I am doing some digging on another issue we are having and I am curious if there is a parameter or some readout for distance to marker pulse of the ecncoder or even how to identify the marker pulse without using a scope.
Does anyone have any knowledge on this for the robot?
Finding Marker Pulse
-
ih560 -
October 23, 2018 at 8:23 PM -
Thread is marked as Resolved.
-
-
dmr_grp[1].spc_count updates with each count. If u jog until this parameter value is equal to zero, you will be on the marker pulse position. I think i read that the count value is actually a couple pieces of data at different bit offsets such as count, revolutions, and maybe error check, so u might be able to calculate how much to increment the spc count for each revolution of the motor.
-
So after jogging the robot around some and watching this number, I am assuming that SPC_count is at 0 when on the marker pulse where the pulse count was established at initial power-up or after batter loss. I suspect this because SPC_count is at 0 several rotations away from where the actually 0 point is, and pretty close to where I remember jogging it after battery failure to re-establish the pulse count.
-
Your understanding is correct. Did that info help with the issue you were trying to resolve? Or is there anything else you can add to describe the issue you're trying to resolve?
-
It was somewhat useful. I am trying to establish a method for quick mastering the robot (after BZAL) in conditions where the robot cannot go to the witness marks and without additional hardware.
-
Youre in luck, i have posted a thorough description of how the factory mastering counts can be restored after bzal without remastering. This will work for your situation with some slight modifications. When u say the robot cannot go to the witness marks, do u mean only all axes at the same time but possibly one at a time? Factory zero is not necessary, you can remaster to whatever position you want and still be able to restore after bzal but its very helpful to have some new scribed or otherwise marked position for a new "zero" if you cant reach factory zero on an individual axis. Check out this post https://www.robot-forum.com/robotforum/fan…7802/#msg117802
-
I have read that post and understand what you are doing there, I'm just worried the process is a little repetitive and time consuming for the average person, who will be low on patience already after a robot "failure" (dead batteries and not the actual robot). The issue that we have is the J2 and J3 axis don't have a great place for a scribed line for a new "0" reference and I am trying to avoid additional hardware for a marker. These robots are spread all over so I will not be able to personally get to each one.
-
The process only seems lengthy for the first attempt at finding the master pulse positions. If you can do this ahead of time as a preventative measure and somehow mark the robot to know where to jog each axis, it can be done quite quickly. The benefit of this restore master method as opposed to any remastering method is that the robot calibration and hence positions will be exactly as before. All other methods except quickmaster rely on human precision and will inevitably be slightly different at best. No matter what you do, if you lose master counts due to BZAL, you will need some reference point to jog to if its not the factory zero marks. Maybe a hardstop? But you said you didnt want to add hardware. The encoder position is the only other thing to look at but you need a starting reference position. Im curious, what was your original idea that involved needing the marker pulse?
-
If your BZAL occurs when you're at a known position (like home, for example), there's a pretty simple procedure for remastering without using witness marks or hardware. On the Position screen, write down the current joint values. Go to master/cal, single axis master. Put the joint values from your known position in the MSTR POS column, enter a "1" in the "SEL" column for each axis. Jog each axis to establish pulse , return the robot to the exact position that you wrote the values down. Execute then calibrate.