> > Since I'm just getting into setting up spindle control I'm interested in > > this > subject. Can you tell us how you are actually electrically controlling your > spindle? The problem you are having at first glance appears more electrical > than software. > > > > Can you put a scope on the control signals to the spindle control? That > would tell you if LinuxCNC is doing something to the I/O lines that causes the > change in direction. > > So, my wiring is an RS232/485 connection from the PC to the GS2 VFD using > CAT6 twisted pair wire. I use the Linuxcnc VFD Modbus driver: > (GS2) here: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?VFD_Modbus
Unfortunately that site the link for the GS2 are all broken. > > My configs are in this zip folder: > https://bgp.nu/~tom/pub/EMCOTurn120P/EMCOTurn120P-ConfigFiles.zip Thanks. I'll take a look. > > Again, I have never seen this kind of behavior in the 5 years it�s been > running. I have seen some oddities but not related to the spindle suddenly > changing direction or stopping. The back to back G96s seem to be the trigger > (if not the cause)�. Modbus is pretty robust with the CRC so messages to set speed and or direction quite often just involve a minimal number of bytes transferred back and forth. The brushless 1.8kw AC Servo with 2500 line encoder I'm evaluating here, once set up, takes a write register operation to PN33 with a value between +/-3000 to set the RPM. The enable can also be done with a MODbus message to PN95 or a discrete input. I've tagged on a scope photo of what the Rs485 signals look like when the enable signal to the motor is asserted. Most of the noise is cancelled out because it's differential noise but some of it appears to be making it through to several different RS485 receivers causing CRC errors etc. So far I'm not happy. On the 3 phase AC motor front my GS1 manual (1HP 3 phase Baldor on South Bend Lathe) says that the Speed Reference from 0.0Hz to 400Hz; the RUN 0=STOP, 1=RUN; Direction 0=Forward, 1=Reverse; are all in one group so theoretically only one message to set the 3 registers is ever required for normal operation. Addresses 0x091A to 0x091C But to be truthful I've not examined the Modbus control under LinuxCNC that closely yet. I suspect your GS2 is probably pretty close to the same. You'd have to add an RS485 dongle and start listening for the messages that set that particular group. If the message says go in the same direction but you see the motor reverse then it's the GS2. If the message says go the other way but the G-Code says go the same way, then it's the LinuxCNC Modbus interface. And maybe something has changed on your system to also create massive bursts of electrical noise at just the wrong time. > > -Tom > > > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users
_______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
