KUKA Mastering questions

1. yes. you can connect robot to some controller and master it. mastering info stays on the robot, not controller. so after this you can power down, disconnect them, transport them, connect them together (or to other controller) and everything will be fine. robot will lose mastering if moved by force (without using controller). an example of this is manually driving robot in emergency using manual release wrench (12mm socket, connects to end of the motor).

2. you can check mastering logs (C:\LRC\ROBOTER\LOG\mastery.log). mastering is selected based on application and required accuracy. if your robot is HA or ABS variant, you do want to use the mastering with offset. if you have a standard robot and variety of payloads or tool changer, you may also want to do the same. if your robot only has fixed tool(s) and always handles same or similar payload, standard mastering is fine.

3. yes and sometimes you have no choice but for sake of repeatability it is better to bring all axes to premastering position, then master them all. that way you get the smallest variance in readings when using check mastering.

EEprom

as long as other controller is setup to use same robot arm you can connect another one. it will ask you if you want to switch over of course but that is it.

something this helps speed up commissioning of several cells when only one controller is hooked up to power (in North America power is not entering KRC through X1 connector, it is cable that need to be fed through bulkhead and terminated at the disconnect.... and this requires tools and time).

yes, mastering info (and bunch of other things) is stored in the EEPROM.

this is possible because KUKA arms have servo motors that use resolvers as a feedback. resolvers are absolute position feedback for one motor pole and motors are three of four poles. saving last position in EEPROM and checking position when powered up again allows determining if axis was moved. now, if somehow one is able to move the powered down servo motor exactly 90 or 120deg depending on number of poles - robot would not be able to detect position change - it would assume that it is the same position. but that is practically impossible to happen.

many other robots use quadrature encoders for feedback. they need constant power to check for movement. and they use counters to keep track of the position. those counters too need to be powered all the time. if the power goes out, position is lost and robot need to be mastered.

resolvers are much nicer in this regard - they can be stored for years without power or fear of leaky batteries or batteries going down. also one can power down robot, disconnect motor cables, check the motor with multimeter for example, connect things back and it will still be fine upon powerup - no loss of mastering. disconnecting resolver cable while robot is powered up will be sensed and axis will be unmastered by the system for safety sake.

Quote from Maverick

Thanks a lot for the help!, regarding to question 1, so doesn't really matter if i plug the robot into a different controller? it will be not a "position mismatch" regarding to the previous robot that was connected there?, how does it keep the information of the servos if the robot does not have batteries?

As already stated, the KRs have an EEPROM in the RDC card located in the base of the arm.

The "trick" KUKA uses is that dual copies of the Mastering are retained -- one copy on the robot hard drive, and one in the RDC card. This data is usually updated when the robot is undergoing a controlled shutdown -- that is, the main power is removed, and the KRC processor keeps running on the batteries for 30-60sec to perform various "cleanup" actions.

On older KRCs, there was a regular issue where Mastering would be "lost" if the KRC was powered down with the batteries dead (which is why replacing the batteries regularly was vital) -- the data wasn't actually lost, but KRC checks the two copies against each other during boot, and if they don't agree, the KRC insists on a new Mastering as a safety precaution. KRC4s are much better about this.

When the KRC detects that the KRC (hard drive) and KR (RDC card EEPROM) Mastering data do not match, it will generate a message asking which one is correct. Normally, this is used so that re-Mastering is not required if the RDC card has to be replaced, or a similar repair. The "correct" Mastering then overwrites the other copy.

For the past several years, a brand-new KRC right out of the crate will throw the Mastering Data Inconsistent fault the first time it is powered up. This is b/c the KRCs are usually built in one factory, and the KRs are built and mastered in another, so the first time they get "paired up" is at the Integrator or the end customer. KUKA performs Mastering on each KR at the factory by connecting them to a KRC that never leaves the factory, and is just there for Mastering and power testing. So, on a brand-new KRC/KR, you always want to choose the RDC copy as the "correct" Mastering.

It sounds like you want to do something similar to what KUKA does at the robot factory -- Master a bunch of KRs using a single KRC, then ship them with valid Mastering data on the RDC. This should work fine, but it will be important that whomever does the first power-up at the final install site knows how to handle the Mastering Data Inconsistent fault.

The good news is, even if they get it wrong, the worst that can happen is that you have to re-Master the robot on site. But having them do it right the first time will save a lot of agravation.