DC configuration

register 0x6060 is usually a single byte/int8 register, yet in your xml, you're trying to write a double byte/int16 value -

If I know but I am not wrong, value from sdoDataRaw data="08" ... 8 is less then 127 and more then -128 ... or?

there is other one value which should be implemented in 0x6060 data object... and that is 0xFA from producer manual.. I should try them

Parsers can do seemingly strange things, so I'd only use "8".
"08" could be resulting in more 0 bits being prepended to the final output than an int8 register can handle.

finally we have success, they was more bugs then I expected... drive is in the OPmode after lcnc turned on... attachment 0

There is minimum required configuration for running it in the CSP mode... Attention for this really necessary facts

• read this thread forum.linuxcnc.org/ethercat/47358-kollmo...-stuff-over-ethercat
• for configuration have to be installed DriveGUI software by this thread forum.linuxcnc.org/49-basic-configuratio...llation-guide#221029
• all necessary stuff for rading is there www.kollmorgen.com/sites/default/files/p...ads/srethercat_e.pdf 
• there webhelp.kollmorgen.com/kas3.03/Content/C...0h%20Controlword.htm
• there www.wiki-kollmorgen.eu/wiki/DanMoBilder/...700help/ethercat.htm
• and there www.wiki-kollmorgen.eu/wiki/DanMoBilder/...0s700help/ASCII.HTML

Turn on drive and connect it via RS232 bus. You can continue after sucessfully running rs232 stuff. Drive will have 2 warnings n18-n27. Which means that not configured DC clock in right way. never mind, it will disappear after beginning of ethercat communication. (There is no place for error. Erros have to be solved separatelly.)

User have to read or customize settings in the softwate though terminal interface in the DriveGUI software with next commands - PTBASE(in the 250us) and PRBASE(bit resolution per revolution) 

For working configuration write in the terminal PTBASE 4 (means 4*250us=1ms == 1 000 000ns of servothread period time) and PRBASE 16(means 2^16=65536 increments per revolution).  attachment 1

save all to eeprom and cold start the drive.. after this is driver ready to ethercat run.. 

After all drive will switch to wait for enable signal. 

Minimal necessary .xml config in the attachment 2

I am working on customizing CiA402 component from mr. dominik braun for incremental output and custom OPmode numbers because they are different for older stuff... I will publicate it under DS402 name

enjoy all stuff and work yourself

I have Beckhoff ax2006 driver, it's a rebranded Kollmorgen S600. I had a lot of trouble with this driver. Never go's in op mode over ether cat with DC. I will try your solution next week. What is min DC cycle time of your drive? I hadn't been succeeding with lower that 4ms. It's too slow for pid loop. I hope it's will work. Thanks.

 

Hello,

I have not good news for you. there are some hidden bugs in your configuration.

1. 3.6 Free PDO Mapping (S300/S700 only) In addition to the fixed PDO mapping the free mapping of real-time objects is possible. This mapping is only available when the parameter DRVCNFG3 bit 16 (0x10000) is set to 1. The configuration is similar to the described sequence for the fixed mappings: 1. The mapping selection is cleared (write 0 to object 0x1C12 sub 0 and 1C13 sub 0) 2. The free mapping is built with the objects 0x1600 sub 1 .. n (receive-direction) and/or 0x1a00 sub 1 .. m (transmit direction) in the CANopen manner. 3. SDO write access to the objects 0x1C12 sub 0 and/or 0x1C13 sub 0 with the number 1. See an example on page 24. The cyclically used data are visible in the PDO assignment window for the Inputs and Outputs of the Sync Managers. Default setting are the fixed PDOs 0x1701 and 0x1B01 (visible contents when selected in the PDO list). 
   You can read it yourself in this document  " www.kollmorgen.com/sites/default/files/p...ads/srethercat_e.pdf " at page 23 from beginning.
   
2. you can run AX2006 in more ways over lcnc(step/dir or analog input in velocity mode).... but in the CSP over ethercat you probably can not because you have to use fixed moapped PDOs which are not in shape which we are using for CSP. You have to use demand position or demand velocity PDO which is not same as setpoint position. it is in the shape of older stuff twincat2 interface.
   
3. I bought mesa 7i77(analog interface) stuff for this purpose because I recently have some S300 amplifiers with older hardware and software revisions which do not support Flexible mapped PDOs. They are great stuff but from times age which worked on other principe.

If you have time take a picture of your hardware and we will be clevered after. you should take a shot of ethercat interface and manufaturer sticker from the side.

My configuration runs in 1ms synchronization DC. You should run it over ethercat in the " n x 250us" samples like I´ve described before.

I think 4ms is the limit of realtime. All of your stuff(pids, motion controller, etc...) have to run in 4ms. It should be too slow from my point of view for any type of control.

I think but we can ask tommylight or bigjohn or other senior users on their experiences.

regards Slav