Help in problem "unexpected switch limit"

More
06 Dec 2010 16:04 #5895 by aike
Hi!
I have problem - unexpected error appears "joint 0 on limit switch error", but switch not reached.
I apply the photo:
What is the problem?
None switch limit pressed, I implement flipflop function to find momentary pressed - unfortunatly.
But the error appears.
What event can occure this error, exclude the switch limit event?
Attachments:

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

More
06 Dec 2010 16:34 #5898 by turbospeedskater
I have had the same problem, but could fix it with some capacitors ( about 220nF) across the switches.
Depending on the cable length, the input resistance of your parallel port and the switching speed of you motor driver, you may have some spikes on the switch input lines caused by electromagnetic interference.

Additional you can use shielded cable between the switches and the parallel port.

best regards,
Martin.

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

More
06 Dec 2010 19:02 #5900 by aike
turbospeedskater wrote:

I have had the same problem, but could fix it with some capacitors ( about 220nF) across the switches.
Depending on the cable length, the input resistance of your parallel port and the switching speed of you motor driver, you may have some spikes on the switch input lines caused by electromagnetic interference.

Additional you can use shielded cable between the switches and the parallel port.

best regards,
Martin.

Hi!
I include this code in HAL file.

net X.min flipflop.0.set
net X.min-out flipflop.0.out pyvcp.led-xmin

In this case ANY short impulse will switch on the flipflop and the LED in PYVCP will flush. Hence I can found the problem.
But none led flushed, but errors appead. May be there is some invulable interval withing limit procedure, but this interval
make break down the system? What poll time to prevent false contacts switching?
I think that problem is in poll switches function. This function is very sensitive (but why the flipflop not switched?)
I use MESA 7I37 card. This is slow time signal card. And may be the front edge if impulse has short interval and cannot to switch on the led, but can to break out the system?

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

More
06 Dec 2010 19:22 #5901 by PCW
You could try the HAL debounce component in series with your limit switch input to see if the error is noise related

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

More
08 Dec 2010 06:57 #5923 by turbospeedskater
Just to test if my idea about spikes on the switch lines is right:

Can you remove the switches with cables from their place on the machine and put them (and mainly the cables) far away from the motor power lines?
Or (if that is easier) remove the switch cables form the input of the controller and make a short connection there (or leave open, depends on what is the status for limit switch not reached).

If everything is running correct then, it must have something to do with spikes on the switch inputs, I think.

Then the only question is: why the HAL-FF is not triggered. But to answer this, I must look closer into the source....

best regards,

Martin.

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

More
09 Dec 2010 11:11 #5935 by Mike_Eitel
I had the same expirence with a retrofit ( good friends fellmann )

After I did a correct grounding-concept ( PC, machine, powersupply, interfaces, etc connected via massive cables to one point ) all spikes disapeared.

I have to point out that the hal.osciloscop helped a lot, a phantastic instrument.

Mike

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

More
09 Dec 2010 12:41 #5937 by aike
I restricted the errors appear till 1-2 per one shift.
I grounded PC, impulse PC power supply, set capacitors in parralel each switcher.
I think that the main problem in 5V noise.
Probably the best way to power MESA doughter card 7I37 via separate power supply and not
via 50pin cable.

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

More
09 Dec 2010 16:05 #5941 by PCW
The 7I37 draws very little power and is well bypassed, its very unlikely that 5V power supply noise is a cause of your symptoms. The 7I37 does however have fairly fast (20 usec or so) and sensitive (~2V) inputs. This is why I suggested using the debounce component. Most likely your noise is impulse noise from PWM motor signals coupled into the opto inputs of the 7I37. cable routing and shielding can help but for low speed signals, the debounce component is a sure fix

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

More
09 Dec 2010 16:48 #5942 by aike
Hi!
I have never used debonce components. Can you help me or send reference how to use it?

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

More
09 Dec 2010 17:27 - 09 Dec 2010 17:49 #5945 by PCW
Heres a link that may help (with appropriate changes for your hardware):

" www.mail-archive.com/emc-users@lists.sou...ge.net/msg17909.html "
Last edit: 09 Dec 2010 17:49 by PCW. Reason: trying to get forum not to munge link

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

Moderators: piasdom
Time to create page: 0.132 seconds
Powered by Kunena Forum