× Forum Header

Mach3 XML and Stepconfig issues.

More
24 Apr 2017 14:02 #91964 by tecno
Todd, so am I as it has been working OK with Mach for many years.

I am more or less planning to replace my home-built BoB to a more modern one and go with PWM to 0-10V conversion if no remedy is possible for this.

In the pipeline there is a MESA setup, have all boards and new AC servos for X/Z and also a 3kW AC servo for spindle but I would like to see my basic parts to get working. Going MESA is no more Mach and I'm a bit uncertain about the lathe GUI in LCNC and in general what I can do in Mach3.

That is why I need to have both working in parallell to learn LCNC and get comfortable with it.

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 14:30 #91965 by andypugh
I don't think we can read too much into the halscope plot, as the video shows that it is sampling in the servo-thread at 1kHz rather than in the base thread. The apparent pattern in the output is probably actually "aliasing".

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 14:35 #91966 by tecno
Please tell me what to do and setup halscope properly.

Cheers
Bengt

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 14:42 #91968 by andypugh
It might be interesting to halscope stepgen.N.frequency
That would indicate if LinuxCNC thinks that it is doing the right thing.

To run halscope sampling in the base-thread, click the button that indicates the sampling rate.

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 14:56 - 24 Apr 2017 14:57 #91970 by andypugh
Reading the LM2907 data sheet, I am wondering if it is struggling with the duty-cycle of the stepgen output being nothing like 50%?

e2e.ti.com/support/amplifiers/etc_amplif...linear/f/18/t/375205 confirms that 50% duty-cycle is what is expected.

I _think_ that if you set the stepgen to quadrature mode then the A-phase will then have a 50% duty cycle. The drawback with this scheme is that the stepgen DIR pin is then no use at all for setting the spindle direction, and this output would need to be generated a different way.

So, try commenting-out the DIR connection to the parport, and edit the loadrt line
loadrt stepgen ctrl_type=p,p,v step_type=0,0,2
Last edit: 24 Apr 2017 14:57 by andypugh.

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 15:01 #91971 by tecno
I hope I´ve got it right.

Attachments:

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 16:05 - 24 Apr 2017 16:06 #91973 by tecno
Got email message with your answer but it is not showing up here.

Looking at the plots short pulse comes steady but step length is very short, so I wonder how it works OK in Mach?

Will try to apply Quadrature and report what happens.
Last edit: 24 Apr 2017 16:06 by tecno.

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 16:26 - 24 Apr 2017 16:27 #91975 by tecno
Halconfig shows flashing yellow ch A

Halscope nothing
Last edit: 24 Apr 2017 16:27 by tecno.

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 16:42 #91976 by tecno
Had to comment out several lines in hal to have lcnc starting

File Attachment:

File Name: Blomqvist_...24-2.hal
File Size:4 KB
Attachments:

Please Log in or Create an account to join the conversation.

More
24 Apr 2017 17:02 #91980 by Todd Zuercher
What does it look like with the spindle speed set to its fastest speed?
Try adjusting stepgen.2.steplen. Try setting it to 2500000 (that would make the step length half the frequency of your top speed.)
Attachments:

Please Log in or Create an account to join the conversation.

Time to create page: 0.142 seconds
Powered by Kunena Forum