On Saturday, May 05, 2012 11:07:28 PM Andy Pugh did opine: > On 5 May 2012, at 23:42, gene heskett <ghesk...@wdtv.com> wrote: > > I think I was in the devel branch, but similar. Right now g76 is > > waiting forever for the index pulse, which is not good, so I > > obviously mucked up something. encoder.0.index is visible in hal > > meter, but G76 can't see it. > > You need to wire up motion.spindle-revs to spindle encoder position and > the index to motion.index-enable.
Thanks Andy. That isn't quite the dotted i and crossed t to what I did, but that is working fine now. I added: net spindle-index encoder.0.index-enable <=> motion.spindle-index-enable net encoder.motion encoder.0.position => motion.spindle-revs I had to add a second scale channel because I was already using scale.0 in the spindle speed for the rps signal, and the sum2.0.gain1 for the encoders speed output I had to set at -1.0 to get the control diff as a pwmgen.0.value input. Otherwise the feedback added & the spindle took off to max speed at the acceleration allowed. So the closed loop spindle speed control is working. I had it running at about 3 rps when the cutoff tool dug in and hung it. Cleared the fuse in about 3/4 second, 4th fuse I've had to put in in over 10 years! Running that slow without the feedback, the stock controller simply isn't stiff enough to pour the good gulf into it and blow the fuse. But I wasted .6" off the end of my 5/8" rod of cold roll (big deal? no) finding out 2 things. One being that it looks like a formula I found in the handbook, where the od - id in the K parameter needs a little further tweaking, because that figure also needs reduced another 20% to compensate for the 10% flat bottom of the groove, and the 10% flat top usually left by std threading taps & dies. Clearance in the threads for this application is a gas leak, at high enough pressures (peaking at 15-20 kpsi) that gas cutting damage will rear its ugly head in about 100 shots. That is my mistake of course, however g76 did something else I am not extra pleased with. I had spec'd an 'l2' and an 'e0.005'. At 200 rpm that may not be a legit value for 'e', but that is minor compared to its inability to initiate the backout a bit earlier as it gets deeper, so the q29.5 advance caused a pretty heavy duty trench in the shoulder as the depth of the thread increased. Humm, thinking out loud here, I have made another mistake in not giving it a starting 'z' that was half a 'p' away from the shoulder I don't want to dig up because if I did, I would still get the withdraw at the end of the last loop exactly enough that its a never mind at the shoulder? The actually cut thread should be identical to what I had in mind 2 hours ago. That is an important and nearly final piece of info to allow me to tame the G76 command I believe. If that's truly the fix for the dug up shoulder, it should be noted on the G76 pages shouldn't it? I don't believe it is now. A 2 line footnote could explain that. Next, I need to swap the X motor for one with some Wheaties, I hung or slipped that little one several times. But the couplings to do that are on a rowboat from Bejing, motive power unknown. I may as well see if I can catch up on the honeydo's till then. Thanks Andy, I owe you at least a pint if we ever meet. Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> Whip it, baby. Whip it right. Whip it, baby. Whip it all night! ------------------------------------------------------------------------------ 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 Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
