How do I know when my machine lost steps. I only found it because the cut looked slightly funny. Holes and pockets were a bit off which was ok but also not really because it shouldn’t have been that way. The last cut came pretty close to the clamp which could have been disastrous for the cutter. So is there any way of getting the program to warn you about it is there a mechanical way of warning? I measured the position in the homing state and it was out by 2mm after running several tool paths over the day. A new ref. corrected it but why did it happen in the first place and how can I find out before things turn ugly. Any thought are much appreciated.
As I understand, you either need to push more current to your steppers or not push them so hard. Not pushing so hard could be accomplished by taking shallower cuts.
I thought that might be the case. I did update my toolpath to the max of what the machine can handle but I might overdone it a bit. It was the time factor that made me do it. I can take it slightly back again but I do actually like it the way it is. I might have to ref it more often. Was looking for an automatic ref after each program but I haven’t quite found a way yet.
Ref?
Referencing/ homing
You know about G54 through G59?
Yep
I hope some of the more experienced users get in on this and give you solid assistance.
Lol. I thought you come up with an awesome plan now. I thought of G28.1 but that didn’t work. Will have to look into some macro calling.
I only know enough to be dangerous. I haven’t operated a machine for a few years and even then it was only on a school setting.
Right. I might wait than until someone else hooks onto this post. But thanks for showing interest anyway.
the only way to know while you are cutting is to have some kind of feedback. Rotary encoders on motor shafts, or scales on your axes. Open loop you just have to run far enough below your machine’s maximum capability that it is not an issue. Or you can beef your hardware up so you push the headroom you have further up. Bigger steppers, better motor drives, higher voltage, better linear actuators, lighten inertial load, etc. Though sometimes stalling is also caused by harmonic imbalance too. Stepper systems often have bad resonant frequencies. That’s all voodoo black magic to me. In the simplest terms there are certain speeds steppers can struggle to run at. But there are strategies to eliminate, or shift that too.
I just went through some wackiness with my machine after an upgrade and it turned out to be a combination of things causing me problems. Some of those things I did not suspect had gone wrong either. The take away from that is never assume anything. Maybe you have a chip hung up in your linear guides? The unexpected can bedevil anyone.
So check your machine for play, or binding. Even if you’re absolutely sure everything is OK. Pilots look over planes before they take off. Even though they figure the plane is probably OK. So give stuff a spin, or a push. Kick the tires.
You might just discover something you had no idea was going on.
As a long term solution, there are drop-in-replacements with servo drives to replace NEMA23 steppers. I’m planning to do that late this year.
The servo drives have an ALARM output that could e.g. be routed to ESTOP. When the servo doesn’t have enough power to reach the position, the alarm will stop the entire operation before damaging the workpiece.
(They also have a position-reached output. Maybe that comes in handy for G31 probe commands to make sure a position is precisely reached before continuing.)
I did check everything over Paul but found nothing. I thought of performing an auto ref home at the end of my hose. For instance, after the code is done the machine goes back to the home position anyway but it doesn’t perform a ref home sequence. As i said earlier after pushing the ref home Butten everything was back to normal. I did try 28.1 in MDI but that didn’t give me the result I’m after. What it basically needs to do is resetting the machine to its original state after each gcode program. That doesn’t really fix the problem but it will at least give me some assurance that the next code is getting executed out of a homed machine.
I think MACH3 supports you to add rotary encoders to your existing steppers…if you have enough input ports left.
I do have the inputs but at this point not the money to upgrade. Well, it looks like a push of the ref home button every now and than will have to do it.
A rotary encoder can cost you less than $10.
There is no way to tell when a stepper loose steps while it’s working. You always find that after finishing the job or if a disaster occurs (which also finish a job).
This is the “problem” when there is no closed positioning loop.
BUT, normally, this kind of positioning cinematic using steppers without any positioning transducer is fairly reliable if, as Kyle already suggested, the motors have to have enough current to perform the needed torque for the cutting forces you are working with. Or, inverse, the axial cutting forces do surpass the ones resulted from the torque delivered by the steppers. So you have to work with adequate cutting regime.
If even with light cutting regime you experience such lost steps, it might be another issues like: blocking due to misalignment of guides or of bearings of screws with the but, or too much speed (in case of ball screws), or even the motor drivers.
It’s almost only about “fine tuning”!
In practice I rarely meet such “lost steps” and only at fast movements (with big machines can you can experiment large inertia forces so must be careful with accelerations tuning).
And, yes, it is a good practice to “home” if not before any new program at list at the begining of a new work session.
@Miguel_Sanchez Thats right, Miguel. You need then a software whitch knows what to do with the infos from the rotary encoders 
@Peter_Spiess you can find many DIY digital readout projects that can do that cheaply too http://www.yuriystoys.com/2012/09/do-it-yourself-dro-with-arduino-and.html