I have an old deckel maho mh1000T. I'm planning to retrofit it with masso. The only problem is that the masso isn't in closed loop. This deckel maho has linear glass scale Encoders which can give square wave signals for position and i wanted to connect them with masso. When will masso release a closed loop version of the board?
Closed loop masso
- Thread starter sourabh
- Start date
zombieengineer
ZombieEngineer
@sourabh
I very much doubt that Masso will ever support closed loop control using linear scales for the following reasons:
Looking at Heidenhain's website (a German manufacturer of linear glass scales) they list the protocols they support include: EnDat, Siemens, Fanuc, Mitsubishi, Yaskawa, Panasonic, PROFIBUS-DP, PROFINET IO & SSI. What protocol are your linear glass scales using?
The reason why I believe that direct closed loop control using linear scales (linked with servo motors) have faded from the market place is due to a failure mode resulting in crashed axis. Should anything go wrong with the CNC control loop (linear scales, cables or controller) the servo motors will drive the axis past the ends - sometimes (frequently?) resulting in damage to the machine.
The current system of Step / Direction control of motors (both stepper and servo motors) means that should the CNC controller hang the step pulse would immediate stop and the whole machine should "freeze in place". The penalty of the Step / Direction control method is that backlash needs to be accounted for (ballscrews can largely eliminate backlash so the penalty is small versus the gain).
The reason why Step / Direction method was not used in the early days of CNC was the speed of the CPUs was not fast enough to generate the step pulse train - you need to generate 10,000+ pulses a second versus adjusting the servo motor velocity 10 to 100 times a second. With the velocity control method if the CPU hang for any reason the servos would keep running at whatever speed they were previously set to until they hit an immovable object like the end of the axis (if the linear encoder cable failed the motor would accelerate to maximum speed for an "improved crash").
I hope this provides some insight into the issues.
I very much doubt that Masso will ever support closed loop control using linear scales for the following reasons:
- There is no standard communication protocol for linear scales (there are at least 6 different interface protocols)
- Direct closed loop control using linear scales and servo motors was a solution for a different era (1980s)
Looking at Heidenhain's website (a German manufacturer of linear glass scales) they list the protocols they support include: EnDat, Siemens, Fanuc, Mitsubishi, Yaskawa, Panasonic, PROFIBUS-DP, PROFINET IO & SSI. What protocol are your linear glass scales using?
The reason why I believe that direct closed loop control using linear scales (linked with servo motors) have faded from the market place is due to a failure mode resulting in crashed axis. Should anything go wrong with the CNC control loop (linear scales, cables or controller) the servo motors will drive the axis past the ends - sometimes (frequently?) resulting in damage to the machine.
The current system of Step / Direction control of motors (both stepper and servo motors) means that should the CNC controller hang the step pulse would immediate stop and the whole machine should "freeze in place". The penalty of the Step / Direction control method is that backlash needs to be accounted for (ballscrews can largely eliminate backlash so the penalty is small versus the gain).
The reason why Step / Direction method was not used in the early days of CNC was the speed of the CPUs was not fast enough to generate the step pulse train - you need to generate 10,000+ pulses a second versus adjusting the servo motor velocity 10 to 100 times a second. With the velocity control method if the CPU hang for any reason the servos would keep running at whatever speed they were previously set to until they hit an immovable object like the end of the axis (if the linear encoder cable failed the motor would accelerate to maximum speed for an "improved crash").
I hope this provides some insight into the issues.
@zombieengineer
Yeah, you speak the truth. But i have seen large cnc manufacturers such as DMG mori, who still use linear encoders for feedback on every machine. I don't think it's outdated yet. And honestly, i personally think that it's a good idea to use linear encoders. I wonder what masso has to say about this. As far as your thoughts on linear encoders causing crashes, does not seem to be a problem now, we can always set a tolerance limit now a days, if the encoders show values beyond that, machine will come to e-stop.
Yeah, you speak the truth. But i have seen large cnc manufacturers such as DMG mori, who still use linear encoders for feedback on every machine. I don't think it's outdated yet. And honestly, i personally think that it's a good idea to use linear encoders. I wonder what masso has to say about this. As far as your thoughts on linear encoders causing crashes, does not seem to be a problem now, we can always set a tolerance limit now a days, if the encoders show values beyond that, machine will come to e-stop.