Closed Loop Steppers? (masso)

I have been running my Elite for some time now and it is great to say the least. I had an issue yesterday that makes me wonder about the closed loop steppers. I was machining the insert for a Vcarve inlay in end grain maple. I was trying out some new bits from SPE, 1/8" Compression and the bit broke about 30 seconds into carve. I hit the estop, jogged the spindle to replace bit and did so, set the bit to Z zero, I rewound the program and thought I should verify X, Y before starting and found it out about .350 inches in both X and Y, so I rezeroed them and began cutting again with more conservative feeds, the second bit broke about 2 minutes later, I repeated the same process and found the same issue with X , Y being off about the same amount but in different directions, did it all over again with a different bit and finished the carve only to find out that it was no good as during the first bit break it cut in a place it shouldn’t have.
I thought the closed loop steppers wouldn’t allow this to happen, do I assume wrong.
Pat

Hey Pat,

just as food for thought: the closed-loop steppers will ensure that the Step and Dir commands they receive are executed. Of course they don’t check if the commands coming from the controller are correct or e.g. corrupted by EMI.

With closed loop it is always important to not forget where the loop is and where it ends. Here it is completely inside the driver/encoder, which sits in the small black box on the motor.

Another thing that can exist as a source of error outside of the loop is a divergence between the stepper axis and the ball screw, i.e. a loose coupler.

1 Like

Aiph5u, thanks for the response, I believe I have done everything to prevent EMI and so far I have had nothing but perfect results, only this instance where I broke a couple of bits have I noticed the issue. There are no mechanical issues either.
Thanks again.
Pat

1 Like

When I researched closed loop steppers a long time ago I seem to remember there being maybe three main systems used. The most common, and probably least expensive, was a step-loss compensation(?). In this case the downside is lost steps are not ‘compensated’ until the end of a given programmed move. At the time, I could find no details on which type of closed loop system the Masso motors used, and I did not ask as at that point I had already decided I did not need/want them for my build.

Maybe this is part of the reason you saw the behaviour you did.

Perhaps others with more expertise with closed loop systems can add to the conversation.

Thanks for the info, I submitted a ticket to Masso as instructed by 1F, see what they come up with and will post it here.
Thanks, Pat

1 Like

Pat,

I’m wondering why you didn’t start the file from the broken bit position? The Masso is known to remember where the E-stop was pressed, and will allow you to start from that line number. Anyway, when you hit rewind, it begins from the start of the file, and although I don’t see why the X / Y would be different, but I also wonder if you tried the MDI command of “go to work offset”, would that have been a more correct position? How did you determine that there was a 0.35 error in position? Could it have been operator error? I’ve been running my Foreman for about 3 months now and have 4 work positions on my spoil board with stops to register against, and I have not reset them and they are always “spot on”. I use G-55 thru G-58 for locations, and they always go to my start position.

P.S. - I doubt that closed loop steppers have anything to do with your issues, unless there is slipping in a drive coupler.

1 Like

I was 20 seconds into the carve and didn’t even think about that feature, my only real concern is that the machine cut where it was not supposed to and ruining the inlay part, I thought the closed loop steppers would have prevented this.
The failure was not from resetting the origin and there are no issues with the drive coupling, I ran the same part with different bit and more conservative setting and it was perfect like all other parts made on this amazing machine.
Thanks for your input
pat

Hi Pat,
The difference in an open and closed loop stepper is simply that the open gets a command to move, but does not report back to the controller if it actually made the move or not. Closed loop report back to the controller that it has “driven” the amount that was requested. That is why the question about the drive coupling being tight, as the motor doesn’t know if position change occurred or not, just that the motor made the correct revolutions as commanded. So, logically, if the motor reported the correct rotation, but the X/Y didn’t move the correct amount, it would result in a file position error. I can’t say what happened in your case, but closed steppers are really not likely the issue and cannot “save” a problem, only report the success or failure to attain the commanded shaft rotation.

I had a case of closed loop working when I was zeroing my Z height after a bit change. I had dust in the collet that prevented the touch plate to magnet on collet from making an electrical contact, and the bit drove slightly into the touch plate, but was obstructed from going further, and before I could hit the E-stop, the Z motor went into alarm and stopped everything. It was an example of the feedback that the motor could not attain the rotation any further. In my Woodworker, it would have kept on driving the bit into the touch plate until something broke, or smoked!

Hey Jim, hey all,

the Elite closed-loop steppers with integrated driver and encoder don’t report back if they moved an amount of steps. They only report back an error condition to the controller if there is a movement discrepancy between their stepper driver and their encoder that they could not compensate by themselves, in this case they trigger the error line and the controller can then stop the program.

The Step and Dir lines are one-way and with no feedback.

That’s why I suggested above that if the Step and Dir signals transmitted from the controller are somehow not in order (EMI, loose contact), the closed-loop stepper will not know about it.

1 Like

Pat and Aiph5u,

Didn’t know that. It sounds like feedback, but in a different way, I guess.

Anyway, Pat, I went out and played around with my Foreman, and was able to kind of duplicate your error. Here’s what I did: I set an arbitrary zero position for X and Y and started jogging in Continuous and max speed. Once it was at max jog, I hit the E-stop and there was “some” coasting of the travel to a stop, but the DROs and machine position froze at the E-stop pressing. I then used the MDI to go to work origin, and it was off by .500 inches from where I had originally set it. My thoughts are, if you had a fair amount of travel when the bit broke and you selected the E-stop, it could have “skidded” to a stop, and that amount might explain the position error, as the DROs stop immediately with E-stop selection, such that the file is not updated with the “coasting” as it were. Interesting situation.

That certainly would explain how I lost the origin, still not sure how it cut in the wrong place though.
I appreciate the effort you are going through with this.
Pat

Pat,

My thoughts are that it cut in the wrong place because the origin is now in error. When you rewind the file, it starts at a different location than it previously had for zero on the material, it is not starting at the lower left or whatever, but at what it now falsely thinks is the start point, and the error was the amount of the “skid”. I don’t think Masso will have anything to offer as it appears to me that it was just an unexpected error due to the situation of the skidding. I also think, my advice of starting at the line where interrupted by the E-stop, probably wouldn’t have been any more successful either. Maybe a comment in the manual that if you E-stop with a file underway, you must establish a full zeroing of material to ensure accuracy. As I said, a very interesting point that you have discovered!

Onefinity, if you are watching, maybe an instructive video showing this situation would help others, even though it is a physical anomoly of any CNC, not just a Onefinity.

I have mine call for a rehome on startup and anytime estop is hit. It is the only way to realign the xy0 point again.

After home a simple go to work orgin will get you back on track

Masso indicated my machine should have done just that and told me where to make that setting adjustment and i have done so.
Thanks for all your input on this thing.
Pat

2 Likes

It’s on the homing screen in f1, says something to the nature of “require rehoming on estop” and the other is “require on startup”

Pretty sure it’s coming from 1f with only home on start up selected