Signal Scaling

Would you please explain:
1- What is signal scaling?
2- Why and when we need to do the scaling?

Thanks my friends

Homework?

The definition in this case would be:
Act of measuring or arranging or adjusting according to a scale

Now apply that to PLC's and measuring devices. Think about it, you have a device that measures say 0..10 BAR pressure. This device "transforms" this pressure into an analogue value say 4..20 mA. Your PLC reads this value as a digital value say 0..27648. In your program, would you want to use a value of 27648 or something more intuitive as 10.0?

So..

In general, analog process instruments use a standard range of voltage or current to represent their measurements.

The most common ranges are 0-10 Volts for a voltage-type analog signal, or 4-20 milliamperes for a current-loop type analog signal.

Because those signals do not directly represent the units of the property being measured by the process instrument, it is useful to convert the analog signal to engineering units that represent the property that the instrument is measuring.

An example would be a pressure sensor that measures hydraulic pressure between 0 and 5000 pounds per square inch and puts out a 0-10 volt analog signal.

0 PSI = 0 V
5000 PSI = 10 V

If the relationship between PSI and voltage is linear, you can use a very simple linear equation to relate the voltage to the measured property.

In some PLC-based systems there are built-in scaling features in the input modules that do this onboard. In others, there are purpose-built scaling instructions, and in still others you have to write the logic yourself.

Thank you guys,

Based on your answers, I understand that the input card first converts analog signal to digital. Then the scaling means converting this digital signal or "count" to the engineering units.
Is this scaling done in the input card or in the processor? if it is done in processor, is there any specific instruction in the processor for this purpose? If so, what parameters do we need to pass to this instruction? how does the PLC know that the relationship between "Count" and the engineering value is linear or not?

Thanks again

the analog to digital conversion is done in your analog input module.

the conversion from "raw" value to engineering is done in your user program, and is basic math (linear example):

X = Input
Y = Output

////Xmax - Xmin
Y = ----------- * (X - Xmin) + Ymin
////Ymax - Ymin

Remember, the PLC (generally) does not know anything about what you throw at it unless you tell what it is or what you want to do with it.

Not all plc platforms allow for signal scaling in the analog input module. Some do, some don't. Of the input modules that allow scaling I don't specifically know of any that will perform anything but linear scaling. That doesn't mean that they don't exist; i just don't know of any.

There are specialty input modules that will perform non-linear scaling. But if you are dealing with a process involving non-linear conversion of a process signal to engineering units you will likely need to do this programatically in the plc.

Keith

You can get better answers if you specify which type of PLC you are using.

If this is homework for a class, please let us know. Asking homework questions is OK, but not if you pretend they aren't homework.

ControlLogix is an example of a platform where the analog input modules have a simple linear scaling function built into them. You tell them two points along the linear scale, and it does the conversion onboard.

SLC-500 is an example of a platform that has a purpose-built instruction; the Scale with Parameters (SCP) instruction is really just a general form of a simple linear equation that lets you convert from integer counts in the SLC-500 Input data table to engineering units in the user data tables.

Because most process signals are linear, most built-in conversions are linear.

If you have a nonlinear process, there are sometimes built-in conversions. A very good example are the built-in thermocouple conversion tables that a thermocouple module might use. You select the type of thermocouple and it chooses the correct built-in conversion table.

Sometimes nonlinear processes require more advanced math. It can be as simple as an offset or a piecewise linear equation, or as complex as a logarithm.

Thank you all,

This was not a homework or something like that. I am studying about the programming and I could not understand the meaning of the scaling. This is why I asked this question.

"Boneless" provided a formula for that which include X and Y. would you please tell me what the Eng. unit of these parameters are?

Thanks again

X depends on your Plc. For S7 it could be 0..27648. Where x is actual value between min and max
Y depends on your required value, say 0..300 Fahrenheit. Where y is your scaled actual value.

Thanks boneless,

you mean X is the "count" value which depends on the number of bits in ADC of the input card?

sayahan2003 said:

Thanks boneless,

you mean X is the "count" value which depends on the number of bits in ADC of the input card?

Click to expand...

Yes

I generally do not bother with scaling but use raw numbers in the program - pick off points. Sometimes I do need to use scaling and use a standard function block that has been developed for the PLC I use. The PLC I use also has a non-linear scaling function that linearises the result.

probably more than you really wanted to know can be found here:

http://www.ronbeaufort.com/y=mx+b_part1.pdf
http://www.ronbeaufort.com/y=mx+b_part2.pdf

I hope that this helps ...