On Wednesday 08 November 2017 11:08:29 Peter C. Wallace wrote: > On Tue, 7 Nov 2017, Jon Elson wrote: > > Date: Tue, 07 Nov 2017 22:07:36 -0600 > > From: Jon Elson <[email protected]> > > Reply-To: "Enhanced Machine Controller (EMC)" > > <[email protected]> > > To: "Enhanced Machine Controller (EMC)" > > <[email protected]> Subject: Re: [Emc-users] PID > > refresher for Gene? > > > > On 11/07/2017 08:58 PM, Gene Heskett wrote: > > Hmmm, your raw encoder signals seem to have a LOT of width > > variation. Since your time scale is 20 ms, there are 20 samples > > between screen divisions, so the sampling rate is not likely to be > > the cause. Generating encoder velocity from timing the edges > > suffers if there is poor accuracy of the state transitions, and > > would be expected to produce a jumpy velocity just like what you > > showed in a previous Halscope trace. A very poor eyeball guess at > > the A-B trace looks like you might get 10 - 20% velocity error > > easily with those duty cycle variations. This ought to show up on > > an actual scope, to verify it isn't in the Mesa -> HAL part of the > > system. > > > > Jon > > > > -------------------------------------------------------------------- > >---------- Check out the vibrant tech community on one of the world's > > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot > > _______________________________________________ > > Emc-users mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users > > Just counting graticle dots I see some places where theres almost a > 2-1 difference in time from edge to edge (take a look at the violet > edges before and after green falling edge roughly 1 graticle line left > of center) The velocity trace seems to be correctly reflect this > timing difference. > > You can also see that some of the error is straight quadrature error > and some looks like code wheel variations. > > With less than 2 counts per sample, the peak noise is proportional > 1/T_err. With 2 or more counts per sample the peak noise is > proportional to 1/T_errN where N is the number of counts per sample > interval. > > So higher quadrature accuracy will always improve things and higher > resolution encoders will improve things if they have 2 or more counts > per sample period at normal machining speeds.
I have a 1000 line Omron encoder on order. I'll fix it to the butt of the motor shaft, and use its a/b, but the z will the index pulse from this one, changing the encoder scale according to the gear selection switches. Or doing it automatically with hal. Hopefully that will put a stop to this male Bovine output. :) I've made the gear switch mounting sandwitch pieces, out of chewing gum that looks like alu, and sticks to the tool really bad, then found the back of the knob is hollow, and runs about 5 thou from the casting when its properly seated on the shaft. So my 8mm dimple will have to become a bump glued into it, and projecting up high enough to engage the switch rollers. But my bottle of rapid fuse, which seems to work better than regular superglue is about spoiled, so I'll have to go get another after I feed the missus. > > Peter Wallace > Mesa Electronics > > ---------------------------------------------------------------------- >-------- Check out the vibrant tech community on one of the world's > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users 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-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
