Range of action of (^) coil and %T

I've seen that a (^) coil sets a bit var for one sweep. Does the sweep end at the end or loop all the way back to the place where the coil is?
(GE Fanuc PLC micro)
%T can be reused, i.e. is temporary. Will the value be retained over multiple sweeps (assuming no power failure)? When should I be using this and not %M?

A pulse bit generated by an instruction, remains active until the instruction is scanned next time.
However, a physical output with no immediate refresh is being refreshed only after the scan end.

When the logic controlling the one shot coil evaluates as true, the memory location assigned to the one shot coil will be true until the next time the rung containing the one shot coil is evaluated. The true state "wraps around" the end of the scan and through the rungs before the rung with the one shot coil.

The only significant difference between the %T memory and %M is that %T memory cannot be retentive, while each %M location can be declared retentive (default is non-retentive).

So if the pulse bit variable is in a subroutine that is only called every 10 sweeps, will the var keep its value for 10 sweeps?

I would guess yes. Logic that is not executed cannot do anything. So if you trigger the one-shot on Scan #1, it will stay high until Scan #11, when its rung of logic is executed the next time.

I am guessing. It would be simple to test. Put a rung of logic in that uses your 1-shot as a normally open contact triggering an ADD instruction. Use a temporary INT that adds to itself. Monitor the temporary INT to see what happens.

Good idea. Thanks.
I still am wondering when I am to prefer %T. Is it to save memory usage?

Oh, forgot to answer that. The %T memory is available for things that don't matter. For example, a momentary button on the HMI. %T memory is not retentive through a power cycle, it always comes up zero. So its a great fit for a momentary button on the HMI. In fact, that's all I use %T memory for. It makes it easy for me because when I see %T I always know it is coming from HMI.

If you run out of $M memory you can use %T. But you can also use %Q and %G... so for me, %T is only for HMI momentary buttons.

%T memory cannot be uploaded to your laptop. For example if you want to upload all data values from PLC to your laptop, you won't get %T data. All others are saved: %M, %Q, %I, %G, %R.

I hope this helps clarify it.

great. very helpful.
Thanks to everyone.

You are welcome.

I thought of another reason we use %T: as rung markers. For a section of logic buried deep within a subroutine, put a rung just in front with a normally-open %T bit turning on the same %T bit coil. Then all you have to remember is the address of the %T bit - search for 101T and bam - go right to the logic you were looking for. In the days of LM90 this was a great help.