Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Haysys on April 16, 2006, 03:24:24 PM
-
Hi All,
I have a behavior (on my machine, not me... at least none I WANT help with) that I need a bit of help with. Running an Industiral Hobbies Servo conversion kit with Gecko 320s, PMDX breakout boards (2) on a 3GHz ASUS P4 with addon Parallel port cards (2).
Individually, the X & Y-axix moves just fine from the cursor-keys. And when I press two of the keys for (X,Y) +,+; +,-; -,+ all moves are smooth and clean. But when I try to move X-,Y-, the Y moves fine, but the X-axis stutters. The screw will turn 10-15 degrees, stop, try to move again, stop, etc. Move the X-axis alone or with the Y in any of the other three combinations and it moves fine.
I am just getting the system to the point of testing, and have not tried running any G-code.
Anyone have any thoughts or recommendations?
Thanks,,
David
-
Try changing the step pulse width to 5us under motor tuning. The Next thing that you may have to look for is noise on the step line from the motors..
-
Thanks, I'll check it out.
BTW, when you speak of noise, are you referring to possible inductive noise on the encoder cables??? They are shielded and grounded on the controller end.
Thanks again Brian!!
-
You have it on the noise... I have had this problem in the past and had to install line drivers...
-
OK. I have a more complete view of what is happening and have solved the immediate problem. My previous description of the problem was incomplete; the issue effected all three axes. However, I want to explain the whole situation, the solution as currently deployed, and get feedback on the whole.
By observing the status LEDs on the PMDX board, I could see that the DIR signals for the 3 axes go HI or LO depending on the most recent direction the axis was told to move (I say TOLD, because even if the axis is not moving, the DRO says it is). If the axis moves in the minus (-) direction, the DIR signal stays HI; move in the plus(+) direction and DIR goes LO. If I move two of the three axes (say X &Y) in the minus direction (2 of the 3 DIR signals are HI) I can move both X & Y simultaneously in any combined direction in the proper fashion. But when I tried to move X & Z or Y & Z in the combined minus,minus (-,-) direction, (Thus all three DIR signals are HI when -,- is tried) one or both stutter and lose their position.
More concisely stated, if all three DIR signals are HI, then the combined-axis moves will fail.
Additional notes:
1) The brightness of ALLthe LED status indicators dim further for each HI DIR signal displayed. (This seemed rather suspicious to me, but I'm a bit ignorant of such things).
2) If all three DIR signals are HI, then the X & Y will move in the minus direction if moved individually. However, the Z-axis will not move; it will only move small increments (a few degrees) in the minus direction.
3) I also noticed that Pins 8 and 9 on the breakout board controlling X, Y, & Z were both HI but unused.
I was a bit curious on Pins 8 & 9, and decided that they realy did not need to be lit up and using current. I went into Ports & Pins/Outputs and configured two of the OUPUTS as enabled, and set to Port1 , Pins 8 & 9, and Active LOW. Then after doing the RESET, pins 8 & 9 were off. OK, great. I then tested the axis moves and all three Axes move in any desired combination; even all three at the same time in the minus direction. ???
I am a bit mysitifed by this behavior, but I'm quite sure someone can explain why this was happening, and I would like to hear that explanation. But more to the point, is this indicative of some other problem I have not uncovered??? Is there some circuit that is drawing more current than it should? I have the 5 volt circuit fused at 1 amp, but maybe that means nothing.
Thanks for wading through the details here. Feedback greatly appreciated on this one!!!
David
-
David post your XML and I will see if you have an odd setting... also what board are you running from PMDX?
-
Hey Brian,
The xml file is zipped and attached.
I'm using a pair of NetMos PCI ECP Parallel port cards driving the PMDX boards, though I get the same behavior when connected to the parallel port on board the ASUS motherboard which is a P5LD2 running in Standard PC mode.
Thanks for taking the time to look this over. I do appreciate it!!
David
-
I had this exact same problem. Turned out to be a bad capacitor connection on the Gecko. Replaced and life is good!
-
David I see that you have your port set to zero... this has to be the wrong XML...
-
Brian,
Well, I may have screwed up sending you the wrong file... If I did, sorry to have wasted your time.
The file I am including here is from the machine C:\Mach3, with a time stamp of 4/17/06 10:47 PM--Just closed the app.
Thanks,
David
-
Here is the first test...
-
Brian,
I took the file and loaded it. First thing I did was test as you configured it. Everything worked fine. Then, I turned off the OUTPUTS I configured on Pins 8 & 9 on Port 1 and tested again. Everything works as well.
I looked for changes you had made in the file, and noticed the 5 micro second entries in both Step and Direction, but did not notice other changes. Are there other you entries you made that I did not recognize??
Thanks for your time and attention. I appreciate it!!
David
-
You Got me :) That is all I did... Other than the fact that I looked for any odd settings.
So you are good to go now?