![]() | |
| Home Page | Mark Forums Read | Today's Posts | My Replies | Classifieds | Reviews | Photo Gallery | Web Links | Share Files | Advertise With Us | Ad List |
| |||||||
| CNCzone Club House Discuss everything in between CNC. THIS IS NOT A TRASH BIN. |
| This forum is sponsored by: |
![]() |
| | LinkBack | Thread Tools | Search this Thread | Display Modes |
|
#2
| |||
| |||
| A system is in closed loop when a part of the output signal is used as a feedback to the input. This way, the fraction of the output signal act as a 'error signal' showing the deviation of the output against the expected output, and the feedback is used to correct that deviation. Is like if the control system knows what happen from its actions and can correct. A servo motor, for example, makes the feedback through position signal using the encoder, so if exist a deviation in position the system is able to correct. Other example could be a ligths system that uses a photocell to test ambient ligth to dim itself and do not waste power. In a open loop there is no feedback from the output to the input, so the system is blind, there is no way to know if there is deviation from the expected. The same above examples with open loop is a stepper that has no control over the position, only tries to move where indicated but really do not know where is at any moment. Or in the ligths system, a dimmer but used from other room, so you really do not know if the adjusts are correct. Hope it helps /F |
|
#3
| ||||
| ||||
Open loop systems in this case stepper motors have no feed back. So if you where to hold or jam one motor the controler would not error out but contiue from there on. Some times called loosing steps. In a closed loop system with encoder feed back (opticaly encoded disks on the motors) the controler sends so many steps or pulses to the motor and then waits and compares the feed back from the encoder. Then if the feed back (pulses from encoder used to represet rotational travel) is not what was expected then more pulses can be sent to the motor to compensate, or an error can be initiated. In this case a broken belt or ball nut is still not detectable. Lastly if the feed back is from a optical linear scale then, this feed back can be used to compensate for wear and thermal expansion as it is a more direct measurement of the movement and would be able to detect motor overload, belt breakage, and wear in the system. Lastly linear scales are ushaly made of more thermaly stabe material than the metal mechanical drive system and can be used with controler to compensate for thermal expansion.A bit long but, hoefully clear. |
|
#4
| ||||
| ||||
| Thanks guys, this is what I had thought. One further question. Since the feed back is to the driver cards only and not the computer does this make it a partial closed loop. If the driving software was also aware of the servos position, would that make a better system? Is that how profesional NC systems work? EDIT: So if I cleaned up this drawing and posted it, would that be accurate description? Last edited by ynneb; 03-17-2005 at 06:25 AM. |
|
#5
| |||
| |||
| yes, that is exactly how a professional system works. Mostly the drives are working on the +/- 10V platform and the encoder feedback is fed to the controller. and the presence of a feedback from the motor to the drive alone does establish a partial colsed loop( like in a gecko g320 ), so as long as you are generating pulses of correct shape and following a reasonably good timing and accelleration profile , it works just as well. |
| Sponsored Links |
|
#6
| |||
| |||
| Just about all digital motion controllers revert back to step and direction at there basic level. (How else could you digitally do it?) The fact that the loop is closed at the drives is no different than on a motion control board. (When it is done on the motion control card you can usually see following error at the control, with Rutex's new SPI interface this should also be possible.) Darek http://www.cnczone.com/forums/attach...achmentid=1712 Last edited by HillBilly; 03-17-2005 at 07:59 AM. Reason: add a link |
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | Search this Thread |
| Display Modes | |
| |
Similar Threads | ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Open Source CNC Controller Specification | gregmary | DIY-CNC Router Table Machines | 28 | 12-04-2005 10:58 PM |
| Servo Dynamics or Rutex | Tig | Servo Motors and Drives | 22 | 06-05-2005 11:17 AM |
| Open & closed loop | totsandman | Benchtop Machines | 3 | 05-07-2005 09:49 AM |
| Closed loop....open loop? | DAB_Design | General Electronics Discussion | 10 | 06-26-2004 05:02 PM |