Micrologix 1400 and / or

Hi all,

I have a series of 7 bits I would like to monitor. If any 'two' of these bits are on I need to activate a flag in my program. I'm trying to avoid having a group of nested instructions to do this. Is there any simple way to achieve this with bits.

Thanks

Use the BSR instruction. Bit Shift Right. Move the channel into another location. Shift the bits 7 times and count the number of bits that are on. That is if the last bit is one then increment a register. After 7 shifts you have your answer.

http://literature.rockwellautomation.com/idc/groups/literature/documents/rm/1766-rm001_-en-p.pdf

Regards,
Garry
http://www.accautomation.ca

A couple of ways that come to mind:

- If the bits are all sequential then you could use a rung with 7 NEQ instructions in series looking at the word as a whole - e.g. for B3:1/0 through B3:1/6, if it is NEQ to 0, 1, 2, 4, 8, 16, 32 or 64 then turn on an output. You would have to make sure that none of the other bits in the word ever get used - perhaps to guard against this, do a masked move first into an N word, to ensure that you are only examining those 7 bits

- On one rung, set an N word to zero. On the next 7 rungs, XIC each of your bits and if true, add 1 to your N word. If N>1 then you have two bits on. Again, if all the bits were sequential you could do this in two rungs instead of eight by also using a pointer and incrementing through (e.g.) B3:1/[N7:0] where N7:0 is your pointer. You could either increment the pointer every scan and it would take 7 scans to check all the bits, or you could use JMP and LBL to create a loop and do it all in one scan. Just make sure to stop incrementing the pointer on the last bit and give it a way to exit the loop!

Garry's method above is also a good one.

Your specification could be interpreted as 'at least two' (2 or greater) or as 'only two' (exactly 2).

The 'count up' method makes it easy to accomodate either interpretatiion because the final test can be 'equal to 2' or 'greater than or equal to 2'. Garry's method and ASF's second method each provide for a count.

I like the first sentence of ASF's second method. This will be understandable by virtually anyone just by looking at it. A nice rung comment of 'Count the ON Bits' on the first of the rungs would prevent any confusion.

bernie_carlton said:

Your specification could be interpreted as 'at least two' (2 or greater) or as 'only two' (exactly 2).

The 'count up' method makes it easy to accomodate either interpretatiion because the final test can be 'equal to 2' or 'greater than or equal to 2'. Garry's method and ASF's second method each provide for a count.

I liike the first sentence of ASF's second method. Yhis will be understandable by virtually anyone just by looking at it. A nice rung comment of 'Count the ON Bits' on the first of the rungs would prevent any confusion.

Click to expand...

It seems to me that if it is "at least 2 are on" that you could create an integer from the bits and if the result is >2 you know that at least 2 are on. I don't work with AB enough to know the method to do it without playing with it a bit but it shouldn't be too hard. However, I suspect that it is "exactly 2 are on"

If the bits are contiguous then the bits can be interpreted as a number (0 - 127).

If we are looking for exactly 2 bits on there are 21 numbers that fit this.

If we are looking for 2 or greater bits on then it is actually a little easier as there are only 8 numbers that DON'T fulfill that.

Jayden said:

I'm trying to avoid having a group of nested instructions to do this.

Click to expand...

Why?

Doug-P said:

Why?

Click to expand...

Probably because, like me, his OCD won't let him make a rung that doesn't all fit on the screen at once

ASF said:

Probably because, like me, his OCD won't let him make a rung that doesn't all fit on the screen at once

Click to expand...

Get a bigger screen.

28" moni lets you have quite a rung..

Nova5 said:

Get a bigger screen.

28" moni lets you have quite a rung..

Click to expand...

I'd love to see the looks you'd get carrying a 28" monitor onto a site to program a MicroLogix

Thanks for all the replies, (sorry for my delayed response)

I ended up needing a quick solution this this on the run so I just moved the Bit status to Integers and computed the SUM of the integers to get it working. I was curious to see if someone had a simple solution that I was maybe overlooking.

Although there are some good ideas I decided keep the integer solution because the bits are not sequential and I also the maintenance guys have trouble with anything other than very basic logic.

As for the reason I was trying to avoid nested, the program I'm working on has heaps of large nested rungs which I am slowly trying to re-write. I hate navigating and fault finding in programs with massive nested rungs.

ASF said:

I'd love to see the looks you'd get carrying a 28" monitor onto a site to program a MicroLogix

Click to expand...

What... you mean a monitor strapped to my back is unreasonable!?

ASF said:

I'd love to see the looks you'd get carrying a 28" monitor onto a site to program a MicroLogix

Click to expand...

Doesn't that new mega-laptop HP is working on got a 28" screen?

Construct the rung with multiple branches, test it works, and forget it... Who cares if it fits on the screen, so long as it works...

There are only 21 possible values the 7-bit register can have if only 2 bits are on - 3,5,6,9,10,12,17,18,20,24,33,34,36,40,48,65,66,68,72,80, and 96

If you want "2 or more bits on", then just test that the number isn't one of the 8 values 0,1,2,4,8,16,32, or 64

21 and 8 - just as Bernie posted earlier !!

daba said:

Who cares if it fits on the screen, so long as it works...

Click to expand...

My OCD cares, and hates you