Soft limits – again!

Home › Forums › TinyG › TinyG Support › Soft limits – again!

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • December 23, 2015 at 8:52 pm #9189
    leversole
    Member

    I guess I am going to have to give up on soft limits…I changed my limit switches out to NO type so I can use the touch plate feature in CP…whenever I enable soft limits, using the touchplate feature alarms the board. I jacked with it all day and, using cool term, figured out what was going wrong…it dies on line 4 below when it tries to move Z2…the alarm is axis beyond limits or something to that effect…the card then has to be reset…
    G21 G91
    G38.2 Z-20 F50
    G28.3 Z6.5
    G91 G0 z2
    I have attached a mm dump from the TinyG…

    TinyGConfig

    • This topic was modified 8 years, 10 months ago by leversole.
    December 23, 2015 at 9:43 pm #9191
    cmcgrath5035
    Moderator

    Without digging into your parameters at the moment, a somewhat philosophical question – What soft limits would you expect tinyG to ‘enforce’ after a probe?
    Seems if you have Zmax=0 and Zmin set to something, perhaps like -40mm, and then run the probe, when the probe resets Z0 to the material surface, the probe will be at the soft limit and you get the result you are seeing.

    I am not sure , either, if the Z axis remains “homed” after a probe, wit soft limits on or off.

    So question is, are soft limits inconsistent with homing and probing?
    Seems to me that after a probe operation, a new Zmax needs to be set, at Z0 + a safe height, then perhaps soft limits reenabled, ?

    December 23, 2015 at 11:44 pm #9192
    leversole
    Member

    Maybe soft limits ARE NOT compatible with touch plate?!? I thought though, that Chillippr was written specifically for the TinyG? Why offer something that does not work? Maybe John L will chime in…

    December 23, 2015 at 11:59 pm #9193
    leversole
    Member

    I hope you get a chance to look at my parameters, as there is still an issue with homing and the touch plate…after a fresh power up of te TinyG, if I use the touch plate, it works as it should, if I home the machine, then run the touch plate, the Z axis runs up (backwards) for a short distance and then stops as if it hit the touch plate, if I run the touch plate again, it works correctly (runs down until touching the plate, then setting the Z to zero)…this is repeatable and happens the same way every time…after the second time, it will work as it supposed to until the next power reset…maybe the touch plate is not compatible with homing?!?

    December 24, 2015 at 9:13 am #9197
    cmcgrath5035
    Moderator

    First, configs
    * Interesting that $sl=0, I’ll assume you had turned it off before $$
    * Try $xjm and $yjm to 10000 (yours are 51)
    *** This will provide ‘snappier’ stops as homing switches are activated
    * You might bump $xlb, $ylb and $zlb up to 20mm
    *** Depending on type of homing switches you have, 5mm may not be enough to clear a switch
    * Other parameters look OK.

    Second, the more I think it thru, the more I wonder if (Homing and Soft Limits) are compatible with (Touch Plate and Soft Limits). You are among the first to extensively experiment with Soft Limits, and I would not assume that the touch plate implementers considered soft limits when developing. There is a chance that Soft Limits did not even exist at the time, it is a realtively new feature, I think.
    The issue I see is that Soft Limits depends on axes first being homed.
    each axis has only one set of min and max values. Using your machine as an example, if you home using the homing cycle and switches, the appropriate Zmax=0 and Zmin=-76.2. But if you use the Touch Plate to find a work zero then force Zhome to be at or near the touchplate level with G28.3, the appropriate Zmax might be +20mm (to give some space for safe tool travel ) and Zmin could be =-56.2 (which of course makes no sense if you are milling a PCB).
    The probe command G28.2 could turn off Soft Limits, set the new Zzero, then change Zmax and Zmin and turn soft limits back on, but I see no indication it was designed that way.
    I believe a TouchPlate with soft Limits widget could be implemented as well, following the above logic.

    Third – The behavior after reset you describe might be a bug.Try setting $zzb=10 and rerunning. You have Zero backoff set to 1mm, which may not actually be enough to clear the switch, The intent of the Zero Back off setting is to ensure that the switches are not activated (are open in your case)

    December 24, 2015 at 8:04 pm #9198
    leversole
    Member

    Thanks for your input Carl…with the jerk settings as high as you recommended, it seems like it is tearing the machine apart, that is why I keep lowering them! 1mm seems perfect as I can here all of the switches “click” on the initial 5mm move…never had one fail to clear…

    I was looking thru the touchplate code, I don’t know java, but I used to be pretty good with VB and C++, so some java is familiar, I found where he is actually sending the commands, and I already thought that if the touch plate turned off soft limits and back on when finished that would work pretty slick…I do not understand the fork or really any of the open source stuff so I will have to learn…if you fork something public like that, how does it affect only your own machine?!? Anyhow, lots to learn and thanks again…

    December 24, 2015 at 10:42 pm #9199
    cmcgrath5035
    Moderator

    I guess, in an analog sort of way, I would say if you can hear the switches click, then you are good to go.
    I use proximity switches, no click.

    Just to be clear, you are most likely looking at javascript, not java.
    I had a tough time as well, the tools are not quite as refined as the C/C++ world and the uber flexibility sometimes confuses. JohnL is a much better sounding post than I am.

    A key question you might want to very clearly understand, which I don’t, is what might cause a given axis to loose it’s “homed” status, thus cancelling soft limits.

    Another thought, not well formed in my head but always bouncing around: What good are soft limits if the predominate cause of violating soft limits is a Gcode that causes tinyG to loose track of where the spindle really is? Soft limits are logical limits, and if logic has failed due to hardware or code deficiencies, what is the probability that a logical soft limit failure will actually help?

    Good luck with your project

    December 25, 2015 at 12:34 am #9202
    leversole
    Member

    Your last paragraph hits the mark! If you are going to crash because of lost steps, soft limits are not going to do any good…I wanted to set them up to prevent fat fingering! I cannot count the times that I have had the jog set to ten inches (for the X or Y), then jogged the Z before changin it to something smller! or been at the far side of the machine and jogged right instead of left!

    Thanks again!

  • Author
    Posts
Viewing 8 posts - 1 through 8 (of 8 total)
  • You must be logged in to reply to this topic.