Advanced Search

Hello,
I tested on an Orange Pi5 Plus and noticed that the jerk was not always maintained.

the ini and the nc code in the appendix.
In addition, an NC code generated by a CAM where the jerk is also not maintained.. 

Edits: Photo link updated:
-Wire labelling and diagramming under way.

****************************************************
Original post, minor edits for clarity.
****************************************************
Hi I think my machine is strictly speaking, a router, but its kinda beefy.
photos.app.goo.gl/pLPEiGzPKKzL9mCc6
I want to upgrade it to have a BC swivel head plus a couple of rotaries. This will involve adding a minimum of 4 servos. I need to rework the wires for sanity and the conduits need replacement (cracked, broken and generally worn). 

I think most of the connections are correct, but, it is clear that the machine has been tinkered with by someone who probably didn't pay a lot of attention to documentation or detail, so I think its important for me to double check connections.

As of now, I am thinking of:
(a) confirming that the Fagor controller works and put it for sale.
(b) Remove all wires one at a time, labelling then, and then migrating the entire contros to Linuxcnc, keeping drives where possible.
(c) Adding a second junction box so that I have more space for fingering
 
Power is a complete other discussion. I do not need a 16hp spindle. Have to confirm that it works, and then downsize. I also do not want a swivel head large enough to handle this massive spindle.

The goal is to have a machine that is a full-sized prototyper. I want it to have plenty of capabilities, but lower throughput. I'm doing higher value, highly customized, runs of 20 to 50 per part. Not more. This is not a run day and night auto-loader setup.

Please, would love some comments and general feedback. Happy to add pictures to the album. 
 

Manche Sachen muss ich noch machen Bsp: Beleuchtung des Bedienfeldes , dies sind alle Ausgänge der Mesa 7i73

Moin in die Runde,ich bin gerade dabei, meine Maho 500 W4 zu retrofiten und habe schon einiges geschafft, stehe aber bei ein paar Punkten auf dem Schlauch. Vielleicht kann mir ja hier jemand weiterhelfen oder mal über meine Konfiguration schauen.Bisher erledigt:

• PID-Tuning der Achsen gemacht
• KSS-Pumpe lässt sich ansteuern
• Werkzeugwechsel funktioniert
• Bedienfeld läuft soweit

Hardware:

• Achsen: über Indramat 3TRM2 mit analogen ±10 V Signalen
• Spindel: über Indramat 6TKR, ebenfalls analog ±10 V
• Mesa-Karten:
  • 7i97t
  • 7i74, daran hängt 7i84U
  • 7i73 fürs Bedienpanel

Meine Fragen / Probleme:

1. Spindel:
   • Ich habe versucht, die Spindel über Tipps von ChatGPT und verschiedenen Forenbeiträgen sauber einzustellen. Ich habe eine 2-Gang-Schaltung (bzw. 3, wenn man Neutral mitzählt), wie es auch User Dowerna in seiner Konfiguration gemacht hat.
   • Problem: Die Spindel dreht zwar, aber es läuft trotzdem nicht vernünftig – sie ruckelt beim Drehen und klingt alles andere als flüssig. Ich weiß nicht, wie ich das noch sauber hinbekommen soll.
2. Mesa-Konfiguration:
   • Ich verstehe nicht ganz, warum hier PWM-Gen-Pins genutzt werden und nicht die Analogen-Outs, obwohl die Mesa 7i97t 6 analoge Outputs hat.
3. Konfigurations-Check:
   • Kann mal jemand über meine Konfiguration schauen und mir sagen, ob das so passt oder ob manche Sachen besser gemacht werden könnten?
   • Beispiel: Ich habe die MPG.hal nach meinem Wissen angepasst, die funktioniert bisher sehr gut, aber vielleicht gibt es noch Optimierungspotenzial.

Freue mich über jeden Tipp oder Hinweis! Danke schon mal vorab.

Englisch below:
 Hey everyone,I’m currently retrofitting my Maho 500 W4 and have made good progress, but I’ve hit a few points where I’m stuck. Maybe someone here can help or take a look at my configuration.What I’ve done so far:

• Completed PID tuning for the axes
• CNC pump (KSS) can be controlled
• Tool change works
• Control panel is running fine

Hardware:

• Axes: controlled via Indramat 3TRM2 with analog ±10 V signals
• Spindle: controlled via Indramat 6TKR, also analog ±10 V
• Mesa cards:
  • 7i97t
  • 7i74, with 7i84U connected
  • 7i73 for the control panel

My questions / issues:

1. Spindle:
   • I’ve tried setting up the spindle using tips from ChatGPT and various forum posts. I have a 2-speed gear (or 3 if you count neutral), similar to the configuration from user Dowerna.
   • Problem: The spindle spins, but it’s still not running smoothly – it rubs/jerks while turning and doesn’t sound or feel fluid. I’m not sure how to fix it properly.
2. Mesa configuration:
   • I don’t quite understand why the PWM-Gen pins are being used for the spindle instead of the analog outputs, even though the Mesa 7i97t has 6 analog outputs.
3. Configuration check:
   • Can someone take a look at my configuration and let me know if it’s set up correctly, or if some things could be done better?
   • Example: I’ve adapted the MPG.hal as best I could, it works fine so far, but there might still be room for improvement.

I’d really appreciate any tips or advice! Thanks in advance.My configuration is linked below.

Working on a switchable kinematics for my lathe, using the millturn example on GH.

(more here:  forum.linuxcnc.org/10-advanced-configura...r-than-x-axis#344888 )

One part of the switchkins process are a pair of M-codes: M128 and M129.  I gather these are used to change the axis max/min limits when the respective joints are swapped.

Problem is I don't understand the Bash language/terms/syntax well enough to make sure I'm modifying it appropriately for my config.

Here's the full M128:


Question(s):

1.  Do I need to change (INI_FILE_NAME) to my particular ini file name, or is this some sort of 'points to the configuration's INI file regardless of name' programming thing?

2.  Under #RESTORE, there is the foreach | $coords... part.
- Does this call, or otherwise ensure, everything below applies to all values after set coords "x y z" at the top?

- Does set L [string toupper $l] somehow change the lowercase "x y z" to upper-case for use in the ::AXIS_[set L]... part?

- What is "catch"?

3.  Could this have been written more simply?
like: hal setp ini.x.min_limit [set ::AXIS_X(MIN_LIMIT)]

and repeated for desired axes?  With no "set coords" or "foreach" or "set L" needed?

Bonus Question:
- Other than to confuse non-programmers, what purpose does this complication serve?  Is the intent to make it so the end-user only has to change the coords "n n n" values and everything still works?

Here's another Halshow screen shot showing the index input.
 

Here's a Halscope shot of it.
 

I have a 4 axis router with step/dir servos using a 7i76EU+7i85 to command the servos and read the encoder feedback.  I have the servos reasonably tuned running closed loop.  But I'm seeing a following error spike on homing to index for the W axis (XYZW) on the first attempt at homing the machine after power up.  After the initial encoder position reset on the homing index it will home fine the 2nd try and works after that. If the machine was parked at the W-zero location when Linuxcnc is turned off and started it does not do the following error.  For some reason it seems like Linuxcnc is not ignoring the F-eror on the encoder index reset for the W-axis like it should.  Any Ideas how to fix this?

Here are my main hal and ini config files.
 

Thanks it should work fine with the 2 RCPWMs and the sserial on TB4

Small update:

I aligned all docs with the actual runtime behavior and added a practical diag.sh guide.

Quick start is now simulation-first and profiles are clarified as declarative contracts.

Some parts were previously implicit (in my head), so this pass was mainly about making those assumptions explicit and removing ambiguity.

No changes in runtime/semantics — just making it easier to use and debug.

Now waiting for real hardware feedback.

Ok, it is reset of latched estop.
And how to realise reset button for external errors, for example drives errors or external devises (torch head or edm generator)?

Also i couldnt startup halui component, but somehow flexgui send power on signal to drives , i want to know how to use the same logic and send signal the same way

It doesn't sound pll-related if you get clunks when idle/standing still.
You can try and record this. If the clunks are from commanded position disruptions
you should see it as disruptions in position feedback.
You can record long sequences to try to catch it. Use the "sampler" hal component.
Possibly add a ddt component to take the derivative of the position feedback.
It will show some spike/peak when a clunk happens, if there is something with the position.
Plot the data, and evaluate.
 

I know Windows 11.1 uses a GPT partitioning. Everything I've read indicates that if I used Rufus to make the USB live stick on a GPT partition that I could do a dual boot install. I turned off bitlocker on the Windows files, turned off secure boot, turned off fastboot under the power settings, and shrank the windows partition to give me 50GB for the install. The install asked me for a missing file related to a sound card and didn't recognize my network (wifi). I selected manual partitioning and did not see any of the partitions other than the USB stick. The live version booted and ran fine. Running gparted showed the EFI partition with an error triangle alongside. I suspect this has something to do with GPT but I can't figure out what.
I'm trying to install this on an HP Laptop. Any thoughts?

Nice, thanks for testing it

And yeah, honestly you did a better job than the docs there
At that moment they were a bit messy and could easily lead in the wrong direction.

I just went through and cleaned them up properly (quick start is now simulation-first, no profile/topology needed, and observability is aligned with what diag actually shows).

So what you ran is basically the intended path, just not well explained before.

If you get time later to try it on real hardware, that would be really interesting to hear about.

Appreciate you taking the time to try it.

The config was previously working just fine. I dont know what changed that caused this.

Ill check dmesg again when im home.
As for the working state after the DRO populates... Its hard to say. I can definately move every axis just fine, but theres a 50% chance that clunks occur. Sometimes more clunks, sometimes less clunks. Sometimes they even appear mid-cycle which is a no-go.

ive swapped out all the ethernet cables now, re-imaged my lcnc host, even used a new host without any success.