Also make sure you are driving your optocouplers with respect to ground, and *not* tying one leg to the 3.3v or 5v lines. The inputs to many drivers seem to hearken back to TTL logic days and are screen printed with the optocoupler inputs listed as pulling down from V+, which besides inverting your output, also does not work well on the BBB. Switching from using [3.3v, pin] to [ground, pin] solved a similar issue somewhere on this forum.
On Saturday, November 5, 2016 at 12:57:15 PM UTC-4, Jeff Pollard wrote: > > > Hi, > > Check the Xylotex hal file in the ARM/BeagleBone directory for an > example of using PID to get rid of the hunting problem on the DIR line. > > Jeff > > > On Saturday, November 5, 2016 at 7:01:30 AM UTC-7, Charles Steinkuehler > wrote: >> >> On 11/5/2016 8:11 AM, Johan Jensen wrote: >> > >> > _4. After completing the first jog command (+/- does not matter), the >> DIR signal >> > starts to oscillate between 0V and 4V continuously. __Not expected._ >> > 5. If I use "disable motors" in Machinekit the DIR signal stops >> oscillating and >> > goes back to 0V. >> > >> > Any ideas on how to correct this would be greatly appreciated. Thanks. >> >> This is "hunting" where the PRU driven step/dir generator is >> oscillating back and forth (by a fraction of a step) around the set >> position point due to the two asynchronous timebases (ARM CPU core and >> PRU stepgen) and the ARM interrupt latency/jitter. >> >> If this is causing you problems, you could add a deadband to the >> hal_pru_generic stepgen driver (PR welcome!), or switch to using a >> standard HAL PID control (with the PRU stepgen in velocity mode) >> instead of the fairly simplistic PID loop found in the hpg driver. >> There are some examples of configurations using the HAL PID + steppers >> for hostmot2 (Mesa) hardware, and I think there have been some for the >> BeagleBone if you dig through the group archives. >> >> A PR for an example BegleBone configuration using HAL pid components >> would also be welcome! >> >> -- >> Charles Steinkuehler >> [email protected] >> > -- website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit --- You received this message because you are subscribed to the Google Groups "Machinekit" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. Visit this group at https://groups.google.com/group/machinekit. For more options, visit https://groups.google.com/d/optout.
