I should add that Darwin will probably not do threading. While Mach4 turning will work with Darwin generally, threading I may decide not to implement. If it cant do it pretty much perfectly, Id rather Darwin not try. I may change that view..but currently that's my line of thought. I did leave myself hooks for encoder based threading, Ill wait to see if there really a need in future.

Art

To provide an additional view on homing, my system homes until the switch closes, then reverses and travels until the switch opens, then continues until the encoder index signal occurs.  Since the encoder is 2000 counts per rev and the index pulse is one count wide, this provides for a very accurate home.  Simple mechanical switches are not that repeatable.  However, this will call for a script in M4.  No big deal, just a little exercise in programming.

Peter

Wow:

Great information Art. I'll have to read that over a few times to understand it fully. So much thought put into homing. I guess I should use homing more. I usually just zero the axis' (DRO's) over a point on the stock that I've designated in my CAM program to be zero and work from there.
Once I understand your explanation better I should probably take advantage of homing properly. I'm not sure my home switches are too accurate or repeatable to any great degree being that they are just micro switches so I may never be able to use the feature properly.
 So are you saying that M4 has purposely been made different than M3? Because right now if I home M3 the main Y DRO goes to 6.4" because that's what I've set (I believe) in the offset.
Why do I have it this way you ask - I try to pull the table to the front for easy access which is a Y+ direction. i.e the quill is over the top left corner of the table at X0 Y6.4. Doesn't have to be that way but that's how I have it. Is that something I should change? It would only mean moving the Y home switch to the back instead of the front on the machine.


Sage

Hi Art

Finally made some head way with the latest Darwin 1.145 plugin. I was able to home without any problem and move off of the limits without having to push reset. Tried working on the step/direction spindle. Darwin reg isn't remembering any changes to the settings in the reg file editor. I need to set the spindlefreqhi to 800 for so my spindle will be close to desired rpm and stepwidth has to be set 5 per the C6 speed board instructions. When I exit M4 and restart the spindlefreqhi goes to 5000 and the stepwidth goes back to 0. I did this several times and had the same result. Tried the estop setting for home switch in limits that still isn't working right. When machine tries to home M4 goes to estop when home switch is tripped. That one must be M4. If you can fix the forgetfull regfile settings bug I think Darwin will be there.

Eric

Hi art still no luck with spindle speed change.
Its getting the pwm signal just not doing anything with it above 5000 rpm.
when it turns on the superpid receives a .5v signal what is the same as M3 for 5000 rpm.
in M3 the superpid gets around .5v per 5000 rpm, so 5000rpm .5v, 10000 rpm 1v, 15000 1.5v there about.
M4 is just not putting out anything above .5v what is only 5000 rpm.
So ever M4 does not like the superpid or something is wrong with M4

Hi Art
so I could not get the m4 to keep its settings so I swopped computers and did a complete new install. And guess what everything is working beautiful.


Mach 4 V1936
Darwin 1.14.5

Thanks for all your help and hard work
Bob

I found when that happens I remove anything to do with M4 and Darwin, so I was doing a clean install and it came right.
I think it might just be a problem with some computers.
as the computer on my machine never really has any problems with M3 and M4 other than when I cleaned it up and removed a program M4 needed put it back on those problems went away.
the computer on my dad machine is a piece of poo it can just run M3 after its started twice and it cant run M4 at all its a newer a fast computer than mine.

Wow.. lots of responce today.

Sage:  Yes, when making Darwin I had talks with Brian about simplifying the homing as a base operation, an then allowing scripts to prsonalise
the process. It just makes better sense. As to if you should change..no..not if your happy with it, I use my systems as I see fit, no reason you
shouldnt. I might mention on some system I home regularly just in case somethign screws up, on others I just zero and cut. There is no wrong way, there is simply "my way".

  Hands:   Ill look into the forgotten settings. I fixed memory loss by oving those setting change saving to the shutdown process, so it may be your
system isnt shutting down properly, I'll check that out..

Dan:

    In the end , from what I read, you need 50Hz being fed to the Gecko which then feeds your PID. I think you need to select "Negate Signal" on your spindle signal. Other than that it shoudl work. BUT.. I added some messages ot find out whats up. Try an M3S2500 and then hit the history button on the
bottom of the screen. Whats listed? It shouidl show each step of Darwins journey to set up a pwm 50hz at a proper percentage. Th results may show us
where your problem lies.

Thx everyone, Great work..

Art

Hi All:

   Darwin 1.15 is here. It fixes the reg's not being saved. It also seems to fix the inputs not showing up till you toggle them.
Note: Until M4 fixes that bug that creates that situation I have told darwin to toggle all inputs for 1 second on startup, this causes
the signals to show up properly.

  Unfortunately I cant fix the charge pump, youll still need to toggle the enable a couple times sometimes to get it to work. Ill bump that
report to the guys. Thanks again for all the testing, I think we're just about there in terms of having a working system.. Ill be hooking
up my laser and further developing an engraving system for myself, so over the next months Ill undoubtedly see places to make
improvements. Report any errors you see and Ill see what I can do.

Art

Mach3 moved off, but it didnt accelerate, it just moved off at full set homing speed. This means a few steps "could" have gotten lost by the "auto move to zero" process. It didnt accelerate because it was the driver doing it, and it wasnt smart enough to know how to properly accelerate, though it tried.

Art

That "though it tried" has me intrigued.

I've always been of the understanding that Mach3 hits the switch decelerates to a stop, (reverses direction), accelerates, comes off the switch and then decelerates to a stop and calls that home.

I just scoped it here and sure enough that's certainly what it looks like - clear accel/decel ramps.

I then compared these homing ramp pulse timings to those used for jogging and they're pretty darn close - fractions of ms different.

Certainly here - "though it tried" translates to "and it darn well succeeds too". 

Ian