CNC Skipping G Code Instructions

More
04 Aug 2020 23:22 #177305 by tommylight

I'd believe it's something wrong that I have done or something I have missed. I'm new to this so I could have done something dumb.

Nobody said any of those things, and we are aware that you are following the guides for setting the timing, but the issue is with the drive manufacturers and their specified timings, namely those given timings work only on very expensive drives. Everything else needs close to twice the timings to work.
And acceleration settings are to low for metric, but to fast for imperial units.
The following user(s) said Thank You: M Sills

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

More
05 Aug 2020 00:24 #177317 by M Sills
So I tested the looser timings but I still have the oscillation in rpm that seems to be independent of timings.
I have of 10,000, 10,000, 5000 and 5000.

I'll change my accelerations since I am using Imperial. A setting of 15 should be good right?

I have any faults trigger the motion.feed-inhibit hal command. If there is a better way, let me know.

From here, I'll send a response if this issue happens again.

Now I just need to fix what I assume is a PID tuning issue. When running at 20 in/min, it has hiccups when performing linear movements. When running at greater than 50, it overcompensates way too much when I stop moment causing oscillations.
This was not an issue before I started messing with my settings. I'm not sure what changed. My P value is still at 1000 for my x and z.

Thanks for the help so far.

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

More
05 Aug 2020 00:41 - 05 Aug 2020 00:47 #177322 by PCW

For the wiring, Both my x and z are on the TB2 header.

X Axis

PUL+ Pin 3
PUL- Pin 2
DIR+ Pin 5
DIR- Pin 4

Z axis

PUL+ Pin 9
PUL- Pin 8
DIR+ Pin 11
DIR- Pin 10


You might try single ended connections
Some drives require quite high current on their inputs

Single ended connections:

X Axis

PUL+ Pin 6
PUL- Pin 2
DIR+ Pin 6
DIR- Pin 4


I would leave the PID settings alone (at 1000 for a 1 ms servo thread)

Note you can also reduce the ferror settings in the ini file to see if the
motion problem is in the hal/stepgen side or the drive hardware or interface
side. I would suggest 0.001" as min_ferror and 0.005" as ferror
Last edit: 05 Aug 2020 00:47 by PCW.
The following user(s) said Thank You: M Sills

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

More
05 Aug 2020 01:06 - 05 Aug 2020 01:10 #177326 by M Sills
Turn out my oscillation issue was that I had accidentally changed my stepgen_maxaccel when viewing the calibration page...

Thanks for the help. I will update if my machine misses a rapid again.
Last edit: 05 Aug 2020 01:10 by M Sills. Reason: clarification

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

More
05 Aug 2020 01:46 - 05 Aug 2020 01:48 #177331 by M Sills
I found the issue.

When my z would start a rapid, there was a very small chance my power supply would cut out for about a 1/4 second.

This was enough for my machine to miss the rapid but for me not to catch it.

I have a switching power supply for each motor and a separate supply for my mesa card so I didn't realize this originally.

I didn't step down the max amperage in my driver config so it would pull roughly 140% of the rated current.
Last edit: 05 Aug 2020 01:48 by M Sills. Reason: clarification
The following user(s) said Thank You: tommylight

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

More
05 Aug 2020 09:25 #177370 by tommylight

I found the issue.

When my z would start a rapid, there was a very small chance my power supply would cut out for about a 1/4 second.

This was enough for my machine to miss the rapid but for me not to catch it.

I have a switching power supply for each motor and a separate supply for my mesa card so I didn't realize this originally.

I didn't step down the max amperage in my driver config so it would pull roughly 140% of the rated current.

That will do that ! :)
Thanks for reporting back.

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

Time to create page: 0.092 seconds
Powered by Kunena Forum