External axis: Mastering, UN-Mastering, and changing gear ratio

    (R30iB+, V 9.3)

    G'day, all. Today, I'm doing my first fresh-from-the-box setup of a Fanuc with an independent (Group 2) 7th axis. It's just running a rotary index table, shouldn't be anything too hard.

    Now, I've managed to Master and calibrate G2A1, and it moves fine, but: The customer wants the axis "zero" to be different from what I originally set, and my default gear ratio is obviously way off -- the turntable only rotates about 10deg when the pendant POSN display shows it moving 100. I'm still waiting on the gearing ratio data from the gearbox manufacturer, but once I get it, what's the process to change that setting in the robot?

    To un-Master, do I just set the DMR_GRP[2].$MASTER_DONE to False?

  • Go to Best Answer

    Ah! Okay, I'm used to KUKAs, where you have to explicitly un-Master an axis before you can re-Master it.

    Anyway, I did the Single Axis Master and it looks fine. Which leads me to my next question: this turntable, as built, has a Mastering vernier/marking-set from the manufacturer... but this zero doesn't line up with the "zero" that the customer wants. I was thinking that, under Controlled Start>Maintenance, the "Mastering Position" setting would allow me to create an offset between the "Mastering" position and the "zero deg" position. But it doesn't seem to work that way, when I tried it. Do I need to just ignore the vernier marks, jog the axis to the 0deg the customer wants, and re-Master it again?

    On a related note, I did find the Gear Ratio setting under Maintenance, and changed it to match the turntable gearbox.

    On another related note: under the Controlled Start>Maintenance menu, does the setting Motion Sign let you reverse the +/- direction of an axis? I got lucky with my axis, but for a bit I was thinking I might have to invert/negate it.

    Yes, the motion sign is a bit of a crap shoot. It works exactly as you think it does.

    If you want to make the Vernier useful without moving them, zero master where the customer wants it, then jog so that the Vernier is lined up. Record this position. Then go back through the maintenance setup at controlled start, and when it ask for the mastering position, enter that value. When you do a single axis master now that will be the default zero position.

    Quote from pdl

    If you want to make the Vernier useful without moving them, zero master where the customer wants it, then jog so that the Vernier is lined up. Record this position. Then go back through the maintenance setup at controlled start, and when it ask for the mastering position, enter that value. When you do a single axis master now that will be the default zero position.

    Hah! Thank you! That worked perfectly (once I remembered to to the Calibrate after the Single Axis Master).

    Quote from pdl

    Yes, the motion sign is a bit of a crap shoot. It works exactly as you think it does.

    Thanks. I'll keep that in mind for the future.

    Normally I would make sure to remind people to take backups and be cautious, but it's pretty obvious from your post in this thread, not to mention your 7k+ other ones, that you know what you're doing with this procedure.

    Besides, I'm sure you had a backup already :winking_face:

    • Best Answer

    If I may add, you don't need any gearbox or reducer data to calculate the gear ratio.

    1) declare and master the external axis in Controlled start under maintenance, use the Vernier marks for zero as this mastering is only temporary.
    2) Boot into Cold Start, go to STATUS > PULSE. Be sure to choose Group 2 from UTIL menu and write down the Pulse count. Your external axis should be aligned to the Vernier marks.
    3) Jog the external axis 1 full rotation until it is realligned with the Vernier marks. Substract the pulse count of the full rotation from the pulse count obtained from the previous step and divide by 1.000.000.
    I am 90% certain, αi/βi series motors use 1 million pulses/revolution encoders but you can verify that on the motor manual and the manual number is posted on the motor sticker.
    4) By dividing the difference in pulses for a full rotation by the ppr (pulses per revolution) of the encoder you get the motor shaft revolutions required for a full rotation, you return to maintenance in Controlled start and modify the gear ratio of the motor you added.

    5) you can now remaster your axis by performing a Cold Start and jogging the axis to your prefered position.

    Quote from ps0f0r

    3) Jog the external axis 1 full rotation until it is realligned with the Vernier marks.

    That's what I ended up doing -- as it turns out, the mfgr data for the gearbox was incorrect or incomplete.

    Although I simply rotated it 360deg until the vernier lined up again, calculated the ratio between 360 and the POSN value the pendant showed for J7, then adjusted the gear ratio again. Worked a treat (well, the second time -- the first try, I got the ratio backwards and made the problem worse).

    Now, at 0 and 360, the verniers line up perfectly. I imagine that if the axis were to be used in an infinite-rotation style (which is is physically capable of), any rounding errors would build up over distance, but this application has J7 limited to between 0 and +360deg, so it's not an issue.

    Quote from SkyeFire

    That's what I ended up doing -- as it turns out, the mfgr data for the gearbox was incorrect or incomplete.

    Although I simply rotated it 360deg until the vernier lined up again, calculated the ratio between 360 and the POSN value the pendant showed for J7, then adjusted the gear ratio again. Worked a treat (well, the second time -- the first try, I got the ratio backwards and made the problem worse).

    Now, at 0 and 360, the verniers line up perfectly. I imagine that if the axis were to be used in an infinite-rotation style (which is is physically capable of), any rounding errors would build up over distance, but this application has J7 limited to between 0 and +360deg, so it's not an issue.

    Pulsecoders never lie. You just need to be sure about the ppr of the motor.

    Quote from pdl

    But a Vernier read by human eyes does lie, if only just a little.

    For the procedure I mentioned above you dont care about the human error involved when you use the verniers to identify the gear ratio from the pulse coder count. That is because even if you miss the vernier by lets say .2 mm you still get the gear ratio by dividing the pulses by the ppr of the motor. The division will give you something like 5,04 or 5,0008 but you have still identified the gear ratio being 5.

    Quote from SkyeFire

    Now, at 0 and 360, the verniers line up perfectly. I imagine that if the axis were to be used in an infinite-rotation style (which is is physically capable of), any rounding errors would build up over distance, but this application has J7 limited to between 0 and +360deg, so it's not an issue.

    Similar to the OP but I have an application where this is used in an infinite rotation style (continuous turn). How would I correct for the rounding errors?

    Quote from Matthew Brincefield

    Similar to the OP but I have an application where this is used in an infinite rotation style (continuous turn). How would I correct for the rounding errors?

    Well, there's probably no way to completely eliminate them. However, they can probably be eliminated over a reasonable working range. Instead of rotating once and setting the ratio, rotate multiple times, check the vernier, and adjust the gear ratio accordingly. It's the same process, but with a longer baseline. The more rotations, the longer the baseline and the less rounding error. At some point, you'll hit the mathematical precision limit of the controller, but that will depend a lot on your physical gearing.

    Quote from SkyeFire

    (R30iB+, V 9.3)

    G'day, all. Today, I'm doing my first fresh-from-the-box setup of a Fanuc with an independent (Group 2) 7th axis. It's just running a rotary index table, shouldn't be anything too hard.

    Now, I've managed to Master and calibrate G2A1, and it moves fine, but: The customer wants the axis "zero" to be different from what I originally set, and my default gear ratio is obviously way off -- the turntable only rotates about 10deg when the pendant POSN display shows it moving 100. I'm still waiting on the gearing ratio data from the gearbox manufacturer, but once I get it, what's the process to change that setting in the robot?

    To un-Master, do I just set the DMR_GRP[2].$MASTER_DONE to False?

    which cable did you used for j7 in the controller?

Advertising from our partners