Advanced Search

The ping times are not terribly important _unless_ packets are dropped due to timeouts.
Default timeout setting is 80% of the servo period so those times look fine for a 1 KHz
servo thread.

Ok so I just checked and yes it has them on both sides. They are contact switches and they both work. I have pictures but I don't know how to send them.

Have you tried isolating the CPU cores?

I have the same processor, but a different motherboard, and the results are better.

It is possible that your problem is caused by something other than delays, but it is worth trying to improve the result.

 

The E-Stop was cool!

Yeah, where to start looking?
When you turn the servo motor (if possible) and look at the relevant pos-fb pin in halshow, does the pin number count up/down? What value does the lcec.0.xxx.controlword have?
Vendor id and product code from "ethercat slaves -v" match the vid and pid fields in ethercat-conf.xml?
Any error messages from linuxcnc at startup, or errors in "sudo dmesg" that can help?
 

Use "Output 3" and "Output 4" on the digital out connector to control the servo.

The MX3660 has its own breakout borad that outputs a filtered (buffered) signal for pins 8 and 9.

Connect your servo to those.

There's no need for an extra breakout board.

S-Curves would be nice.

Constant acceleration causes machine wobble.

Explain.

Did ya fix it?

Hello, first, you need to get yourself a CMSIS-DAP probe. You can buy one on eBay, Amazon, or covert an ST-link probe. Here is a line to the thread how to do it.

forum.linuxcnc.org/18-computer/44828-rem...rd?start=2050#321860

Once you have the probe, you need to upload the firmware, I will send you the commands later when I have a moment to take a look into my notes.

Once you have the new firmware, then you can follow the instructions from my post above:

forum.linuxcnc.org/18-computer/44828-rem...rd?start=2140#327240

Happy to help

Hi can someone tell me if mine are broken.
i cant understand these. They are stock part on the cnc machine.
And would like to keep it stock. 

here is one of the switches. Attached.

I use an new 7i76U board. And can easy just trigger the end stop pin with a wire from positive source (+)

Since this gives reverse. I did think to add a relay.
but the relay just stays on.

How do these work. I get constant 24 volt.
but when messure resistanse it does from 4ohm to 40ohm.
so something is happening. When it switches and led goes on/off too.

Anyone know these.
https://photos.app.goo.gl/jcqMZvWbBJQ8ZE1p6https://photos.app.goo.gl/jcqMZvWbBJQ8ZE1p6

I've mentioned a few times that my DIY mill is very old (32 years), and probably some of you have looked at my pictures in the past.  But bear with me -- this is the last post I'll likely make about this mill.
 

I'd been making a bunch of improvements lately (new RPi controller, mesa card interface, M3/M5 spindle control, M7/M9 vacuum control, some new hold-down techniques, automatic homing, better E-switch, probably some other stuff).  It had been a while since I'd spent much time on it, so I was having a lot of fun.

My wife asked for help on a project, and I'm always happy to oblige, of course.  We designed what she needed, and I set about cutting it.  After a couple of false starts over a couple of days, the rough draft carve looked good.  She needed two good copies, and the first of those looked great! 

The second one, sadly, broke my Z axis.  Not the project's fault, and not the mill's -- we'd have to blame the designer:  me.  Not only did the Z axis driven coupling come loose from the carriage, but because of the poor design, there was no way to fix it without a really huge amount of dis/reassembly.  Which I really couldn't face.  One too many times, straw that broke the camel, etc.  I was done.

So I've ordered a new mill -- I'm expecting a Sainsmart Genmitsu 4040 Pro Max on my doorstep any day now.  I'm very excited about it.  And now if it breaks, I can blame someone else!

But I had to make space.  This only took an hour or so, and it was actually very cathartic.

 
 

A note about the design.  Using drawer slides to support the X/Y carriages, as the original plans called for, worked far better than you'd expect.  High quality slides are specifically made to keep drawers from sagging...  at all.  So if you put them on their side, and use them over a shorter distance, they're not true "linear rail" quality, but they're very, very good, especially for a DIY'er.  They didn't work well on the Z axis because unlike X and Y, Z undergoes forces in all directions, not just side to side.  I eventually rebuilt that axis using actual linear rails, but the rest of the Z design wasn't so great.  Ahem.  As we've learned.)

Anyway, I think one of the things I'll miss most about that mill is the limit switches.  You've gotta admit, this is cute, right?  
 

Good day all. I seem to be having a bit to trouble getting started. I have an NVEM V5 controller and RPI 5 with LinuxCNC installed. Is there anywhere on planet earth that has a start to finish, all in one place, all on one page, set of instructions or video that will tell me what version of what software from what source to use in what way using what commands and what wires to hook up where to get it all done with the hardware I have. I have gone through 100's of posts and can grasp the general concepts of what needs to be done but it's a small and sometimes incomplete piece of info from 1 post and then one has to read 5 or 20 more before the rest of the info is found. In the end, and totally a flaw of my own, I seem unable to pull all of this scattered info together and have it make any sense as to how to get from start to finish without burning something up or creating an unsalvageable mess.

I found a video by Scott 3D on youtube but it is 3 years old and several key points are just blown by with no explanation and no info on exactly what versions of what from where is needed today.

Is there a single set of complete instructions with valid links in existence anywhere for setting up an NVEM v5 with LinuxCNC?

This is my last effort to get the ping down. I set the Cool Option to thermally optimized, a cool&quiet option in BIOS. Other settings would bee acoustic optimized or full throttle. I hope it does not affect anything of cpu clocks.
 

rtt min/avg/max/mdev = 0.105/0.126/0.259/0.002 ms  what do you think?

Greetings, Hanno

 

Hi, I tried a lot the last days.
I set the P to 500 as you said, I made the realtec optional kernel settings, i took a look for Energy efficiency and did not find anything of this, I set ff2 back to 0, and in the first try (an air run of my last part) it seemed to run - but 2 min after start the motor-0-position error came again, right in a direction change of the path. As before.

That was very disappointing, so I decided to get another PC from the shelf, a Lenovo m910q with 16GB, i7-6700T 4C/8T and put the SSD from the other in it.
The machine started at once, no problems to fix, I declined all BIOS settings that are described in the latency thread. This jounger PC has WLAN for the Internet and a I219-LM Intel network card, that should be better, somebody told me.

I did the coalescence workaround in etc/interfaces, I made the ethtool working (sudo ethtool -C enp0s31f6 rx-usecs 0), I made the kernel options:   quiet irqaffinity=0,1 isolcpus=3 intel_idle.max_cstate=1 i915.enable_rc6=0

The outcome was not bad, but worse than before, regarding the ping times. The realtec gave me 0,056-0,070, the intel gives out:
rtt min/avg/max/mdev = 0.165/0.189/0.307/0.052 ms which is bad in my eyes.
Even if i kill the WLAN adapter, it does not much better:
111 packets transmitted, 111 received, 0% packet loss, time 112114ms
rtt min/avg/max/mdev = 0.108/0.164/0.199/0.011 ms

What do you think about it? should I go down to the machine and try it out or should I do sthing else?

Thanks for your advice, I appreciate that much!
Hanno