My vote would be for 2 as well. However blended moves by definition are inaccurate so a slight timing inaccuracy is probably unimportant. Assuming the machine has sensible acceleration, 1 and 3 should be close enough together for it to make little difference. I suppose for my plasma example the perfect solution would be to turn on the ouput at 1 turn off at 3. If the machine has very low acceleration you are onto a loser no matter what you do.
Les Alex Joni wrote: > It's a bit far from perfect at the moment. > There are basicly 2 different kinds of outputs: > M63/M64 Turn on/off digital output synched with motion > > M64/M65 Turn on/off digital output immediately > > It's been a really long time since I touched these, I know at some point > they used to behave properly (they probably don't atm, and they would > benefit from looking into them). > M64/65 are basicly executed when the command reaches the motion controller, > so it's not at the exact time when the next motion starts. > M63/64 should probably get attached to motion commands and go on the TC > queue. That way they would be handled by the trajectory planner at the right > moment. > Although the right moment for a blended move is a bit hard to define. > 1. Is it where accel starts on the next move? (we're starting to decellerate > on the current move at the same time). > 2. Is it when the next move is halfway through accel, and the current move > is halfway through decel? > 3. Is it where the current move decellerates completely? > > I would probably go for 2, although I'm not sure how easy that is to > implement. > > Using exact stop it's easier to define the moment/position when to enable > the output. > > Just some thought to keep the discussion going.. > > Regards, > Alex > ------------------------------------------------------------------------------ Check out the new SourceForge.net Marketplace. It is the best place to buy or sell services for just about anything Open Source. http://p.sf.net/sfu/Xq1LFB _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
