AB micro logix scan time

Hey all,
a little info on the project:
I have a VFD driven "buffer" roller in a printing press that is removed and replaced on a daily basis, the operator simply removes the old "buffer", puts the new one in and secures the lovejoy coupling that ties it to the motor. While the press is running this buffer turns the opposite way the web is travelling. apparently the operator forgot to attach the lovejoy couple to the motor and a lot of scrap material was ran. Now the plant manager wants me to come up with a way to tell if the buffer is running the right direction. I have to be able to tell that the buffer is physically turning so whatever I do has to be tied to the buffer.See .bmp for details
My solution is to attach two pieces of reflective tape at A and B on the buffer shaft and have the two photoeyes A and B look for the tape in a sequence to tell if the shaft is turning in the right direction.
The code would be such
Sensor A Sensor B B3:0/0
---] [-------------]/[--------------( L )
The shaft is 2 inches and will have a max rpm of 1600
My question involves scan time,would the AB micrologix 1200 be able to execute the this code quickly enough to know if the shaft is turning the right direction.

It depends on response time of the photo eye... ML1200 ladder execution depends on length and type of instructions contained. I would think an encoder to a HSC would be better.

I don't think it's fast enough for real-time, but you could read it on a one-shot during start and compare the time stamp.

Charlie

You will get one revolution every 37.5msec. To get the best chance of catching the input you should run the tape around half the shaft circumference, giving you a 50% duty cycle. This gives you 'tape time' of 18.75msec. The two pieces of tape are separated by 1/4 revolution so you have 9.375msec between input on events.

Make sure your sensor can respond to the reflector in the time you have. Some photoeyes are pretty slow.

If you have a DC version of the 1200 you should be able to use the HSC function and two of the fast inputs. If you set it up right two of the HSC function file status bits are CU(count up) and CD (count down). Evaluate these to determine direction.

If you do choose to do this in logic, you don't want to look at two levels to determine direction. Evaluate one level on a oneshot of the other. Level sensitive logic will be true once a cycle no matter which direction you are going.

Keith

You'll definately have to one-shot the leading edge of 'A' since your current logic (A and NOT B) will occur regardless of your direction of rotation.

The ML1200 has the ability to configure an input as a latched input. That means that a very short pulse that is less then one scan long is latched by the system for one program scan, that way you see the pulse, even if the pulse is faster than the scan.
Check the manual for more information on using this feature.

duhhhh

Thanks all for the input
A little more info for you guys, The buffer roller is always against the web and doesn't come on until the press speed is greater than 500fpm. So while the press is running below 500fpm the buffer roll is turning the direction the web is going. This is caused by the web turning the buffer. At greater than 500 feet the drive starts up and turns the buffer the opposite direction of the web, Hence the need to be able to distinquish between the two directions. Anyway, what I will do is when I get the signal to turn the buffer on, first go through a "direction detect sequence" that runs the buffer in an very slow mode before ramping up to full speed or may put in a long accel time, this will give the plc an adequate chance to test the direction of the buffer. I realize there could be complications to this but its in the details.