Centroid oak and delta servos

[Houseman303]

My name is Heiko and I live in beautiful Switzerland. After 12 years of Eding CNC, I have treated myself to something real. Centroid oak. I have worked with Fanuc and Heidenhain controlled machines. Since there is no documentation for Delta ASDA B2 servos, I made the connection cables myself. i need a little help.

setup for the servos by tech bulletin 264

P01-44 160000 Encoder output pulses**Delta encoder line count

the servos move, but there are errors, see pictures.

[cncsnw]

Your expected-velocity (VExp) vs. absolute-velocity (VAbs) suggests a mismatch in both direction and scaling, between what the Centroid software expects the drive to do, and what it actually does.

The red trace (VExp) is what Centroid thought it told the drive to do.

The green trace (VAbs) is what the drive actually did, as indicated by its encoder feedback.

If the drive moves minus when Centroid tells it to move plus, then you need to either:
1) Reverse the polarity of the command pulse train (e.g. swap the command A phase with the command B phase), or
2) Reverse the polarity of the returning encoder signal (again by swapping A and B), or
3) Configure the drive to interpret command pulses with reversed sense, or
4) Configure the drive to send out encoder feedback with reversed sense.

If the drive moves too far in response to command pulses, then you need to either:
1) Change the "electronic gear ratio" so that the motor does not turn as far in response to a command pulse, or
2) Change the output encoder resolution so that the drive sends out fewer encoder counts for a given amount of motor movement.

[Houseman303]

VExp and VAbs now point in the same direction. Now I just have to get the errors under control. I now understand some of the settings in the CNC12, but not the settings in the Delta asda B2. It would be good if someone from Centroid would comment on the topic.

Thank you for your support

[Houseman303]

new day, new problems. drive enable for 0.5sec makes my no happy.

[centroid467]

What errors were you referring to in the post from yesterday?
What did you mean by the settings in the Delta ASDA-B2? Are you asking for help interpreting the drive's parameters list or something else?
Drive enable for 0.5 seconds? Does this happen when enabling all drives or a specific drive? I would recommend changing to the standard PLC as mentioned in the other thread before troubleshooting this issue.

[Houseman303]

Hello guys, I need the setting values ​​for the Delta asda B2 servo. My axles have a gear reduction. per motor revolution 7.08333333mm distance(geardrive 34:48 and 10mm pitch).

P01-00 0x1000 Encoder is AB pulse output
P01-01 0x0100 Precision mode control
P01-40 3000* Motor max commanded RPM. Depends on motor
P01-41 200 Max analog torque command
P01-44 ?
P01-45 ?
P01-46 8192 Encoder output pulses**

Encoder output pulses in CNC12?

[centroid467]

I would set P1-44 and P1-45 as recommended in TB264 unless the motor encoders somehow do not match that. The gear reduction and screw pitch is accounted for in CNC12 in the F1 Setup > F2 Config > F2 Machine > F2 Motor menu. You would insert the 7.08333333mm / revolution in the third column of that table.

P1-44 and P1-45 set the ratio of input pulses from the OAK to the Delta motor's encoder counts. This translates the OAK's 8192 pulse per revolution outputs to something the ASDA-B2 drive can use to position the motor. P1-46 then applies a similar ratio to divide down the encoder pulses per revolution to match what the OAK expects. You set "Motor steps / rev" in CNC12 Motor menu to 32768 because those 8192 pulses per revolution are in quadrature and therefore the real count is 4 * 8192 = 32768.

Did that help clarify those drive parameter settings?

[Houseman303]

TB264 for Delta asda A2, not Delta asda b2.

P1-44 320000 for Delta asda A2
Delta asda B2 max. 160000! for P1-44 posible.
if P1-44 320000 and P1-45 8192 for asda A2,
then P1-44 160000 and P1-45 4096? right?

[centroid467]

Leave P1-45 as 8192 and set P1-44 to 160000.

[Houseman303]

8192 * 320000/8192=320000 (asda a2)

8192 * 160000/8192=160000 (asda b2)

This is not the same result and explains the errors in the procedure