Analog ~ what is your choice for a PID

Just wondering what everyone is using and why? im building a new trainer and it will have analog controlling a PID and feedback, I use 0-10vdc now but wondering if it would be better 4-20, 0-20 or its all the same as long as it works

Cost is also a factor (as always) so if there is not much advantage in going either way then price wins

Voltage might be a better choice for a trainer, but for "real life" I always prefer 4 to 20mA, and it might be a good thing for training to learn real life stuff.

+1 OkiePC

You can always use a resistor and get a voltage from a 4-20 signal.

I always go with 4 to 20 ma ....

all of my "maintenance tech" type students have either:

(a) worked with ... or ...
(b) at least "heard about"

4 to 20 mA signals ...

on the other hand, VERY few have to deal with analog VOLTAGE signals once they're out on the job ...

personally, I would't use 0 (zero) to 20 mA - because you'll want to be able to demonstrate the benefits of the "Live Zero" function of the 4 mA signal (alarms for "broken wire" etc.) ...

on another subject:

how fast does your trainer oscillate? ... specifically, when you put the Proportional action REALLY HIGH (with no Integral and no Derivative) the system should go into sustained oscillations ... if the period of oscillation is more than a few minutes or so - you'll have a tough time using the trainer in a classroom for "tuning" exercises ... that's mainly because any change that the student makes to the tuning parameters, will take too long to "settle out" and make it obvious whether the change was "too much - too little - and in the right or wrong direction" ...

good luck with it - and let us know how it turns out ...

Current transmission is less sensitive to noise than voltage transmition.

There are 4-20mA transmitters that use the output current to power its internal electronics and can therefore be connected with only 2 wires. Simplifies wiring.

My vote is also for 4-20mA

Over distances, you will have voltage drop along a cable and at each terminal. the voltage drop will change as the system ages, due to loosening terminals, oxidisation etc.

also, noise is picked up as a voltage, so a good current regulator can be noise free.

0-10V is only used if you don't really care about the value, OR if you are using remote IO and keep the cable distance short and the noise immunity high (no nearby vfd or AC cables).

As one example, I have a sensor from company 1 which is £100 and a sensor from company 2 which is £3000. the £100 sensor is 0-10V and the £3000 one is 4-20mA. I really care about the value, so I mounted a RIO adapter 0.3m from the sensor.

Ideally you would have a trainer with both 4-20mA and 0-10V, an RF noise source, and a switch to increase the resistance (simulating years of oxidisation and terminal loosening). This is a very important lesson to learn, that most people learn the hard way.

Use 4-20mA and have them break the circuit with a meter to determine the current being supplied..

This will teach them what people see in real applications, and how to verify the signal. Seems simple, but I doubt most people have actually done this.
Also like most people have said they can see the result of a broken wire. (0mA)

If you could please explain to your students the actual function of the transmitter you're using to measure.
At a previous employer I got into an argument with "Super Tech" about the fact that his calibration of a transmitter at their facility has nothing at all to do with the calibration of said transmitter at my facility.
" We don't need to calibrate the pressure sensors. I already did it at the shop."

I like 0-10V because it is easier to troubleshoot as far as using a meter.
BUT..4-20mA seems to be more accurate in distances and you can see error on loss of connection in less than 4mA.

Ditto the comments above about volts vs amps.
However, a critical item is feed back resolution and consistency of the sample interval.

AustralIan said:

Ideally you would have a trainer with both 4-20mA and 0-10V, an RF noise source, and a switch to increase the resistance (simulating years of oxidisation and terminal loosening). This is a very important lesson to learn, that most people learn the hard way.

Click to expand...

That would be ideal!

As a systems integrator, I avoid 0-10V like the plague wherever possible in real life applications. There are just so many benefits to 4-20mA (live zero, better noise immunity, no voltage drop concerns) that I'll push 0-10V to the outer darkness (where there is weeping and gnashing of teeth) as quickly as I can.

From a trainer perspective, I would recommend going the 4-20mA route:
1. I can't be the only one that sees all the benefits to 4-20 over 0-10, so surely 4-20mA is becoming more common everywhere, not just in the factories I work in
2. 4-20mA is slightly more complex to test than 0-10V. Anyone can whack a DC voltmeter on two terminals, but when you have to find a low-range clamp meter or work out how to put your meter in series, and work out what the process implications will be of disconnecting the loop for that time to insert your meter, it requires a lot more thought
3. You can demonstrate the advantage of loop powered transmitters

Of course, the ideal thing would be to have a card that can do both, and be able to demonstrate the pro's and con's of both methods. Assuming you can find a "pro" for 0-10V

4-20mA is slightly more complex to test than 0-10V. Anyone can whack a DC voltmeter on two terminals, but when you have to find a low-range clamp meter or work out how to put your meter in series, and work out what the process implications will be of disconnecting the loop for that time to insert your meter, it requires a lot more thought

Click to expand...

A nice feature of the trainer would be to run the 4-20 mA signal across a DIN rail mounted fuse block. When you need to measure the signal put your meter on the terminals, pull the fuse and the meter is in the loop with no disruption of the signal and no upset to the process. Just don't forget to put the fuse back before you disconnect the meter.

You don't need a current clamp, or put the meter in series with the loop.
The Analog input card is voltage input, with a precision resistor across the input terminals to make it a Current input. Most use 250 Ohms. That means you would read 1-5 volts at the terminals.
500 Ohms is 2-10 volts.
The resistor on a P2-08ADL for an AD P2000 is 124 Ohms. That's .5 - 2.5 Volts.

That's true - but in the real world, that assumes you're at the PLC card, not out in the field, and have access to the information about the particular analog input that's there.

And for the purposes of the trainer, learning all of the above and calculating your mA reading based on a voltage reading at the card would be another great learning exercise that you couldn't do with a 0-10V input!

You can measure Ohms at the field end if you can disconnect the wires. If not, then measure voltage, and assume a 250 Ohm load. If the input impedance is different, readings will be half or double what's expected.

Putting meter in current mode and breaking the line is the right way to do it. If that's not possible, there are several more options.