Which signal to choose?0-10v or 4-20ma?

Hi,

I've done a research here for that and didn't come with an answer so here I am.

We are building a plastic machine and it would be my first project with analog signal.

In the setup of the machine, it will have linear transducers for the clamping and another for the injection ram (shot size). Also, it will have an hydraulic pressure transducer for the injection pressure.

For the injection, we will use pressure reducing/relieving valve and the pressure transducer will close the loop.

My question is : as I have 3 analog inputs and 1 analog output, what would be the best signal for me.

As far as I know, liner transducer are less expensive if we use 0-10v but for the rest, I don't know which one to choose. I'm a bit affraid to choose the wrong one and we all know proportionnal valve and analog stuff are expensive so I don't want to make a mistake.

All the control of the machine runs on 24v and it will have a motor drive somewhere on the machine so maybe noise could be a problem?

Shed some light on me please.
Thanks!

With 0-10VDC how do you know if you are really reading 0 or have lost the signal and do you care?

Ok I understand. If I use 0-10v I would arrange things to not have any zero reading.

Anyone have other suggestions?

Philippe said:

Ok I understand. If I use 0-10v I would arrange things to not have any zero reading.

Anyone have other suggestions?

Click to expand...

Invalid assumption. Often, if a 0-10VDC line goes open, the input will see a random non-zero value, as it is a high impedance input, and very susceptible to noise, or even adjacent channels on the A/D convertor (Phantom Signals). This can be circumvented by ALWAYS putting a 10k to 47k resistor directly across each input.

Still, as an output, you have no knowledge of a cable break. In general, for almost all process control, a 4-20mA signal is the overall best bet. It has much higher drive capacities, it is much less troubled by noise, and provides inherent fault detection. If you aren't running multiple items in series on a loop, you can even use an extra terminal, and wire in a dedicated 250 ohm resistor in series, so you can still do quick troubleshooting with a voltmeter, without having to break the loop.

Also, many 4-20 devices are HART compatible, which might be an advantage in the future.

rdrast said:

If you aren't running multiple items in series on a loop, you can even use an extra terminal, and wire in a dedicated 250 ohm resistor in series, so you can still do quick troubleshooting with a voltmeter, without having to break the loop.

Click to expand...

Instead of using the 250 ohm resistor,
What I do is fit a forward biased 1 amp diode, then to test the current flow, I connect the multimeter leads across the diode leads, setting the scale to milliAmps, and measure the current flow in the current loop, as the current will take the path via the meter, rather than the diode whilst I am using the multmeter.

If you may, just use 4-20mA signal instead of 0-10V. First of all you will be able to detect open wire, and second - current signal is more stable to el. noise then voltage one.

Go with the 4-20ma signals - much less trouble

Gil47 said:

Instead of using the 250 ohm resistor,
What I do is fit a forward biased 1 amp diode, then to test the current flow, I connect the multimeter leads across the diode leads, setting the scale to milliAmps, and measure the current flow in the current loop, as the current will take the path via the meter, rather than the diode whilst I am using the multmeter.

Click to expand...

I have never seen that done. Interesting...

Thanks, Paul...

Although I don't have much experiance either I have had terrible times with diffrent sensors, and I'm usually stuck using things that go to other machines to keep ordering easy and the supply room strait.

I recently used a current ring that scaled out 4-20 and it was great.

I have been using some temperature sensors that have been imposible to figure out the scale I have even spent hours using various temperature water (32 - 212) trying in vain to find some kind of scale.

So go with whats easy.

If anybody wants to take a crack at one of the sensors let me know I might send one to you, and I can't remember the name (type)right now but will look tommorow.

Clint

4-20mA is superior for long distances, like those encountered in a steel mill, oil refinery or chemical plant that covers, literally, square miles - where
- signals have to go long distances with assurance of a live zero (4mA, rather than 0mA),
- good resistance to noise
- wiring needs to be only 2 wires per circuit, where the power to the transmitter is supplied over the signal wires.
- smart functions like re-ranging with HART is important for inventory and processs reasons.

They use process transmitters like this:


Those criteria are not as critical (if at all) when instrumenting a molding machine, where 6 or 7 steps covers the length of the machine and 2 or 3 steps cover the width of the machine. You don't drive a pick-up truck to get out to the transmitter. Running 3 wires, rather than 2 wires, over a 'long' run of 10m is no big deal.

Note that most 'smart' process grade pressure transmitters that are 4-20mA devices produce an output change 3x to 10x per second, depending on brand & model. Is this response adequate for the injection pressure?

In comparison, industrial pressure transmitters, frequently 3 wire 0-10V models are either analog or low end digital uP models with 1mS-10mS response, limited scaling (tweaking a zero or a span pot), but they work great as a dedicated device. Many models come in either voltage outputs or current outputs.

Industrials tend to look like this:


As long as you know your pressure range up front, one of the industrials would serve you well.

Dan

Thanks guyz,

After all replys, I'll go for 4-20ma.

So now, do you know a good brand of linear transducer not much expensive that will output 4-20ma? I know pepperl-fuchs, balluff, it seams that it's in the range of 200$ for a 9" stroke. Of of our plastics machine which is an Uniloy (US machine) have some novotechnik on it but how does it work? On voltage? Ohms? It seams thats our machine get volts from them...

On the flip side of the coin, techs usually have a way easier time understanding and diagnosing 0-10V, so if it is for something simple like speed control for a motor, or the run isn't too long or is in a low noise environment, I'd stick with with 0-10V.

4-20mA is generally the better choice from the engineering perspective though.

Is this common, that a current analog output is generally slower then a voltage output?

Plc_User said:

Is this common, that a current analog output is generally slower then a voltage output?

Click to expand...

It really depends on the spec, but I would read the spec sheet real closely for a 4-20mA output device to see what its frequency response is, if frequency response is a concern.

History shows that the process industries invented and standardized the 2-wire, loop powered (4-20mA) current loop.

Generally speaking, noise reduction is of bigger concern than frequency response, to many large scale processes.

Many processes are either not changing at a high rate, or the output end of the control loop can't provide sufficient energy/medium to react to an input change, so process readings tend to be slower than what one can find in discrete manufacturing. Filling a 2cc bottle can be much quicker than filling a 30,000 gallon storage tank.

That doesn't limit 4-20mA current output to only slower reponse, but historically that's where it comes from.

Fast responding pressure transmitters can have either a voltage or a current output, either can respond within the response spec. But there are distinct differences between slower responding process pressure transmitters and their industrial pressure transmitter counterparts.

Dan

Thanks for all the reply.

I understand that 4-20 would be great for some security issue but I don't find any linear transducer or position mesurment sensor that are not too expensive (balluf) or that they are not the type I want (pepperl-fuchs indusctive position sensor).

Anyone have an idea who makes 4-20ma position measurment sensor? If not, I think I will be constraint to have 2 analog inputs cards in my plc rack and use novotechnik position pot for position and 4-20ma sensor for pressure.