A-M-C A12A100 Servo Drive with SouthWestern Ind ProtoTrak Plus

More
15 Feb 2021 19:20 #198963 by PCW
FF1 is the main tuning value for velocity mode servos
but is typically not appropriate for torque mode servos

Can you set P and D to about 1/2 their current values
to get rid of the Oscillation and then try adjusting I again?
The following user(s) said Thank You: new2linux

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

More
15 Feb 2021 19:57 #198967 by new2linux
PCW, Many thanks for your help! Attached are pics of adjustment as per your suggestions, many thanks.
Attachments:

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

More
15 Feb 2021 20:02 #198968 by PCW
The next step is to dial in FF2
The following user(s) said Thank You: new2linux

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

More
16 Feb 2021 14:38 - 16 Feb 2021 17:23 #199066 by new2linux
PCW Thank you for your help! Attached is just 1 pic, this is best yet using the 50 unit scale. The feed rate is slow, but if you increase the f error will appear.

many thanks!

Edit: Additional pics with higher FF2. It seems to matter if I "save to file" the settings that are changed (even the few FF2 changes made recently), it will f-error if not careful.

many thanks!
Attachments:
Last edit: 16 Feb 2021 17:23 by new2linux.

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

More
17 Feb 2021 15:28 #199160 by new2linux
Many thanks for all the help!! Attached are several pic of the FF2 only changed. I have many others not posted here, if you want another setting I may have it. Starting with FF2 at .255 going to .290 . next .395 stop at .445.

many thanks
Attachments:

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

More
17 Feb 2021 15:56 #199161 by Todd Zuercher
.25 is too much FF2, You had about the best FF2 results around .2-.21.

I do think you may need to try to do something about the steady state error, In a velocity mode drive FF1 fixes that, but in torque or current mode, it is not going to behave the same way. I think some FF1 may help, but its results will probably be very speed dependent, in other words the setting that works well at one speed will likely be very wrong at another speed. I'm not sure what the answer is to that. I wonder if setting a value for "pid.N.maxcmdD" could help by limiting the amount of FF1 used at a certain point?
The following user(s) said Thank You: new2linux

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

More
17 Feb 2021 16:26 #199162 by new2linux
Todd & PCW many thanks for your help!! I will try FF2 at the .2-.21 range, monitoring the spike at the beginning & end of trace, as to reduce the value of the error in the trace. Adjust FF1 for the following error in the "cruse", the short time between the start & stop of trace as to reduce it as well. Keeping in mind that feed rate will suffer somewhere in the range, so try to sacrifice at the faster end of feed rate.

many thanks!

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

More
17 Feb 2021 16:54 #199163 by new2linux
Todd & PCW many thanks for your help! This attachment has the FF1 & FF2 adjusted, in addition to the gain is in the 20unit/div range, feed still 18.5. many thanks!
Attachments:

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

More
17 Feb 2021 19:46 - 17 Feb 2021 19:48 #199174 by PCW
Look like still not enough FF1
You should be able to just about zero the error during constant velocity.
If you have too much the error will be below zero

Note: zero for the following error is the straight white line.

Once FF1 is dialed in, I would add I until you get oscillations
and then use about 1/2 the value where oscillations started
Last edit: 17 Feb 2021 19:48 by PCW.
The following user(s) said Thank You: new2linux

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

More
18 Feb 2021 15:00 #199260 by new2linux
Many, many thanks! Attached are some current pics of the tuning. The FF1 shows before and after crossing the "0" or other side of 0.0, I think. Feed rate suffered 15"/min, the last pic has the "I" at 1/2 of the value that oscillation started. Will error out if feed rate is faster than 18.5"/min. Welcome your thoughts.
Attachments:

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

Moderators: piasdom
Time to create page: 0.170 seconds
Powered by Kunena Forum