Advanced Search

– ETHERCAT RC SERVO MOTOR INTERFACE
– 4x RC SERVO MOTOR CAN BE CONNECTED SAME TIME
– 24VDC SUPPLY
www.egerobotics.com/ethercat-pwm-rc-servo-interface

That sdodataraw input field to sdoconfig isn't the easiest to use, unfortunately.
After using an online ieee float to hex converter one ends up with 0x43480000 for 200.0
Swap the byte order, the number becomes SdoDataRaw="00 00 48 43".

An alternative is to set the value outside of linuxcnc
ethercat -p 0 download 0x2001 0 200.0
and check the value with 
ethercat -p 0 upload 0x2001 0
The value needs to be set at every power cycle of the board, it isn't saved.

I gave MacAcceleration a default value of 1000000.0, you might want to give it a reasonable value.
 

Hi,
Any update on the project?

 
Thank you for your reply!

I don't think the GUI and/or the connection method matters much in this situation - it's more of a HAL question than anything else - but if you're curious, I'm retrofitting a Carvera Air desktop mill, details can be found here , it's a bog-standard 3-axis Cartesian machine for all intents and purposes. Probing is the last part before the setup can be called functional.

I don't know what kind of picture would be helpful here, so I'll try to explain better.

The machine has a tool setter/TLO sensor on the bed, it's a spring-loaded button that closes the contact when pressed.
There's also a probe that you clamp in the spindle to touch off the top of the stock, again, it closes the contact when the tip of it is pressed.

I need to measure the length of the probe using the tool setter. If I move the probe down against the button and trigger the measurement when just the tool setter button closes the contact, I get an inaccurate reading - the probe tip requires additional ~0.05mm of travel before it triggers (I'm wondering - is it maybe common to ignore this error? That would explain a lot). So the length needs to be measured when both the probe and the tool setter contacts close. I can combine the HAL inputs with boolean AND operation, feed the result into motion.probe-input, and it works perfectly.

However, after doing that I need to touch off the stock with the probe, and that quite obviously doesn't work with ANDed inputs - the tool setter button is not pressed when the probe triggers against the stock (unless I reach out and press it down with my hand but that's obviously not a solution). The same goes for measuring the end mill length.

I can replace AND with OR, and then I can measure my end mills and touch off the stock, but I lose the ability to accurately determine the probe length, as described above.

I don't mind writing a new HAL component if needed, but I don't know how to isolate the situation of measuring the probe. There's motion.motion-type pin, as I've discovered recently, so maybe I can cook up something based on that, in combination with the fixed tool number for the probe, but I'm not sure if motion.motion-type would be set to 4 (tool change) when tool length measurement is happening, I guess I'll have to experiment.

One should always RFTM, it's a good practice to follow unless you are quite financially well off, have plenty of time to spare and have the patience of a saint.

Or some other poor sod is setting everything up and you have 100% trust in them.

Do you have a link to the drive manual or at least the the I/O connections?

You can use MesaCT to generate 7I97/7I97T compatible configurations
plus theres a sample 7i97/7i97t configuration. I think this is in the 7I97T
software support zip file, if not, I can post it here.

Yes, that should be correct.

Thx, this is how my setup looks like... 



So if I understood your answer correctly - I replace R1 (currently 10k) with 110k, which gives me 10k (110k & 11k in parallel). 
Will try that tomorrow....

thx!!