I think I found a bug for Mach4. 

Don't use forward slashes in signal descriptions in Darwin.  The forward slash is used as a delimiter separating the port and signal name in the Mach4 profile configuration file. 

When setting a mach4 configuration item, the drop down will show the signal name with a slash and appears to accept it when you save and exit.  When you go back and confirm that the setting was set it will be gone.  In some cases the selected port will be also missing.

nice.. good to know.

Ok, I have my X3 moving in all three axis.  All my input signals are recognized.  I tuned my motors and they seem to be behaving just fine, no stutters or erratic behavior.

Issues I'm seeing at the moment:

  * Had to disable Charge Pump on board the C11G.  Works in Darwin when configured for 100Hz, but Mach4 doesn't seem to do anything with it (ie. toggle it on/off as needed).  Mach4 is configured to see it.

  * When initiating homing from Mach4, each axis moves in the correct direction until it hits the limit switch at which point I get an "hit limits" error without the axis backing off.  I have tried with "e-stop trip" turned on and off.  Is homing suppose to be functional?


-Freeman

Freeman:

   The charge pump should turn on and off depending on if your in EStop or not. If in ESTOP, no frequency should come out, if not that 100hz should be present, Ill check that
and report back.
  As to homing, its the specification of M4 that in homing the axis should move to the switch. No other action occurs and the home is considered complete. I was told originally that scripting or other user operations would move the axis back off the switch.....or not. However , just as in M3, if you share a home and limit, the limit safety trip is turned off until you jog, script or otherwise
make the axis move off the switch. Ill ask Brian how they handle that in other hardware , it may be a macro or other operation that moves off the switch. The Axis are all zeroed at the switch trip position so really whats needed is just an execturion of a G53X0Y0Z0 after a home to get it to move to the machien coordinate of 0,0,0.

Art

Hi Guys:

   Found the problem with Charge pumps and weirdness with other signals. M4 was telling enable  to turn on before re-enabling the control.
Its always been my habit to refuse IO when in EStop, so the enable got refused in many cases. Thats been fixed. The Charge pumps were a typo
on my part and are also fixed. Finally, frequency based spindles were on my list. I cant find a way to get M4 to give me any sort of frequency vs rpm
information, so I guess the plugin has to figure that out. SO Ive put a workaround in there for the moment. If you have a PWM spindle, use an even numbered
output signal and proceed as you did in previous versions. If you have a frequency spindle, use an odd numbered signal. (For example if PWM set signal #6 to
the name Spindle, and set your PWM base frequency in the freq box in darwin. All will work properly., if you have a frequency driven spindle set signal #7 to the
name Spindle. This will put out a frequency equal to the S word. S5000 will put out 5khz, etc.. ). If you just want On/OFF , then map the signal SpindleForward
instead of Spindle in Mach4. Using the Spindle output will map you to PWM or freq, depending on if the mapped SPINDLE signal is even or odd numbered.

  When you config the pulser, version numbers are now displayed for driver and dll , both should read 1.07, if they dont your driver or dll didnt update
properly. Soem people find they must delete the old driver or it will reinstall itself instead of installing the new one.

Art


 

Installed Darwin 1.07 under Mach 4 build 1889.  I had to use pnputil to remove the 1.05 darwin driver since it constantly got re-installed. 

Charge pump now seems to get set correctly when enabled.  I verified this using a PMDX-111 serial monitor that has provisions for monitoring the charge pump on pin 17.  I  added a LED for Charge Pump #1 on my diagnostic screens and it never activates, even though the PMDX-111 says it is.

Art, in your configuration video you configured your limit switches to be homing switches and had them trigger e-stop.  I have done the same. I have all my limit switches tied to pin 11 and they are used for homing. When homing, should the e-stop be triggered by darwin when the limit switches are reached? 

When I home my axis and the limit switches are triggered, I get stuck in a situation where I cannot back off my axis due to not being able to re-enable the machine.  I have to physically back off the axis by removing power and using the hand dials I have on the X & Y.  The Z axis I have to disable e-stop trip and then jog off, since I don't have a hand dial on it.

Oddly the diagnostic screen doesn't show any input signals causing the problem (such as the emergency signal).  Should darwin tell Mach4 that the emergency signal was triggered?  Better yet, should Darwin not issue the e-stop when it is homing.  Does darwin even know it is homing when Mach4 homes?

Is the homing order column in the Homing/SoftLimits work?  No matter what values I use in this column to set the order of which axis homes, my mill homes all axis concurrently.  I would assume (Danger Will Robinson) that setting the Z to a smaller number than the others would cause it to home first.  Setting any value to zero would disable homing for that axis.

-Freeman

Installed Darwin 1.07 under Mach 4 build 1889.  I had to use pnputil to remove the 1.05 darwin driver since it constantly got re-installed.

>>Ive heard of that, the other solution is to find and delete the old one from windows/system32 or whereever your system put it.

Charge pump now seems to get set correctly when enabled.  I verified this using a PMDX-111 serial monitor that has provisions for monitoring the charge pump on pin 17.  I  added a LED for Charge Pump #1 on my diagnostic screens and it never activates, even though the PMDX-111 says it is.

>>Im thinking charge pumps are only used in printer port systems so they never hooked it up.  For my part I actiavet it when the plugin is running,
really thats the only spec on a charge pump, run when the program is running. Mach3 used it as well for Estops, M4 will not, it will just put
out its frequency when running. Im not sure why it doesnt show in led's, Ill check that out.

Art, in your configuration video you configured your limit switches to be homing switches and had them trigger e-stop.  I have done the same. I have all my limit switches tied to pin 11 and they are used for homing. When homing, should the e-stop be triggered by darwin when the limit switches are reached? 

>> No, Darwin igores the limits when homing.. Im think perhaps m4 doesnt .. Ill check this as well.. Darwin works as m3 did in that area.


When I home my axis and the limit switches are triggered, I get stuck in a situation where I cannot back off my axis due to not being able to re-enable the machine.  I have to physically back off the axis by removing power and using the hand dials I have on the X & Y.  The Z axis I have to disable e-stop trip and then jog off, since I don't have a hand dial on it.

>>yes, not good, Ill figiure out whats up there..


Oddly the diagnostic screen doesn't show any input signals causing the problem (such as the emergency signal).  Should darwin tell Mach4 that the emergency signal was triggered?  Better yet, should Darwin not issue the e-stop when it is homing.  Does darwin even know it is homing when Mach4 homes?

>>>I only found there is a difference between emergency and disable in m4, so Ill see about what coudl cause this and fiux it up ..


Is the homing order column in the Homing/SoftLimits work?  No matter what values I use in this column to set the order of which axis homes, my mill homes all axis concurrently.  I would assume (Danger Will Robinson) that setting the Z to a smaller number than the others would cause it to home first.  Setting any value to zero would disable homing for that axis.

 >> Thats an M4 thing.. I think.. I get a call for homing.. and it homes. If all three go together Im not sure where I coudl stop it. But Ill ask the guys why that might be.. perhaps the plugin is supposed to hold them to an order but that seems a bit weird..

Thx
Art

Hi Guys:

  Here is version 1.08 which fixes many many things. First, homing is done in the order you declare in m4. You may set a home or limit to Estop tripable in Darwin and it will work. If, when trying to re-enable a limit is found to be tripping an Estop, it will be
disabled and the control will enable to allow you to jog off, at which point it will auto-re enable.

  Charge pump will work, though the led may or may not, the guys are looking into it.. Most everything works in this version, a couple of bugs remain, I will turn on a register to allow you to calibrate the spindle frequencies if not in pwm shortly, also M4 doesnt show the IO on its diag until after the signal changes.. Ill fix that up with a delayed refresh after startup...but other than that all seems to work well and a great many bugs were repaired. Thx for the reports, it really helsp to get these things out of the
way as quick as I can.

No limitations in this plugin at all at the moment..Please let me know what you find. The driver test should show 1.07 on each.
 

Art


 

Darwin 1.08
Mach 4 build 1900

Charge Pump works much better now.  Although I noticed that I have to cycle Mach 4 through disabled-enabled-disabled before the Charge Pump comes on.  After that it behaves.  The LED still doesn't light.

When homing, the e-stop is handled correctly.  I can now back off the axis.  Although Mach4 doesn't leave me in an e-stop, it does have the limits triggered and the axis is not zeroed.

Are there documents available for developers of motion controller drivers?  I would expect it would outline the division of responsibilities between Mach 4 and the motion controller driver.  It might also show points for where script drivers would interject their own Lua code to control behavior like homing, probing, etc.


-Freeman

Freeman:

  No, there is no doc, its why it will take a little while to stabilize plugs, its still early for the developers in terms of docs.
Things are getting better though. When you home, the machine coordinate gets zeroed to the offset value. For example,
if you have an offset of 10, the X axis will read -10 plus and decel distance observed. So the machine coordinate should
read the actual distance from the trip point of the switch. Doing a G53X0Y0Z0 would then move you to machine zero. If you then zero your fully zeroed in the work coordinates.

   Ill fix up the Charge pump activation, I noticed M4 doesnt show a signal on the diags till its toggled as well.. minor issue..

  The limits triggered will happen if you map the home switch to the X-- for example on X. The switch being triggered will stop any joggin in that direction as a side effect.

Have fun , preciate the feedback..
Art