimum spindle speed setting in the wizard from previous 115 to 0. I left the max rpm at 5320.
2. Using the Rigid Tapping Setup directions, I adjusted the servo analog instruction gain (similar to the frequency setting of a VFD) until the spindle was turning 640 rpm when I entered the mdi command M3 S640. I used a handheld tachometer and got withing 1/2 rpm.

The result was that when commanded to spin 640 rpm, the spindle turned almost exactly 640 rpm, however when commanded S1000, S2000 or S3000, it was as almost 200 rpm off. Also, instead of reaching a max rpm of 5320, the max is now about 5000 rpm.

Lastly, it is still very hard to read the spindle rpm as the numbers bounces up and down and are not steady. Is there a buffer setting of some kind for the encoder readings?

I hope this helps... Richard

[tblough]

It doesn't matter what the exact rpm is. CNC12 uses the encoder to synchronize the Z-axis with the actual rpm as read from the encoder.

[cnckeith]

It doesn't matter what the exact rpm is. CNC12 uses the encoder to synchronize the Z-axis with the actual rpm as read from the encoder.

yep!

this is always a common misunderstanding, newbie's think CNC12 is somehow automagically controlling the exact position of the spindle like it is a servo motor when in actuality there is a much better approach that we use for Rigid Tapping.

a demo of cnc12 "electronic gearing" spindle encoder position electronically geared to Z axis.

is here at 21:26

[RJS100]

ok... i understand the concept of the exact rpm not being that important. But should the rpm's read like this when commanded to spin at 640 rpm?

https://www.dropbox.com/scl/fi/uttv7rm2 ... qencu&dl=0

Richard

[cnckeith]

in the wizard under spindle setup menu which selection do you have it set to?

[RJS100]

Yes... that is how I have it set.

[cncsnw]

Either your actual spindle speed is bouncing around erratically, or your spindle encoder is giving erratic readings.

If the spindle speed really is bouncing around erratically, that could be due to noise or distortion on the analog command signal; or it could instability or control problems in the spindle/servo drive you are using.

Can you run your spindle/servo drive from its keypad (e.g. in a test jog mode) so that its speed is as steady is the drive can make it, without use of the analog command? If you do that, is the displayed RPM steady or does it still bounce around like it did in your video above?

Is your "Omron" encoder actually an Omron encoder, or is it a Chinese copy from Amazon? If you paid less than $200 for it, it is probably a Chinese copy.

When you measure spindle speed under the same operating conditions, using your handheld tachometer, does the reading also bounce around? Of course, you have no way of knowing what kind of averaging window the handheld tach uses, but if the handheld shows similar fluctuation, then you can reasonably assume that the spindle speed is really changing.

In my experience, using decent VFDs and induction motors, the encoder-derived spindle speed in the Centroid status window will not vary by more than 3-5 RPM while running at steady speed.

If you have an oscilloscope, you could use it to look at the analog command signal (to look for noise and interference); and you could use it to look at the quadrature channels coming from the spindle encoder (both for distortion and interference, and also to see if the speed really is fluctuating erratically).