Hi Russ,
What Software Version are you using?
What Limit Action Type are you selecting?
You left the bit numbers blank. I assume you are entering 42 for both directions?
When you set parameters on the KMotion Screens you must Download them to KFLOP for them to have an effect. Are you doing this? You might watch this Flash Video: http://dynomotion.com/Help/FlashHelp/Parameters/index.html
How are you testing if Limits are active?
Regards TK
| Group: DynoMotion |
Message: 9603 |
From: Russ Larson |
Date: 6/2/2014 |
| Subject: Re: KFLOP question [1 Attachment] |
TK, Software is version 4.32 The action selected was kill motor drive. Yes, the blanks were both set to bit 42. We also tried bit 43 with the same results. These are both set to inputs. How did we check things. We issued a move command in the step response screen and if the limit is enabled, it will not move, and reports on the console screen the axis is disabled, even though the state in the digital IO screen shows it is not set high. Yes, we have always downloaded the parameters on the Kmotion screen to the kflop. If you change the limit bit to one on a different connector like Bit 35, when set to output, the move commands work. If you change the state of bit 35 during a move the axis stop immediately and the console screen reports the axis is disabled. The issue seems to be limited to the bits on JP5. We are kind of scratching our head here.... any ideas what we might be doing wrong? Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Monday, June 02, 2014 11:58 AM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] KFLOP question [1 Attachment] Hi Russ, What Software Version are you using? What Limit Action Type are you selecting? You left the bit numbers blank. I assume you are entering 42 for both directions? When you set parameters on the KMotion Screens you must Download them to KFLOP for them to have an effect. Are you doing this? You might watch this Flash Video: How are you testing if Limits are active? We have the KFLOP partially working now. The motors move and we have copied the configuration info into the provided version of InitStepDir3Axis.c We did this for thread 2, compiled the code and ran the code and that seems to work. We actually added a small chuck of code at the bottom to activate a transistor driven relay that enabled the Gecko G540. We know this part of the initialization code runs because when we run it after we compile and download it, we can hear the motors energize. The next issue was to get the limit switches to work. They are all wired together in series and are normally closed contract connected to ground. They all attach to JP5 pin7, which is IO42. We have used 5.6K pull up resistors for all the IO on JP5. The issue is when we use JP5 and test the Limit switch options listed below for bit42, nothing happens. We can see the bit change on the digital IO page by pressing any of the limit switches from low to high, but not from the configuration flash screen with the options listed below. We then moved to another IO pin on another connector and it works correctly. Are we doing something wrong using the JP5 connector as standard IO? Stop When Low Stop When Low
Posted by: Tom Kerekes <tk@...>
|
|
| Group: DynoMotion |
Message: 9617 |
From: Tom Kerekes |
Date: 6/4/2014 |
| Subject: Re: KFLOP question |
Hi Russ, Sorry for the delayed response. One possibility is noise. KFLOP JP5 pins are very high impedance (many megohms). 16 of the IO on JP4 and JP6 have pull down resistors (although IO 35 does not) so those will not be floating like JP5 inputs. How are things wired? You can not expect floating inputs to remain always low. By looking at the Digital IO screen you are not likely to see noise glitches because it is only updated several times per second. But KFLOP checks limits every 90us. HTH Regards TK
| Group: DynoMotion |
Message: 9618 |
From: Russ Larson |
Date: 6/4/2014 |
| Subject: Re: KFLOP question [1 Attachment] |
TK, We had pretty much come to the same conclusion. I will setup a scope to see if we can see noise trigger these inputs. They are pulled high with a 5.6K resistor and normally held to ground until a limit switch opens which pulls the input high with the 5.6K pull up. We also decided this is happening so fast that the digital IO screen cannot response fast enough to be seen. The interesting thing is these inputs are pulled directly to ground, so we are surprised a noise spike could activate the input. I would understand if the input was floating but this is pull in each direction which would have seemed pretty safe even on high impedance inputs. We are looking at reversing the logic and see if we can avoid the noise spikes or possibly put filter caps on the inputs. Thanks Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Wednesday, June 04, 2014 2:45 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] KFLOP question [1 Attachment] Hi Russ,
Sorry for the delayed response. One possibility is noise. KFLOP JP5 pins are very high impedance (many megohms). 16 of the IO on JP4 and JP6 have pull down resistors (although IO 35 does not) so those will not be floating like JP5 inputs. How are things wired? You can not expect floating inputs to remain always low. By looking at the Digital IO screen you are not likely to see noise glitches because it is only updated several times per second. But KFLOP checks limits every 90us.
HTH
Regards
TK The action selected was kill motor drive. Yes, the blanks were both set to bit 42. We also tried bit 43 with the same results. These are both set to inputs. We issued a move command in the step response screen and if the limit is enabled, it will not move, and reports on the console screen the axis is disabled, even though the state in the digital IO screen shows it is not set high. Yes, we have always downloaded the parameters on the Kmotion screen to the kflop. If you change the limit bit to one on a different connector like Bit 35, when set to output, the move commands work. If you change the state of bit 35 during a move the axis stop immediately and the console screen reports the axis is disabled. The issue seems to be limited to the bits on JP5. We are kind of scratching our head here.... any ideas what we might be doing wrong? What Software Version are you using? What Limit Action Type are you selecting? You left the bit numbers blank. I assume you are entering 42 for both directions? When you set parameters on the KMotion Screens you must Download them to KFLOP for them to have an effect. Are you doing this? You might watch this Flash Video: How are you testing if Limits are active? We have the KFLOP partially working now. The motors move and we have copied the configuration info into the provided version of InitStepDir3Axis.c We did this for thread 2, compiled the code and ran the code and that seems to work. We actually added a small chuck of code at the bottom to activate a transistor driven relay that enabled the Gecko G540. We know this part of the initialization code runs because when we run it after we compile and download it, we can hear the motors energize. The next issue was to get the limit switches to work. They are all wired together in series and are normally closed contract connected to ground. They all attach to JP5 pin7, which is IO42. We have used 5.6K pull up resistors for all the IO on JP5. The issue is when we use JP5 and test the Limit switch options listed below for bit42, nothing happens. We can see the bit change on the digital IO page by pressing any of the limit switches from low to high, but not from the configuration flash screen with the options listed below. We then moved to another IO pin on another connector and it works correctly. Are we doing something wrong using the JP5 connector as standard IO? Stop When Low Stop When Low
Posted by: Tom Kerekes <tk@...>
|
|
| Group: DynoMotion |
Message: 9619 |
From: Tom Kerekes |
Date: 6/4/2014 |
| Subject: Re: KFLOP question |
Hi Russ,
Grounded far away at the switch may not stop sub microsecond noise spikes due to cable inductance. I like the filter cap idea. Try 0.1uF Ceramic caps to KFLOP GND near KFLOP.
Good luck TK
| Group: DynoMotion |
Message: 9651 |
From: Russ Larson |
Date: 6/10/2014 |
| Subject: Re: KFLOP question [1 Attachment] |
Tom, Success. After adding high quality .1uF caps on the J5 inputs to ground we had success. We had to keep the leads no longer than 4 inches, but we are getting repeatable results. Thanks
Russ
Sent from my iPad
Hi Russ,
Grounded far away at the switch may not stop sub microsecond noise spikes due to cable inductance. I like the filter cap idea. Try 0.1uF Ceramic caps to KFLOP GND near KFLOP.
Good luck TK
| | | | | | | |
