Advanced Search

It is possible to set up a linux system with no passwords, but it isn't completely simple or recommended (especially on systems with any internet exposure.)

First you can do away with a root password, by leaving that blank when installing. The user password is required while installing, but you can configure Linux to boot directly to the user interface by enabling automatic login, and the system will start without needing to input a password.

If you configured the system without a root password, the installer will have given sudo privileges to the user, and will promp for a password when a sudo command is input at the terminal.

With automatic login and sudo privileges for the user, you would only need to enter your password to perform administrative tasks that probably should be password protected. Otherwise you shouldn't normally need to user your password.

There are ways to even remove the requirement for a password for sudo tasks, but I've never tried and would have to google instructions how to do it. (not to mention that it's generally probably not a good idea.)

I installed linuxcnc 2.9.4
and problem solved,
lol

Sorry about Pxxx/Moning about passwords I am leaving M.S. because of how they handle the passwords. I have been building boxes since Aug of 1977 (Almost 50 years) and over 100 boxes's
why cant they just have a check box do you need a password yes / no and leave it at that
1 frustrated old man
Ken (5280)

Thank you for the reply (link to the Debian 13 install link after building a USB boot install flash drive and doing the install on the (older Dell laptop 8 gig memory)thee install goes fine till the password section which i have tried any number of password from 4 letters to 20 letters / numbers .? nothing get me past the enter login on the reboot . Every thing looks fine on the setup "but" ? I guess I am not to have this I have never understand why since I am the only person "WHY " I need a password" and the only person who will use this people have gotten crazy with the requiring passwords

Thanks for the help

Thanks to everyone for their comments. The overall take away from this is that you want a solution with minimal hassle, go with Mesa and pay accordingly.

Let me ask a related, more targetted question: what is the most solid Remora implementation both in terms of solid operation once configured *and* using HW that is solid and has a reasonable support expectancy. Novusun (sounds iffy), esp32 based? RP2040 (sounds experimental), some kind of 3D printer driver board? Other?

forum.linuxcnc.org/18-computer/38838-har...ency-tests-used-pc-s

forum.linuxcnc.org/18-computer/39371-res...ith-linuxcnc?start=0

Ok, I got this searching the forums and internet. So would be needed to test it to see eventually how that latency would be affecting it. Another question is if there is an better suitable industrial PC to buy ?

I mean do you have any recommendations on what motherboard and chipset is preferable for use with LinuxCNC ? I understand that this would be any motherboard
that Debian or linux kernel supports, but is there a some preferred hardware list or suggestions for a good LinuxCNC industrial PC ?

Tommy thanks. So Yaskawa would be preferred correct ? What kind of latency issues are there with EtherCAT ? Is it on the LinuxCNC side or due to network ?

1) yes

2) If you have a simple switch probe you simply connect a wire from the input on your board to one side of the probe, and run the other wire to the ground on that board. There are some basic instructions here for the Hal programming.
wiki.linuxcnc.org/cgi-bin/wiki.pl?Touch_Probe

There is not any real programming needed in Linuxcnc, it is ready to go once the Hal lines for the probe are added to your configuration. You might want to make some G-code probing routines to automate your probing process.

3) Most any probe will work. If you have access to a 3d printer you can even make your own probe like this one.
www.thingiverse.com/thing:4672365
Obviously the better the probe quality, the better your results can be.

Depends on budget, good ones are from Yaskawa and Delta, all the way down to cheap china ones that do work, but getting there might be very hard as usually they are not tested or tuned to the motor you get.
I am still skipping EtherCAT due to feedback latency issues.

Hello, good afternoon. I came across your post and wanted to ask you for some advice because I’m dealing with a very similar situation.

At the company I work for, we have an old Arburg 220-75.250 injection molding machine from the late 1980s. The floppy disk drive is no longer working, and I have been asked to find a way to get the machine back into operation.

I’ve been researching the machine, its control system and possible solutions, and from what I’ve seen so far, I’m starting to think that instead of trying to repair the original control, the better option might be to replace the control system and modernize the machine, potentially using something like LinuxCNC or another modern industrial control platform.

However, I have quite a few questions about how to approach a project like this.

First of all, how would you approach the retrofit from a technical point of view? What information would you consider essential to identify before starting?

I’m particularly unsure about the electrical side. How do you determine what the existing connections and signals are supposed to do? Is there normally an electrical schematic or I/O documentation that can be used to map the original controller’s inputs and outputs? If the original documentation is missing, how would you go about identifying and tracing the signals?

Could the existing sensors and switches on the machine be reused with a modern controller, or would they normally need to be replaced? For example, can the original position sensors, pressure sensors, limit switches, temperature sensors, etc. be connected directly to a modern I/O system, or does it depend on their signal type and voltage?

I’m also wondering about the hydraulic system. Would the original hydraulic valves and proportional valves normally be controllable by a modern system, or would these components also have to be replaced? How would you determine what type of signals the original valves require?

The same question applies to the injection and screw-position control. How would you interface a modern controller with the existing servo motors, drives, encoders or position feedback systems, if the machine has them?

Another important point is safety. How would you normally integrate the existing safety circuits, emergency stops, door switches, safety relays, etc. into a modern control system? Would you keep the original safety architecture wherever possible, or would you recommend replacing the safety system as part of the retrofit?

I’d also like to understand how you determine the required I/O. Is there a standard procedure for creating an I/O list for an old injection molding machine before designing the new control system?

Finally, if you have done a similar retrofit, I’d really appreciate any advice about the practical side of the project: what components you would keep, what you would replace, what information you would try to obtain from Arburg, and what mistakes or problems you would specifically look out for before starting.

If you don’t mind, I’d also be very interested in seeing any wiring diagrams, I/O lists, documentation, examples or photos from a similar retrofit that you could share.

Thanks a lot for your time. Any advice would be greatly appreciated.

Moved, thank you Chris.

Thank you for your detailed answers; i have ordered RS422-TTL converters to try everything out as soon as possible.

I checked the stepdir generator again in Verilog/GTKWave and tested a few edge cases.
I noticed that a pulse can be strtched if the velocity drops to 0 at the wrong moment, so I fixed that.
But that shouldn't have anything to do with your problem; it's more of a cosmetic change than a bug fix.
(in the dev branch)

can you post your json config please, i will check it.
you can also try to invert the step pulse in rio, maybe thats the problem.

Another possible cause could be that the deadband value is too high. Does the zero position drift, or does it remain stable with the same offset? How large is the deviation from the actual zero point (approximately 1, 10, 100, or 1,000 steps)?