On Thursday 25 September 2014 10:50:50 Sebastian Kuzminsky did opine And Gene did reply: > On 9/24/14 3:30 AM, andy pugh wrote: > > There are a number of solutions to mutiple-gear headstocks, including > > the sistributed "gearchange" the significantly more useful "spindle" > > from Les Newel (which handles more than 2 gears) and my own approach > > of detecting gear by the ratio of input and output speeds and > > selecting gains automatically. > > I've never used any of these gear changer components, and i dont know > much about what would be useful in one vs what's just cruft. I can see > that you all (Andy Pugh, Sam Sokolik, Gene Heskett, Les Newell, etc) > know much more about gear changers than i do. > > I would like it if you distilled your desires down into one component > (including documentation and preferably tests) that we could review and > merge. > > Maybe this is Les's component from [0], or maybe it's Andy's components > with the reviews & testcase i gave in [1], or maybe it's Sam's > component[2], my point is i don't know. Please clue me in and i'll > help with reviews, or the build system, or merging, or packaging, or > whatever i can. > > 0: > http://thread.gmane.org/gmane.linux.distributions.emc.devel/4043/focus= > 4239 > > 1: > http://thread.gmane.org/gmane.linux.distributions.emc.devel/4043/focus= > 4233 > > 2: > http://thread.gmane.org/gmane.linux.distributions.emc.devel/13921/focus > =13923
For my viewpoint Seb., some of this code I have looked at is oriented to control the gear changer itself, and some of it is intended to adjust the pid's output to get the same speeds (within the range available in that gear reduction) at a small pid.error and the correct speed in both gears, by changing the gear manually, and then switching in out out a gain multiplier between the pid and the servos target, in my case the pwmgen. So there are 2 situations, and its the software version that changes the gains that interest me the most. And the way I'm leaning to do it could be done by autogearchange if it can be verified that it can handle the spindle when running backwards too, a basic requirement for the G33.1 rigid tapping operation. Others will have different requirements, won't even have a pid in their configs, depending entirely on physical gear changes to get the speeds to use. And some might be a mix of both as in the autogearchange where I believe there has outputs than could be used to trigger a near stop, the gear change itself, then restore the previously requested speed. >From that, the statement can be made that it is plain that there is not a "one size fits all" solution. So I would propose that the best outputs of our programmers, many of which are far more capable than I, would best be by including them in the Contributed Components section of the wiki, pretty much in the as contributed state so it doesn't depend on whether or not you have the hardware to test it with, but make sure the authors email is included and valid and permissible to be used by the author as a support channel. This is being done right now, but without the authors specific ok's to do so. And, most of us have little or no problem making changes to this code to fit our exact circumstances, so it doesn't need a thorough vetting before making it available. This is not the heart and core of LinuxCNC that has to be 6 months of uptime stable, its configuration related stuffs that can be extremely volatile without impinging on the core stability of LinuxCNC itself. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ That said, there has been a huge, exploitable bug discovered in bash, so I recommend that you run a cat5 to every machine you all own and run the update-manager on an ASAP basis. Some distro's haven't packaged it yet, but these ubu boxes of mine are all netted, and already updated. Once again amazing me at the level of support ubu is offering for a 4.5 year old LTS release. I had already been offered, and installed the updated packages before reading about it on the net last night. Apparently it can be exploited by any means used to get a new file into your working boxes. Google called it "catastrophic". And if you are anything like me, 1/2 or more of the work this machine does as background tasks, is driven by bash scripts I wrote, running as daemons started at boot time. 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> US V Castleman, SCOTUS, Mar 2014 is grounds for Impeaching SCOTUS ------------------------------------------------------------------------------ Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
