Advanced Search

I can see min and max around 70 us. I do have this numbers all the time with a lot of lenovo thinkcentres tinys (m73, m93, m710q etc.) without having troubles.

As long as you do not have spikes in the range of ms (realtek network issues) or added delays like the intel network delays everthing is fine. As long as you are happy with 1 ms servo period of course. Ethercat runs in the servothread, so 1 ms is totally ok.

I always make a plot with the hal scope of the hm read time (if we are talking about mesa cards and added ethercat) to visually see everything. One can see network times and jitter on the fly while doing things (gears, surfing, intel nic delay playaround etc.). Just keep in mind that you see cpu ticks and you have to devide the numbers (100k to 400k range base and peaks little higher caused by jitter) by the cpu tact.

Spindle Speed Control.
Now I have the spindle encoder in place, I set up a PYVCP read out for actual spindle speed, which agrees precisely with a handheld rev counter. Even registers  negative in reverse. so I believe the encoder is working ok in quadrature.
Problem is that I now see that the commanded speed and actual speed are way out, at s2000 I am seeing almost 3000rpm.
(The spindle speed override is at 100% in the GUI)

My motor has max speed of 3600 and the VFD outputs max spindle rpm of 3600 at 8 volts. Increased volts are ignored.
I've checked the relationship between rpm and volts between 0.24V(350rpm) and max speed, and it is linear.
I have tried changing the ini file [Spindle_0] OUTPUT_SCALE and this helps, but then the slower speeds are shifted.

So I think its clear the controller feedback isn't functioning, and I'm failing to see why.
I don't think I need super dooper speed accuracy, but I do want to be able to rigid tap.

Would an expert please cast their eyes over my attached config files and point out where I am being stupid 
Thank you.


 

By experimentation, the Y and X offsets can be entered pressing tool touch off with the wrong axis selected. On a 3 axis mill, as far as I can tell numbers in these columns are ignored.

I just added file load buttons for frequently used files like probe routines so the path shows up in the plotter before you run the routine.

Also don't forget we have a IRC channel #flex_dev on Libera.Chat.

Flex GUI just keeps getting better and better and more flexible.

JT

Move the motors by hand while LinuxCNC is running but not enabled, do the numbers on screen (Gremlin) change?

if you have two (or more ) Tangnano, you can try it, you need the 'dev' branch and
2 free pins on each tangnano (RX / TX).
For short distance, you connect this pins direct (RX<->TX) or using RS485/RS422 converter for long distance.

Rod, he has huge excursions on both sides of the graph, those are not shown unless using --show, and those are the ones causing latency errors and link drop.

Look forward to seeing how that goes, Tang Nano's are about 30 bucks AUD, two of them are still cheaper than any FPGA with a similar total pincount. I'm being selfish but that also means I stick with the same FPGA hardware for every machine I can possibly imagine making. 

Let me know if I can help with testing, I'm currently retrofitting my three machines to closed loop to the controller, so there's room for some hardware experimentation. 

A broader opportunity could be to try and get involved in plant automation. This will get you using a lot of fieldbus technology like ethercat which is transferrable to CNC machines which are only a very small subset of this automation industry. 

Removing the Wifi board is probably good for latency.
The DKMS driver can usually be installed with 

sudo apt install r8168-dkms
but you do need to have non-free in the apt sources. This is included in the Linuxcnc ISO

Seems the card uses a R8168 interface. The R8168 requires a DKMS driver to get acceptable RT  performance. Not sure how that goes with a Pi. I would not use it for ethercat but would be fine for general network use.

Sometimes the shape is because of the number of sbins samples. You have 20 mand most use 1000

try
latency-histogram --nobase --sbins 1000

My observation is that Ethercat is faster than Mesa hm2eth. 90 usec will be fine If your pings are good

Problem with "Joint 0 Tracking Error" and Real-Time Delay: Resolver-Based System

System:

LinuxCNC with PID control

Resolver-based feedback (no encoders)

3-axis machine (X, Y, Z)

The HAL/INI structure includes: Ibarmia core servo.hal, Ibarmia MotencMotion.hal, motenc.ini, etc.

Main Problem:

When attempting to move any axis (especially the X axis), I immediately receive the following critical errors:

Joint 0 Tracking Error

The axis does not move, and the error appears as soon as I attempt to move it, regardless of the velocity or acceleration settings.

Measures already taken:

Latency Test

Maximum jitter on the base thread: around 16,600 ns

Increased the BASE_PERIOD value from 50,000 to 100,000 to reduce RTAPI errors.

INI Tweaks

Reduced the MAX SPEED and MAX ACCELERATION values ​​on [TRAJ] and [AXIS_0] to very conservative values:
MAX SPEED = 5.0
MAX ACCELERATION = 300.0
FERROR = 100
MINIMUM FERROR = 10

It appears to boot, but after 10 seconds, it fails.

Try 2.4 GHz. 5 GHz is not as reliable in an industrial setting and has much shorter range. I turned off all my 5G wifi

Failing that use a wifi USB dongle that has an external antenna and get a SMA bulkhead mount cable to extend it outside of the cabinet

Enjoy your trip, and no, not network, the Axis GUI.