From the original Synthetos forum (2011–2025)
Okay, I followed the homing setup instructions to a T! The machine homes correctly, but the limits do not seem to be working, I had a G54 zero set wrong and the machine drove itself into the limit switch and kept going! Also, as a test, I have the X axis maximum travel set to 22 inches, issue a G1 X22.5 and it went right to it...
$Home shows the machine is homed, and $sl is set to one...
Dont know what I am missing?
Oh yea, it also (as it was completing the job, it hit the switch repeatedly...I know you can set the xmin switch as homing AND imits, but I thought once homed, you could not go past the soft limits...
Hmmm, good question.
Can you capture a $$ parameter dump for analysis?
Can you describe how this happened:
I had a G54 zero set wrong and
Messing with zero set likely would reset "homed" status.....
I checked and setting G54 zero does not affect teh homed state...anyhow, here is the parameter dump. also, touch plate is doing weird things, like not going to the right height?!? eg,,,plate is 6.57mm thick, after running the z axis could be anything from 3.0 to 7 or 8, and after words, if I return to zero, it is 10+ mm away from the workpeice?!?
Hmmm, your parameter dump is a bit hard to decipher, I forgot to request you put tinyG into mm mode first so that parameters are shown in mm.
In this dump, $sl=0 and hard limits are turned off.
I'll assume you changed at some point.
A couple more fundamental questions I skipped earlier:
What OS are you running (It may affect proplem resolution suggestions){Winxx,MacOS, Linux}?
How do you communicate/send commands to tinyG? {CoolTerm, Chilipeppr, other}?
I suggest you review, carefully, this wiki item
https://github.com/synthetos/TinyG/wiki/Homing-and-Limits-Description-and-Operation
You have your Z axis set up wrong, if in fact you are homing your Z axis.
The tinyG implementation of Homing homes to Zmax, not Z min. And Zmax is set to zero, so for your set up, $ztn=-2.953in and $ztm=0, if you choose to Home Z axis.
The following would be helpful in proceeding:
1. Reset your tinyG
2. Make sure it is in mm mode
3. Provide a $$dump, call it "after_reset"
4. Run your homing cycle
5. Retest a move that you violate Soft limits, if you do enable them.
6. Provide another $$dump (again, mm mode) all it "after_home"
Grabbed two dumps for you, fresh from startup, then after homed...I see that $sl is off, maybe that is my issue, dont know what you mean by homing the Z wrong, I followed the wiki to a T, I have the limit wired to max as instructed, all three axis home properly.
Using Windows7 and chrome to run Chilippr...the dumps attached came striaght out of coolterm...
Here is another useful wiki reference:
https://github.com/synthetos/TinyG/wiki/Homing-and-Limits-Setup-and-Troubleshooting
In this reference you will see explicitly that $sl=1 is necessary for soft limits to be in effect.
Also note the settings for Z axis
[ztn] z travel minimum -95.000 mm
[ztm] z travel maximum 0.000 mm
By procedural design, when the Zmax switch is operated during a homing cycle, that position is set to Z=0 for G53 coordinate system, and G54 will also be 0 as you have no G54 offset.
So $Ztm needs to be set to 0.00 and Ztn needs to be a negative number for soft limits to work
In the $$dump you first posted, you had
[zsn] z switch min "1 [0=off,1=homing,2=limit,3=limit+homing]"
[zsx] z switch max "0 [0=off,1=homing,2=limit,3=limit+homing]"
I see that you have corrected that in the latest two $$ dumps.
I believe that if you correct the $ztn and $ztm values and set $sl=1,, then rehome, you should find that soft limits will work. Do note that Z movement must be to positions <0.
What I envision from your parameters is a XYYrZ machine with GT3 belts and an 8mm leadscrew for Z, perhaps an Ox or ShapeOko2 or Xcarve.
I appears to me that you may have entered your parameters in inch mode, which makes comparison to the recommended (or start-up parameters) somewhat difficult. What I am referring to is, for example, $zvm=787, rather than $zvm=800, a typical recommended value. I also note that you X and Y velocity settings are rather low (5000 or in your case 5080) compared to typically recommended value of 16000.
I also question why your $xjm = 25, vs. a typical recommended starting value of 5000. In fact all your jerk settings seem low.
Is there a particular setup guide you are following?
I am a bit concerned about the bogus value of $atn, which should be -1.0 deg.
You are not using the A axis, so it might be benign, but sometimes crazy values like this are an indicator of corrupted parameter store in EEPROM.
What tool did you use to set parameters?
Finally - I was able to read your latest two parameter dumps, but as bitmaps (rather than the text format you posted first time) they are difficult to manipulate. For example, I will often 'diff' your parameter set against a typical or recommended set (I work in Linux space), letting the machine, rather than my eyes, spot potential issues.
Thanks for your help, I will make the recommended changes...I used cool term to set the TinyG up, I thought I followed the wikis but may have missed something. How do you capture a readable text file from coolterm? Every time I try to capture, it makes the file all in lines instead of each row per parameter? I think I used JCNC to capture the first on in a nice two column format.
Well, I am so frustrated I cannot see straight! Spent 6 hours today trying to get limits to work. Here is the scenario, and it is repeatable (until the last problem)
From a fresh power on, with limits on $sl=1, and the new settings you mentioned (file attached) the machine home's as it should. If I use the touch plate in chilipeppr (G38.2 Z-20 F50), the TinyG goes imeadiately into alarm...the wiki describes the error messages, but does not list the command to return the error message! $stat just reports "Alarm"...messed with this for several hours, checked all of my settings (I cannot find anything wrong)...Now, if I turn soft limits off, the machines homes as expected, then I jog the z down closer to the table 25mm or so, but when I try to use the touchplate (g38.2 Z-20 F50) the Z goes the wrong way and stops after a few mm...If I run it again, it goes the right way and works as it should. So I gave up on using limits, reset everything and was attempting to run a job that I ran before, then ChiliPeppr went berserk (most of the text on the screen turned to machine code, and now the Spindle direction is flashing and no matter WHAT I try, I cannot make it stop or communicate with it (power off, unplug USB, reset, etc...)
I am at the end of my rope!
Please advise
Are you using the configuretinyG widget to make your parameter changes?
Don't. Use the serial port Command window.
Are you in mm mode or inch mode?
I suggest only changing parameters while in mm mode until we figure this oput.
You can run $$ in the CP Comand line widget, then highlight the buffer and copy, then paste.
Helps to clear the window first.
If a reset button on tinyG does not reboot it (after 10 second wait, your flash image is corrupted.
I'm not sure that the SPJS FW upgrade will work if tinyG is stuck in bootloader.
Try this:
https://www.synthetos.com/topics/help-please-spindir-flashing/page/2/#post-9096
You are not alone, several other threads with same garbled CP then death issue.
Root cause still eludes the devs
I was using CoolTerm to configure...i will try the FW Upgrade...
I have 32 biot windows...will this still work?
Thanks
I sort of doubt it.
CMD window avrdude (32 bit) should work, have that?
You can also try this:
1. Download this archive:
https://dl.dropboxusercontent.com/u/50261731/tinyG_WinUpdater/tinyG_440.20_Update.zip
2. Unzip it into a clean folder
3. browse to that folder in cmd window.
4 Read ReadMe.txt
Will give it a try in about thirty minutes, does my parameter file look okay?
Can not get it to run...even run as adminstrator for command prompt...here is the error
Ah so, that must be Win64 version of avrdude.
Hmmmm
This topic is from the original forum (December 2015). Your reply starts a new discussion linked to it.
Back to TinyG Support or latest activity.
