PLC to turn on & off a device 10 times per second

[pg207]

I need to write a PLC program that sends pulses to a device 5000 times and not more than 10 cycles per second. can I simply use timers to accomplish this or is there a better way? I'm using Allen Bradley SLC 5/04 CPU. Thank!

 

[Helliana]

Timers should work fine. 100ms cycle honestly isn't that fast. I would try to pick a solid state type output card over a relay, but either one will work.

 

[dbh6]

Using a timer every 0.1 of a second you can trigger an output, then you can use that with a counter such that when the trigger counts up to 10 pulses your done. Just make sure your switching device if fast enough to actuate that quick as helliana mentioned a solid state type output should be quick enough.

 

[James Mcquade]

welcome to the forum.

is this a homework assignment?

you might want to look into the error associated with timers.
it will prove to be useful.

regards,
james

 

[geniusintraining]

I'm using Allen Bradley SLC 5/04 CPU.

I would make sure its a solid state output... I am sure a relay can handle it but I think it would shorten the life a lot

I had a temperature control valve that the solenoid was open and close... they did it with a relay output, once the PV was close the SP it would work a lot and the card would fail after about 6 months

It worked but not a very good design

 

[boneless]

Would a regular output, even solid state be quick enough to completely turn on and off, 10 times a second? You might want to check that. Yes I know that solid state is much quicker than that, but what about the effects of the surrounding electronics.

I think it might be better to look at specialized pulse output module, these are made to be cycled at higher frequencies.

 

[TConnolly]

A solid state output will switch plenty fast enough. If it is an AC output the triac will switch off within 1/2 AC cycle, or 8.3 milliseconds or less at 60hz, 10ms or less at 50hz. A DC transistor output will switch off much faster.

 

[shooter]

dont use a timer but use the internal time, as a timer needs time to reset, and a internal one goes on , until it overflows.

 

[OkiePC]

dont use a timer but use the internal time, as a timer needs time to reset, and a internal one goes on , until it overflows.

You can overcome this by, instead of resetting the timer, subtract the preset from the accumulator and unlatch the DN bit with OTU. It applies the over run to the next interval, and keeps the logic simpler than using free running clock, unless the frequency is an exponent of 2...then pick the bit in the free running clock and go with that.

For longer term counting:
If you need the number of pulses in a day to be more exact than the squareness of the pulses during any given minute, then the SLC real time clock is going to be off by up to several (as in as many as 30 in my experience) seconds per week, nothing you can do to program around that, except frequently correcting the real time clock, and then your pulses may be fubar for one cycle if you have calculated the on or off period from that clock. The free running clock could be used with a LIM instruction where you compare old and target numbers and trap for overflows, as you said, that is a lot more work than a SUB, OTU added to normal timer logic.

 

[Doug-P]

If a square wave output is acceptable I guess you could use a counter and pulse it at twenty hertz.

 

[Donnchadh]

Hi

I would loot at the STI in the slc. This is a time interrupt and I fell would work better for this

Donnchadh