Advanced Search

I made a few minutes to take a look at this tonight.  

I do see the behavior you describe, although the order is not the root cause. 

In this line:  github.com/LinuxCNC/linuxcnc/blob/bd9034...ialog_widget.py#L201



it's this part that is causing issue:



The icon value is set correctly to 'WARNING' on the go-in, but because but because icon is of type str, the second condition "isinstance(icon, str)" evaluates to True, and therefore never gets to check for 'WARNING'.

Perhaps it’s meant to be a catch all, in that case I agree it should move to last, but I want Chris to weigh in before I change it as I'm not quite sure of the purpose (if typo should we tell the user there’s a typo?).

Yes, I am using (I used) Virtual Box as VM for test.

Is the NativeCAM path correct in the ini file?
(you could post your ini file)
Is NativeCAM runnig standalone from the command line?

As suggested by HansU, you could try to run LinuxCNC from the command line to see the errors during startup.

Wollongong. Not too far!

There are no limit switches. Further, I run a 3 axis file successfuly and used this addon to obviate the need for wiring the Y axis together. I believe it uses 2 stepgens successfuly.
Here is what I left out:
HOME_SEARCH_VEL=0
HOME_LATCH_=0
HOME equals to HOME_OFFSET
HOME_SEQUENCE=-2

Hal file:but further downandNot going further before clearing some stuff up....or
-does the machine have home/limit switches?
There are none in hal, so if there are no switches, you can NOT use tandem axis/joints
-you can use the same stepgen wired in hal to both Y drive pins (as you have them now)
-you can use a normal XYZA config and wire the two Y drive inputs together in parallel
--both versions above are 4 stepgen/4 joint configs, not 5
-use StepConf wizard to make a new XYZA config, then just add the pins in hal, NO editing beyond that.
If there are switches, and you want to have tandem axis/joints, due to limited number of inputs on a parallel port, that requires using at least two inputs as the tandem joints can not be on the same input. Also requires using LUT5 component (i think) to make the homing of all joints that are on the same input.

My guess is that HOME_SEQUENCE is missinghttps://linuxcnc.org/docs/html/config/ini-homing.html#sec:homing-section

(you need as negative number for the Y joints)

Plus, there is a typo in the DISPLAY section:

JOUNT = 5
 

Yes, I call it by connecting STATUS 'general' message and the ACTION.CALL_DIALOG:

A LcncDialog widget with matching launch-ID is defined in the ui-file. I have already checked whether the value of the 'icon'-variable is set correctly right before the statement I have shown in my first post (first code box). It's all fine. Looking at the original code, I think, because 'icon' is an instance of string in all cases, the 'WARNING' and 'CRITICAL' cases can never be reached. So I changed the sequence in order to give WARNING and CRITICAL also a chance getting active. Am I right with my consideration or have i missed something?

I have this
forum.linuxcnc.org/39-pncconf/32400-mesa...pport?start=60#89231
A box with 17 resistors and 17 LED's, a male and a f-male port, nothing else.
Not versatile as yours, but it has also the option of connecting between the PC and machine and monitor in real time.

If anyone needs [2] a simple circuit to test parallel port pin by pin as either input or output, this is good solution.

No ICs and no transistors required.

More about here: forum.linuxcnc.org/18-computer/55803-a-s...el-port-tester-board



 
 

 

I wanted an easier way to test a new PCI parallel port card in all three modes (OUT, IN and X) in order to help with this topic here:

forum.linuxcnc.org/18-computer/54368-pci...-work-out-of-the-box

So I came up with this simple, yet "powerful", parallel port test board. The core circuit for each pin is very simple and uses only 4 components. No IC and no transistors required.



Every single pin can be tested as input and output. It is passive (no buffer) but I have also added a USB conector to provide 5 volts to 10k pull-up resistors (some cards won't have pull-up resistors for every pin). I used different color LEDs to represent each register: Red for Data, Green for Status and Yellow for Control. However I would suggest not to use yellow LEDs because they are not as bright as red and green ones.

It is handy to diagnose single damaged pins. Sometimes a given card has only one or two damaged pins and maybe they have got only input side damage (I have never seen this last case, but I believe it can happen). So you can swap pins positions knowing in advance what is still working.

(Btw, if parallel port masters here could suggest better values for resistors I'd be happy. Maybe 20k for pull-ups? I know some cards require more current to trigger control register pins when working in X mode, so maybe 490 ohms would be too high. But I also don't wan't to destroy non open-collector pins while investigating a unkkown card in X mode)

I didn't have any three position switches laying around, so I used four pin headers instead.

I will be making a single layer PCB version soon.

As it is: 
The LOWER position sets pin as OUTPUT and the led will lightup if STATE = HIGH
The MIDDLE position sets pin as INPUT with STATE = LOW (0 volts through 490 ohms resistor)
The UPPER position sets pin as INPUT with STATE = HIGH (it is actually connected to nothing, so it will just let the 10k pull-up resistor do it's job)





(I know, if one single atom of iron gets between this traces the will short circuit, but I had take as much time as possible to assemble it, so my wife wouldn't ask me to rearrange forniture... I also find it terapeutical to join blobs of solder, ONCE you learn the proper technique)

Here is my latest iteration of my attempt to straighten this out.  Some progress:
1.) Progam opens up (that's a start)
2.) XZA axis home.  Y axis won't home.
3.) Movement keys are changed; X same, Y-page up down, Z [] brackets up down and + - sign move z up and down
4.) If I start joging the axis I will get an error " must take out of E-Stop.  It's not in it for any reason.

I spent a lot of time going through the ini and hal set ups and am clearly missing something.
Would greatly appreciate any insight you could offer.

Thanks again

1. should be connected to z minimum limit in hal, will check later if i can
2. yes, they are OK
3. yes, remove one of the big power supplies, one of those is more than enough for 3 drives.
-
1. yes, use 8 or 16, not 10
2. no, those should do just fine.
3. no, it is still in development, make the machine work first, fancy stuff later.

Thank you for reporting back.

No, nothing new, it's been very quiet on the business side for quite some time so it's causing me huge mental stress after investing all that time and effort and money, but keeps me cool knowing i have plenty of stuff ready.
Just came from the old shop, was looking at starting to build a mill (finally) and a small lathe, and maybe if i can organize the rails and ballscrews a small wire EDM.
I have to much stuff that does not fit together nicely so i need to convince myself to grab the angle grinder and have a blast with it till stuff fits together!
As for servos, i still have some unused brushed motors, a huge brushless Siemens/whatever their name is, but no drives.
I do have plenty of RC servo drives, though, up to 8S/70A ones, but only small motors!
And to top it off in the "stuff that does not fit" i have analog RC googles but only digital RC camera!

With some help of the producer I got to the problem.
For me, my old drives that were working, were on software version 3042.
The drives that did not work, were on software version 30D1.

Apparently, you have to switch A+ and A- when you are working on software version 30D1. All works well after that adaption.