Ethercat HAL driver

nice that this works.

Scaling should be no big problem, i think that the grad /mm thing is only an Indraworks Gui (bug) thing. the parameters for linear axis and scaling exist in both sets soe and coe. maybe it is only needed to attach them to the initcmds.

next step: setting up an EoE gateway in lcec, for using indraworks in conjunction with Lcnc.

but i am out of the shop no till late evening.....

db1981 wrote: nice that this works.

Scaling should be no big problem, i think that the grad /mm thing is only an Indraworks Gui (bug) thing. the parameters for linear axis and scaling exist in both sets soe and coe. maybe it is only needed to attach them to the initcmds.

next step: setting up an EoE gateway in lcec, for using indraworks in conjunction with Lcnc.

but i am out of the shop no till late evening.....

This were really helpful, though i have a second PC for Indraworks.

It is only a Test Setup at the moment, my actual CNC runs under Windows with CS-Lab controller. I want to change it over Time to linuxcnc, this is first only to get to know it and learn how it can be done. Because of that i don´t have time pressure.

But i dont know much about the init commands or the Indraworks parameters.

And thats why i really appreciate your help

db1981 wrote:
If you halpin(float) is 5.123 with :
scale =1.0, Value to to the node = 5;
scale =1000, value to the node = 5123;
scale = 10, value to the node = 51;
scale = 0.1, value to the node = 0 ( float 5.123 * 0.1 = 0.5123 -> cast to INT = 0);
halpin = 51.23, scale 0.1, value to the node = 5.

Very thanks for these explanations ... I haven't read the drive code carefully. I looked at it I realized it could go I tried and it worked. The problems arose only with the "scaling" to be able to have the correct output positioning values ​​which, as mentioned, I still do not understand what relationship they have with the available revolution pulses and the reduction ratio. I sensed a certain linearity but not the formula which is evident has nothing to do with the calculations that are usually done ..... then I had not considered the cast and now I understand that everything works equally just because in any case the number of pulses per revolution on a 17bit ecoder are exaggerated having a mechanical reduction 1 (motor): 37.5 (output) .... so the decimals in fact are not noticed at all. However if I had a 1: 1 mechanical ratio and a very poor ecoder (512 puls / rev) I would probably have some positioning problems.

Now I have no time at all, but I will try to write the right drive for my drive so that I will understand better.

regards
bkt

@ Corbett

In regards to EL5101

I want to add spindle feedback on as have been working on 2 other setups...

I've had this working for a few weeks but haven't had the time to update you. It required a short hal component to calculate the rpm in rev/sec and rev/min from the count/frequency and then apply a +/- sign based on the spindle direction. I haven't done a ton of testing yet but it does seem to be working and satisfying linuxcnc's requirements of coordinated motion.

Dan

@corbet too

I have just connect my el5151 to slaves .... <slave idx="6" type="EL5151"/> .... ok terminal command "ethercat pdos -p6" & "ethercat slaves -p6" show me the correct info .... but when test in Axis gui "shox axis configuration HAL" the slaves nr. 6 is not seen ... all other is ok obviuslly.
Is there something i forgot to do? in the .hal file and in the .ini file I have not written anything, but in any case I should see the slaves ... which however is not there .... my encoders is npn at 24V and yet the leds do not light up in any case.

regards
bkt

@bkt

-does the slave 6 goes to op state, after starting linuxcnc? (ethercat slaves)
-"shox axis configuration HAL" is not clear to me... Do you mean "show hal configuration"? Then looking under Pins lcec - 0 - there should be an "6"
-check dmesg after starting linuxcnc, if there are errors with initialization