fast track/CTI 2500

To all, please help. I am a little confused about the difference between special functions[programs or subroutines]. I have some logic already prepared, and if someone happened to know a little more about this than I, I would sure like to get some help. For example, if I need to do any floating point math, does this all have to be in a special function(program or sub)?

Thanks

Here is the pdf of what I am looking to do.

Thanks again.

Not sure I understand, but in the CTI2500, scaling would be handled in a SFPGM.

SFPGM's can also do high order math. The file you posted is for a A-B plc. What exactly are your goals and issues?

The maths Instructions that can be used in ladder are for Integer only, so if you want to do floating point then you have to do it if special functions.

I know what logic I need (the AB pdf), I am just uncertain about how to translate it into the special functions.
1. I have 12 flows that I need to scale into the PLC.(What is the best way to scale analog inputs?[Ken you mentioned using a SFPGM) Is it through the analog alarm edit screen? Is there a way to monitor over-under current on analog channels?
2. I need to be able to take these flow rates into floating point registers and perform the appropriate math to come up with basically a 0-100% signal which is going to 12 different grudfoss metering pumps. I assume that V-memory is allocated in 16-bit registers. If I need a floating point, do I need to plan on using two registers (V_x & V_x+1)?
3. What is the difference between using a speical function subroutine or program? Does anyone happen to have an example? Basically, if I am thinking about this, is it possible to take rungs 4-11 (from the pdf) and put this in a special function sub or program? I get a little nervous about what I can and cannot do when I read about this T-memory and how it is not saved.

Your help is very much appreciated.

Thanks

1: Have a look at the SCALE and UNSCALE instructions (in SFP) typically they are used to 'scale' a value from V memory to raw Analog Out and the Unscale will "unscaler" from raw Analog input to V memory.

2: Floating point still uses V memory just uses a period (.) to differentiate between INT and Float. It uses two consecutive registers. eg V100 is a 16 bit INt located at location V100
V100. is a 32 bit float located at v100 AND V101

3on't know much about the SFS's BUT I "think" they are like a sub-routine and can be called from within a SFP. Have used the SFP's though they are reasonably straight forward. If yoyu have a copy of tegh programming software try experimenting with it OFF-LINE to get some confidence.

Hi
This an example : (See attached file)

Bardes

1. I have 12 flows that I need to scale into the PLC.(What is the best way to scale analog inputs?[Ken you mentioned using a SFPGM) Is it through the analog alarm edit screen? Is there a way to monitor over-under current on analog channels?

Click to expand...

Three questions in (1) "The best way to scale analogue inputs?" If all you want to do is scale them, then use the SCALE instruction in an SFPGM as shown in the previous example. "Is it through the analog alarm edit screens? Is there a way to monitor over-under current on analog channels?" Well, yes, but now you're asking to do more than just scale them. Each analogue alarm function takes a raw analogue value in, scales it internally according to limits you provide, and then monitors automatically (at a frequency you specify) a bucket-load of alarm conditions: HiHi, Hi, Lo, LoLo, Yellow deviation (+ or -) from a target, Orange deviation (+ and -) from a target, broken wire (<4mA for that type of signal) etc. It is also possible to associate an additional SFPGM to be run in sync with each analogue alarm function to add your own custom code if the standard function isn't adequate. What you do in each of these SFPGMs is, like all the others, entirely up to you.

2. I need to be able to take these flow rates into floating point registers and perform the appropriate math to come up with basically a 0-100% signal which is going to 12 different grudfoss metering pumps. I assume that V-memory is allocated in 16-bit registers. If I need a floating point, do I need to plan on using two registers (V_x & V_x+1)?

Click to expand...

I think ConSys' explanation is clear on that. V-memory is V-memory and you can access it in 16-bit or 32-bit chunks as you please.

3. What is the difference between using a speical function subroutine or program? Does anyone happen to have an example?

Click to expand...

Remember the age of the PLC this kind of programming originated in (the TI 565) and the kind of computer languages around all these years ago. An SFPGM is a main self-contained piece of code and can't have any parameters passed to it. An SFSUB is a sub-routine, designed to be called repeatedly from one or more SFPGMs and can have parameters passed. The issue of T-memory (temporary memory) is quite sensible. It's possible you will require interim or temporary registers within the SF while calculating another result. Once you've got that result the interim results are of no further interest. Rather than have to allocate fixed V-memory addresses to all that temporary data, each SF has a set of T-registers associated with it. These can be used as a scratch-pad for internal calcs and then forgotten about until the next cycle of execution.

Regards

Ken

Thanks for all the assistance. I will let you know how it goes. You guys are more help than Fast-Track tech support.

Again thanks.

This is what I ended up going with. It seems to work (in the simulator at least).


Thanks

I would recommend putting a comment at the beginning of the subroutine that explains exactly what you're doing.