Matthew Herd wrote: > I've gotten my Bridgeport BOSS 5 mill running fairly well. I'm using the > Pico Systems USC board and LinuxCNC 2.4.6. > > First off, I observed that the default acceleration values in the > Pico-Systems sample configs were too high for reliable operation of the axes > on the BOSS. Particularly, the Z would stall randomly. I suspect it has > something to do with the microstepping on the Gecko 203V's, along with the > high inertia and detent torque of the old NEMA 42 motors. First I dialed the > rapid speeds, then the acceleration values way down and have had reliable > behavior. I haven't touched the PID terms and have had no issues there once I > got the Input Scale & Output Scale values set the same. Once I get the > acceleration parameters tuned a bit better, I think I can turn the max > velocity up to closer to factory values (120 IPM). Can anyone suggest a > method for optimizing the acceleration values? > > With an open-loop system, there is no measurement available other than your ears or a dial indicator. As long as it doesn't lose steps, you are OK. When first starting to machine actual work, watch out for missed steps. > Second, I have mounted my spindle encoder (512 CPR quadrature w/ index), > wired it up, and copied the appropriate lines from the threading example > config (from USC_motion.hal). I wanted to confirm that the 'scale' parameter > in the line "setp ppmc.0.encoder.03.scale 6912" should be the counts per > revolution, or in my case, 512 (or is it 4x512?). It should be 2048. Use halmeter to read ppmc.0.encoder.03.position, turn the spindle exactly one turn and see if the value changes by 1.000 +/- a small amount. It should count up or down one integer value per turn if scale is correct.
For final sanity check, make it run a tapping cycle in air with a tap in the spindle, and put your fingernail lightly in the grooves of the tap. If the whole operation works right, your nail won't move up or down. You may also be able to see this with a light shining behind the tap. > I'm looking for a way to confirm that the encoder is responding properly. > I'd like to find a way to test the rigid tapping functions prior to actual > use. > > Third, a series of questions relating to spindle speed control. By way of > background, I have solid state relay control of the solenoids which operate > the vari-speed drive air motor and the Pico Systems DAC board for adjusting > VFD frequency. I plan on writing a component which will turn on the spindle, > adjust the vari-speed drive as required, and trim with the VFD. > This is using one variable (the S word) to adjust two things. It is possible, but with the varispeed, there's no need to change motor speed. Precise spindle speed is rarely important, +/- 100 RPM won't make much difference except at the lowest speeds. There are some things on the wiki about closed-loop speed control components. > At this point, I'm not sure I understand what I'll be getting for > ppmc.0.encoder.03.velocity as I have not defined user units for the USC's 4th > encoder channel because it doesn't correspond to a machine joint. > Yes, the encoder.xx.scale sets that. if the scale is set to 2048 for a 512 count/rev encoder, then velocity will read 1.0 at 60 RPM (one rev/second). > I'm also wondering how I might be able to determine the direction of rotation > of the spindle. Is the velocity pin negative when rotating counter clockwise > and positive when rotating clockwise? I'd like to use this to determine > whether the backgear is or is not engaged. > Depending on the way you wire the encoder A and B, it will be signed. If the sign is wrong, you can switch the A and B wires or make the encoder.03.scale a negative value. > Not that I've read through my hal files thoroughly on this issue, but I'm > wondering if it would be feasible to make sure that motor rotation direction > is automatically corrected for backgear engagement (to ensure that M3 rotates > the spindle clockwise regardless of whether the head is in backgear or high > range). I can see an instance where I forget to switch from/to backgear and > break a cutter. Been there, almost done that. I think the best way is to put a microswitch on the high/low range lever, but you could do what you say. Jon ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
