Advanced Search

Oh, I already did that as I mentioned a few posts ago. However, I didn't mention in the latest post that trying it in Legacy boot doesn't get me past the initial root filesystem error. No clue...

I will attempt to try some different RTAI versions and will also see if I can try it out on a different computer altogether. In the pinned RTAI thread I saw some talk of unstable versions and on differing behavior across different motherboards.

Will give it a few more tries before I look into buying new hardware, lol.

EDIT: Did not see your last response as it seems you posted it as I was typing mine. I have no clue about EtherCAT on RTAI at this point as I am simply trying to boot w/ RTAI for the sake of seeing if the latency spikes issues are resolved. The idea was to go back to EtherCAT testing as the next step after this is settled. Or do you mean that RTAI in general is entirely incompatible with EtherCAT and I shouldn't pursue this line of thinking at all?

Wait, i just saw the tittle, EtherCAT did not work with RTAI at all, are you sure it works now?

Try setting the BIOS in Legacy mode or with CMS enabled.

Unlogic wrote:

While doing the setup I noticed that the four analog inputs on the 7i84 in mode 1 can still be used as normal digital inputs which makes the change from mode 0 to mode 1 very simple.

It's hard to not fall in love with these Mesa cards. So far they have managed to handle everything I've thrown at them.

All that, and the best support on planet earth from PCW.

7i73 has low voltage IO as it is made for control pannels, so get the 7i84, it should also do 2 of MPG in Mode 1 or 2.

SteveatSteelCut wrote:

It's a lot faster than the float switch, ...

Yes it is and that is unfortunate as the probing is done twice every time at very low speeds.
I am beating a dead horse, as i have insisted from the very start of PlasmaC that should not be implemented at all, only a single probe.
On a plate with over 200 pierces, probing time is in hours, no matter how fast the Z axis is.

i also think there are 2 different versions of the closed loop drives. i see a price change on these.
i think the more expensive of the two has more tuning features and i have the lesser of the two which does not have some of the same settings. I only have 6 settings in the Gain adjustment subcategory. i have a torque feed forward setting that is there to take some of the strain off the PID loops

so all i have in the gain adjustment is
position loop Kp
velocity loop Kp
velocity loop Ki
torque feed forward
position loop filter
velocity loop filter

it seems like the torque feed forward is a % value and taking a guess it looks like it matches closed loop holding current in the extended setup section.

it seems like the closed loop holding current setting is a way to take load off of the PID at the expense of operation temp of the motors and efficiency on the power consumption.

at his point i am not seeing anything that would adjust the following error unless it would be the filter settings? i have no settings in the speed, torque control of the main section? maybe because this dive is not the higher end version?

I managed to figure my way around the conflicts, I was interpreting the docs incorrectly probably. Also not sure what exactly I did differently, but after a few reinstalls through varying methods, I got past that error and onto a new problem.

Upon booting to the RTAI kernel, the display manager service fails to start entirely. I am able to press Ctrl+Alt+F1 to log into a tty, though and uname -r does tell me I am booted to the correct kernel. I am still unable to start the display manager or the desktop environment at all, even manually. Interestingly, when I start latency-test, a kernel panic occurs and the system freezes up entirely. I am attaching an image.

All this on UEFI boot and on both a vanilla Debian 12 install and an install of the LinuxCNC 2.9.3 ISO as well.

Quite honestly, I have no idea what to think.

So basically it looks like it does what I am looking for. 

The biggest limitation is my lack of understanding all the functions of the software.

Are there any help files or tutorials for viaconstructor?

It probably happens like this. When I go to the mdi page, the required position is sent to the servo driver and if the servo is not in that position, it goes to that position very quickly. I think I should try to synchronize the position by examining the manual you sent.

I tried but the speed does not decrease. Too fast and the driver fails.

i'm using matching software for the Stepper- online drive's. but the quick start PDF Leadshine has goes into a lot more detail that is the information i need.at this point i have done far more reading than tinkering.

the Leadshine pdf gives you a good understanding of the layout of the software. The tree on the far left helps you sort the 8 main sections of drive settings the first main section looks to just basic info pulled for dip switch settings and the last main section if factory settings. so the main focus should be the 6 other main section and the subsections in those 6 main sections.

it looks like you might be able to record the wave form from test moves. I am guessing helps adjust the PID but that is a bit beyond my understanding at the moment.

bottom line is there is a lot of stuff going on in these drives and probably why you don't see to much on the topic.

Have you tried lowering the 'MAX_VELOCITY' value in the [JOINT_1] and [JOINT_2] section of your ini file?

gmoccapy sets it itself, I did not use any method. For example, when I enter the mdi page, if the positions of y1 and y2 are not equal, the y2 motor moves.

Currently, the y2 motor is going to the y1 position very fast. I should reduce this speed, how can I do this?

It's not clear to me what method you use to move the y2 motor to the y1 position.