Safety PLCs and Safety Distance Calculation

baref00t25

Member
Join Date
Feb 2006
Location
Central Wisconsin
Posts
177
Are there any documents or standards for calculating the response time for a controls system?

For example:
Light Curtain response 20ms
Safety Input module 30ms
Safety Program Scan 20ms
Safety Output module 40ms

Watchdog of Safety PLC 500ms
Discrepancy time of input 500ms

Are the 4 simply added together to assume the worse case, or would you have to consider the safety program scan as double because you may not know when the output image is being set in the current scan or the next. Do you need to factor in the watch dog of the Safety PLC or the discrepancy time of the input?
 
Are there any documents or standards for calculating the response time for a controls system?

For example:
Light Curtain response 20ms
Safety Input module 30ms
Safety Program Scan 20ms
Safety Output module 40ms

Watchdog of Safety PLC 500ms
Discrepancy time of input 500ms

Are the 4 simply added together to assume the worse case, or would you have to consider the safety program scan as double because you may not know when the output image is being set in the current scan or the next. Do you need to factor in the watch dog of the Safety PLC or the discrepancy time of the input?

Siemens has a spreadsheet that helps you calculate it for their components, including different network delays, things like that.

https://support.industry.siemens.com/cs/ww/en/view/93839056
 
Every manufacturer will provide worst case timing info, including how the IO and cycle time line up and fieldbus cycles if that's a factor.

We just went through an exerciser on this on a machine and the numbers you're listing are HUGE compared to the ones we were dealing with. You don't need discrepancy time because that is only used when pulling out the E-Stop/Closing the door, not when activating them. The PLC watchdog is also not part of the safety reaction, it's a separate thing. The OSHA arm movement speed is 1.6m/s, at 110ms, you're looking at 7 inches, but you need to check if you need backplane time as well.

For instance, the switching time of the input we used was worst case at 1.6ms, the output was 0.8ms, the safe PLC cycle time is 2ms, and the light curtain we used had an 8ms reaction time (it said it was 2ms with worst case of 8ms, so why even list 2ms!). There was also the requirement of 3 fieldbus cycles at 0.4ms each. Total time was 13.6ms for the signal. Since the guarded part was a laser, it's off time was also small, at 10ms. 23.6ms is 38mm Safe Distance. The guard was placed 10cm away and we called it a day.
 

Similar Topics

Do all of the so-called "Safety PLCs" have a software-generated watchdog or heartbeat, so that if the software locks-up, crashes, or enters an...
Replies
3
Views
921
Hi All, I am looking for a Safety PLC which can execute the I/Os and logic in around 25mSec. This is for about 1000 I/O points. Any suggestions...
Replies
9
Views
2,262
Hello, A new process line is being installed, it has its own safety PLC monitoring various E Stops / Guard switches etc... Within the line is a...
Replies
10
Views
2,931
I have been hearing a lot about "dedicated safety PLCs". I am curious as to why you would need a dedicated PLC to perform safety tasks, and if you...
Replies
10
Views
2,986
Hello, Does anyone have a good overview/camparison of Safety PLCs? Looking at an application with about 15-20 inputs and 5-10 outputs. There...
Replies
38
Views
21,798
Back
Top Bottom