[thosj]

I have Keling provided encoders, US Digital: E5-500-250-N-S-E-G-2, that say 500 CPS. Do I use 500 to calculate in Mach3, or 2000? The US Digital site says quadrature something or other 2000, so I don't know which to use.

[TOTALLYRC]

I have Keling provided encoders, US Digital: E5-500-250-N-S-E-G-2, that say 500 CPS. Do I use 500 to calculate in Mach3, or 2000? The US Digital site says quadrature something or other 2000, so I don't know which to use.

If they are 500 cpr then they are 2000ppr(500x4). You would set them up as 2000 steps/rev.

Mike

[g-codebank]

Hi,

i bought 4 500 ppr (pulse per revolution) rotary encoders. i will make a MPG one of them and the others for axises. But i don't know how can Mach3 read them to moving exact coordinate.( no more step loses ) Somebody said me "you need to write a brain or macro" Do you know any example of this code?

[Torchhead]

Hi,

i bought 4 500 ppr (pulse per revolution) rotary encoders. i will make a MPG one of them and the others for axises. But i don't know how can Mach3 read them to moving exact coordinate.( no more step loses ) Somebody said me "you need to write a brain or macro" Do you know any example of this code?

1. You cannot eliminate lost steps on steppers motors using an encoder feedback into MACH. At the very best you can e-stop the machine if the error between the read position and the DRO settings reaches a preset value. MACH will not do closed loop control. The method to read, compare and e-stop requires some custom programming. A 500 line encoder (2000 pulses per rev) would generate 20,000 pulses per second at 10 rev/sec. That would probably be too fast for reading through the typical parallel port inputs (high speed inputs) and out of the question for systems with non-realtime feedback (i.e USB or other serial solutions)

2. Steppers lose steps because you push them past their torque/RPM curve. Since to correct position you normally have to add in more torque (power) the only way to add more torque on a stepper is to slow it down. If you slow down one stepper you have to slow all of them down by exactly the same amount or the coordinated position is compromised.

The other issue is that you will need at least 8 inputs to read the 4 encoders and that means at least a second parallel port or more complex external hardware.


You can use EMC2 to monitor encoders and do closed loop but the proper action for torque correction for steppers may still be an issue.

By the time you add the encoders, the interface circuitry and software to accomplish true position control you may find just doing it with servo and servo drives might be easier.

Properly designed stepper systems running within their rated torque-RPM window do not lose position. If you have a system that is losing steps the cause needs to addressed.

[g-codebank]

Thanks for your answer Torchhead.

i don't think to use EMC2 right now. But closed loop controlling is needed CNC applications. i am working metal working industry. we made a router with 7N steppers. it sometimes works 2 hours very well but at the end of the milling there is serious problem occur. i have changed handmade breakout board after these step loss and for problems at the end of the milling. i think the electricity couse these problems. Big welding machines working at the same electricity. i used an online UPS to filter electricity of the CNC but UPS's input damaged because of voltage of current noise at the electricity.

Anyway, i understood that i can not build a CNC with closed loop controlled with MACH3 and incremental rotary encoders are not usable to use with MACH3 directly. Maybe only it can be used for making an MPG.

You mean if i can correct the speed/torque than i will get step lossless milling.