Gene, that makes sense about the software ADCs possibly being too slow to keep up with the rapidly changing input.
I have pot overrides on my machine and they work great. I start all untested programs with feed and rapid ORs set to 0% and then creep up rapid. When the gets to the first feed move it stops and then I creep up feed all looks well. I really reduced my stress testing programs this way, as the 300ipm rapids on my machine scare the heck out of me even on programs I know are good! I have the 7i76 set to software mode 1 so I can read inputs 0-3 as analog, and have pots connected to those pins. The ADCs in the 7i76 are only 8bit, but I guess I don't know if they are software or hardware. Either way the seem to be plenty fast enough. I do get some dithering of the OR value read from the pots, I have a custom HAL comp that adds some hysteresis to settle this down. I agree with you that if and number of IO allows encoders would be preferred over pots. They can be had for a lot less then $20 each, more like $2.50: https://www.ebay.com/itm/232735252724 The nice thing is they also have a push button, which could be hooked up to set the override to 100% just by taping the knob! For feed, speed and rapid ORs I use the halui pins set to take a direct value count input. Here is my HAL file that handles the ORs: https://github.com/KurtJacobson/RF45-config/blob/master/overrides.hal When using the halui pins the sliders in most all the UIs update to match the pot setting, with is very nice. If I try and change the ORs setting with the slider nothing happens, as the slider can't change the position of the pot so easily :D Another advantage of the encoders instead of pots is you could adjust the ORs with the sliders, and the encoder counts would be reset. Cheers Gene! Kurt On Wed, May 30, 2018 at 9:43 AM, Gene Heskett <[email protected]> wrote: > On Tuesday 29 May 2018 09:36:14 andy pugh wrote: > > > https://forum.linuxcnc.org/gmoccapy/34610-potentiometer-bug#111106 > > Most software A/D's need a stable input for the duration of the > conversion, and successive approximation types are extremely sensitive > to noise. Because either is slower than mollasses, taking many servo > cycles to finally clock out their result, I'd never use them with > linuxcnc. Too slow. > > What I would use rather than a pot, is one of mpja.com's 20$ encoders, > and feed the quad signals to a counter as I have done for the jog wheels > on the Sheldon. The counters can be limited both ways, so a rapid spin > CCW can reduce the counter to a zero state quite rapidly. > > I am doing this on the pi, and because its rather complex in the hal > dept, its all in a 200hz thread, and have effectively a real time > control of how much "gain" per click, but on more speedy stuff it could > be done in the servo-thread. Zero reason this same code string, with > some changes in the setp's to control the output range, couldn't be used > for control of feed speed in effectively real time. This would fail for > very slow speeds of course, depending on how low you have the vfd > programmed. I have mine set to shut down below 10HZ > and the low speed boost set to limit at the motors nameplate FLA. Seems > to work quite well despite the nearly full loss of any cooling fan, and > its never got hot enough to consider an aux fan for cooling. > > However, for feed rate control, I do not see a handy input pin. Yet the > axis slider works. ??????? Needs more docs. Or a better description of > how the pots arm is connected to linuxcnc. The forum post doesn't say. > > Cheers, Gene Heskett > -- > "There are four boxes to be used in defense of liberty: > soap, ballot, jury, and ammo. Please use in that order." > -Ed Howdershelt (Author) > Genes Web page <http://geneslinuxbox.net:6309/gene> > > ------------------------------------------------------------ > ------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers > ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
