the choices of NIC for 6 DOF robot with linuxcnc

Hello everyone:
Recently,my teachers and I intended to control a 6 DOF robot for stone carving with linuxcnc and the real robot can move simplely,etc writing the word of LinuxCNC. But it is not very stable in all running times. I have some questions about the system and linuxcnc:
1.Now,we just chose a PC as the controller,the system is Debian 7.9 and kernel system is 3.4.9-rtai-686-pse, linuxcnc is axes-joint13 for robot and the kinematics module is genserkins.
when I run the ec_master and it begin to scan the bus,there are always some questions, just as following(the linuxcnc is not running at this time):

[ 25.299137] EtherCAT ERROR 0-5: Failed to read mapped PDO entry.
[ 25.299138] EtherCAT ERROR 0-5: Failed to read mapped PDO entries for PDO 0x1600.
[ 25.363120] EtherCAT ERROR 0-5: Reception of CoE upload response failed: No response.
[ 25.363122] EtherCAT ERROR 0-5: Failed to read mapped PDO entry.
[ 25.363124] EtherCAT ERROR 0-5: Failed to read mapped PDO entries for PDO 0x1A00.
[ 25.363126] EtherCAT 0: Bus scanning completed in 9168 ms.
[ 25.363128] EtherCAT 0: Using slave 0 as DC reference clock.
[ 28.402384] EtherCAT WARNING 0: 36 datagrams TIMED OUT!
[ 28.402385] EtherCAT WARNING 0: 36 datagrams UNMATCHED!
[ 28.434447] EtherCAT ERROR 0-0: Reception of CoE SDO description response failed: No response.
[ 28.494439] EtherCAT ERROR 0-1: Reception of CoE SDO description response failed: No response.
[ 28.546589] EtherCAT ERROR 0-2: Reception of CoE SDO entry request failed: No response.
[ 28.590422] EtherCAT ERROR 0-3: Reception of CoE dictionary response failed: No response.
[ 28.642573] EtherCAT ERROR 0-4: Reception of CoE SDO entry request failed: No response.
[ 28.734382] EtherCAT ERROR 0-5: Reception of CoE SDO description response failed: No response.
[ 427.372489] EtherCAT: Requesting master 0...
[ 427.372493] EtherCAT: Successfully requested master 0.
[ 427.374225] EtherCAT 0: Domain0: Logical address 0x00000000, 162 byte, expected working counter 18.
[ 427.374228] EtherCAT 0: Datagram domain0-0-main: Logical offset 0x00000000, 162 byte, type LRW.
[ 427.374240] EtherCAT 0: Master thread exited.
[ 427.374243] EtherCAT 0: Starting EtherCAT-OP thread.
[ 427.374267] EtherCAT WARNING 0: 24 datagrams TIMED OUT!
[ 427.374269] EtherCAT WARNING 0: 25 datagrams UNMATCHED!

2.When I exit the linuxcnc,there are some PLL errors occuring,like bellow:

[ 1518.040795] EtherCAT 0: Releasing master...
[ 1518.040797] EtherCAT 0: Released.
[ 1518.040853] EtherCAT WARNING 0-1: Changed to SAFEOP + ERROR in the meantime.
[ 1518.043620] EtherCAT ERROR 0-2: AL status message 0x0032: "PLL error".
[ 1518.043735] EtherCAT 0-2: Acknowledged state SAFEOP.
[ 1518.055581] EtherCAT ERROR 0-3: AL status message 0x0032: "PLL error".
[ 1518.055752] EtherCAT 0-3: Acknowledged state SAFEOP.
[ 1518.059595] EtherCAT ERROR 0-4: AL status message 0x0032: "PLL error".
[ 1518.067537] EtherCAT 0-4: Acknowledged state SAFEOP.
[ 1518.071643] EtherCAT ERROR 0-5: AL status message 0x0032: "PLL error".
[ 1518.071862] EtherCAT 0-5: Acknowledged state SAFEOP.
[ 1518.083578] EtherCAT 0: Slave states on main device: PREOP, SAFEOP + ERROR.
[ 1518.087628] EtherCAT ERROR 0-0: AL status message 0x0032: "PLL error".
[ 1518.087998] EtherCAT 0-0: Acknowledged state SAFEOP.
[ 1518.111605] EtherCAT 0: Slave states on main device: PREOP.

3.I have the real robot moving in a simple way,just as writing a word of linuxcnc,but I found that when I run linuxcnc in simulation mode with the same kinematics module，genserkins, and the same parmameter of ini file. DH parameters.the tool displayed in the AXIS interface is not continuous and the trajectory does not meet the trajectory descripted by the Gcode.
4.The NIC I am using is r8169,is there any relationships with the above questions? Can anyone tell me what type of NIC is best choices for our system?

thank you very much!!!!!!!

I am really new with linuxcnc. It is a great pleasure if someone know the answers,please tell me.Thank you

chuchengbing wrote:

Hello everyone:
Recently,my teachers and I intended to control a 6 DOF robot for stone carving with linuxcnc and the real robot can move simplely,etc writing the word of LinuxCNC. But it is not very stable in all running times. I have some questions about the system and linuxcnc:
1.Now,we just chose a PC as the controller,the system is Debian 7.9 and kernel system is 3.4.9-rtai-686-pse, linuxcnc is axes-joint13 for robot and the kinematics module is genserkins.
when I run the ec_master and it begin to scan the bus,there are always some questions, just as following(the linuxcnc is not running at this time):

[ 25.299137] EtherCAT ERROR 0-5: Failed to read mapped PDO entry.
[ 25.299138] EtherCAT ERROR 0-5: Failed to read mapped PDO entries for PDO 0x1600.
[ 25.363120] EtherCAT ERROR 0-5: Reception of CoE upload response failed: No response.
[ 25.363122] EtherCAT ERROR 0-5: Failed to read mapped PDO entry.
[ 25.363124] EtherCAT ERROR 0-5: Failed to read mapped PDO entries for PDO 0x1A00.

Hard to guess what is wrong without looking at your ethercat configuration xml.

My guess:
CoE slaves have to be configured.
This can either happen

1. with RAW SDOs commands as in github.com/sittner/linuxcnc-ethercat/blo...cat-conf.xml#L23-L33
2. or with a shell script as in github.com/sirop/EtherCAT_HAL/blob/master/setPDO.sh before starting your component

In rare cases "Slave Information Interface" (SII) EEPROM can be invalid: www.etherlab.org/en/ethercat/faq.php

Try to search for your errors on the etherrlab lists:
www.google.de/search?q=Failed+to+read+ma...e:lists.etherlab.org

4.The NIC I am using is r8169,is there any relationships with the above questions? Can anyone tell me what type of NIC is best choices for our system?

Yes, this simple NIC is actually a much better choice than e1000e.

Thank andypugh and sirop:
sirop,you mean the r8169 is a suitable choice for our system? Is there any more better one?
Thank you

chuchengbing wrote:

Thank andypugh and sirop:
sirop,you mean the r8169 is a suitable choice for our system? Is there any more better one?
Thank you

I would say that r8169 is a better choice than e1000e, but maybe RTL8139 is even better?

The Intel chipset is designed for high throughput with as little CPU
interaction as possible. It has things like interrupt throttling,
buffers and the works - poison for any real time application.

Although we have not experienced these issues and are successfully using
this driver, we try to use the most dumb (!) network cards available:
RTL8139 chipset is a good example. Paradoxically, they work best: when
they receive a packet of data, it is available immediately; when a
packet is sent to the card, it is transmitted immediately - no
buffering, no throttling.

lists.etherlab.org/pipermail/etherlab-users/2016/002961.html

This is my EtherCAT configuration file.

chuchengbing wrote:

File Attachment:

File Name: 6_STEP.xml
File Size:7 KB

This is my EtherCAT configuration file.

So what are the SDOs you configure your CoE slaves with?

Hello,
please start with a simple configuration and only with one slave.
If this is running, you can take all slaves and more registers.
Maybe try a config like this:



And try first without distributed clocks, disabe line <dcConf assignActivate="300" sync0Cycle="*1" sync0Shift="0"/>

Please tell us a litte bit about your hardware, manufacturer, type ....

If you start the PC, what say "ehercat slaves" ?

regards Frank

Thank sirop very much!!
All the things I need can be got through the pdos. Is it necessary to configure the sdos? What is the reasons for the PLL error? Does it have the relationship with the sdo just as you said?