Raw Range / Engineering Range

[tcpdump 0]

Hello, The PLC I am programming at the moment has a concept of a "raw range" and an "engineering range". How it works is that, the instrument provides a signal of 0mA to 20mA, where 0mA refers to a raw value of 0, whilst 20mA refers to a raw value of 1000. These raw values between (0,1000) are then converted to a value between (0,100) and then both of these values raw and engineering are given to the MS. Why do we bother in having the raw/engineering values?

[Nathan 7]

A 4-20 mA signal is standard. I'm guessing that 1000 discreet values represents your accurate resolution. Given that, your chosen value range is 0 to 100 in steps of 1/10. You could have just as easily picked 1000 to 2000 in steps of 10. It makes sense to me to first get a discreet value from your input before running it through your linear interpolation. Perhaps there's a standard that I'm missing...

[PdL 71]

If you have a 12 bit resolution analoge input channel, your raw input range is 0-4095, 0 representing the minimum value of your analogue signal, 4095 representing the maximum value. If you have an 8 bit resolution input channel the raw input range is 0-255. Of course the analog input card has to be suitable for a specific range. If you set it up for 0-20mA, and you connect a 4-20mA sensor, you won't have 0 as a raw value when the sensor output is mimimum (4mA), but 819. (1/5 of 4095). If you want the 4-20mA to represent a temperature value of -10,0°/60,0°, than that is your engineering range, also referred to as scaled range. The PLC instruction set you work with probably has a SCALE function or some sort, that does the math for you if you define the min/max raw and min/max scaled values. When executed it will constantly calculate an engineered value according to the raw input value.

[comeng 0]

Simple answer the wondeful world of sensor electronics can only deal with currents and voltages 0-20ma 4-20ma 0-10V -10+10V etc the real world deals in temperature, pressure, length, speed. Thus raw and engineering units why pass both to the MS only if someone does not believe PLC can perform the math correctly

[tcpdump 0]

thanks 4 your replies. they were very informative, and explain WHAT is going on, but to some degree they lack as to WHY we have a raw and scaled range. Why not from the 4mA to 20mA signal, we convert to the 0KPa and 100KPa straight away???? This reduces processing time, storage capacity. etc... Now, what is your experience with these kinds of things?? Do your PLCs have both a raw and scaled range for measurements????

[panic mode 246]

why not everyone drives bentley to work? it's not a bad car... it's all about money. the more you pay the more you get but - you still want to share common roads or 0-10V / 4..20mA range. the better cards will have same standard range (0-10V or 4-20mA etc) like their cheaper counter parts but the resolution, response time, temperature stability, noise immunity, overload range etc all play a role... so how many 0..100kPa sensors you want to buy and what is resolution going to be? one count = 1kPa? so your reading will jump 1, 2, 3 etc. instead of showing 1.000kPa; 1.001kPa etc. how do you represent values 0..100 and 0.000...100.000? and if you need range 20-120kPa next day, do you expect to pay big $$$ for someone to develop it quickly since all transducers on the market are ... well... 0-100kPa... maybe your analog card will now have to be new one too since 4..20mA = 0..100kPa then 20-120kPa should be something else? what if you want to connect more transducers of different type on a same analog card? or do you think that every transducer should have separate input card specially matched for particular transducer? good luck with purchasing such system that can do process with 500-1000 PID loops.