Knowing the advantages of 4-20 mA, why would anyone use 0-10 vDC as a control signal?
In my field of building automation, I notice that a lot of end device manufacturers now use 0-10 vDC or 2-10 vDC control signals.
Are they cheaper or simpler to build, or electronically more reliable?
Plus, many controllers now have either fixed 0-10 vDC outputs, or outputs that can be changed to 4-20 mA by changing a jumper. I am curious as to what circuitry is bypassed or switched internally when the jumper is added. Does anyone know of a website where I could get a peek at the electronic schematics of a switchable transmitter?
(I have found schematics of 4-20 mA transmitters that allow a jumper to change the range to 0-20 mA, or the range of the internal inputs, but my searches were fruitless for the above scenario).
0-10v signals are potentially more precise. To convert a 4-20 ma signal to a voltage prior to A/D conversion, a precision resistor must be used. This resistor may be temperature dependent, which would introduce inaccuracies.
They might be using the precision resistor method to convert the 4-20 ma signal to a voltage before sending it out, or they might be generating the voltage using totally independent circuitry from the
4-20 ma loop circuitry.
-Robert Scott Ypsilanti, Michigan (Reply through this forum, not by direct e-mail to me, as automatic reply address is fake.)
0-10 Vdc, +/- 10 V dc, 0-5 V dc, 0-1 V dc and others are standards that developed out of different places and industries, along with the later 0-20 (I think) and 4-20 mA standards. A book could be written in this subject, and some have been. In some cases vendors reinvented the wheel to beat patents and as improvements.
I myself like the 4-20 mA as it more immune to noise and often intrinsically safe. It can also be used for voltage I/O by adding a resistor.
Because it is easier to understand by people that don't know anything about instrumentation. You've got to admit, 4-20 mA is not the first thing that pops into your mind. Also, if someone is inventing a new type of transmitter, making it operate as a two wire, 4-20 mA device adds a level of sophistication that may seem like a totally unnecessary complication.
In systems with very short wire runs and in a low electrical noise environment, the advantages are not evident.
This is not a valid reason. Firstly, I assume you mean "accurate" not "precise". Good resistors are available. Furthermore, every circuit is full of other resistors and components that affect accuracy. It is the presence of the resistor in a current loop that makes the loop immune to a number of error sources that affect voltage loops.
We prefer to use/design voltage outputs because they require less power. We can drive the voltage outputs easily with a DC to DC converter. 10 volts driving a 10K input requires .01 watts. 20 ma across 250 ohms is .1 watts. That is a big difference in power. If I had my way analog signals would be sent on a RS-485 type line as signed
16 bit numbers similar to the way SSI transducers report positions. Then the lines could be easily optically isolated too.
Actually, no, I wouldn't buy them. The fact that the signal uses less power is not relevant. The instrument as a whole still needs power and I suspect that it needs a lot more that .01 watts. What type of instruments do you design? How much power do they require to operate? How do they get it? It requires a rather high degree of sophistication to design a measurement instrument that operates on the 4 mA * 12 volts = 48 m Watts that is typical of 4-20 mA instruments.
The fact that the entire instrument power is carried by the signal itself provides high degree in noise immunity. What is your typical wire run length?
If you used an RS-485 signal, how would the device receive its operating power? That would mean adding another pair of wires. The user's costs start to run up.
I am pefectly aware of field buses. A field bus allows for multiple devices on the bus. However, updates can be slow with a lot of jitter and the microsecond level. If I have a motion controller that is controlling a valve then I may want to update the command to the valve in a much more deterministic means than is available with a field bus. Some of the big hydraulic manufacturers have tried using CAN to control valves. I haven't seen any success here in the US except maybe in relatively static applications.
If the valves could support SSI type control signals then the controller can SIMULATANEOUSLY update many servo valves. The would be only one device on each cable just as with 4-20ma now.
Some fieldbuses are deterministic in their timing, CAN not being one of them. RS-485, being asynchronous, will not be repeatable to 1 microsecond below 1 Mbit/sec, a speed rarely obtained over any distance. Field buses with deterministic behavior are non-trivial, which is why "fast enough" tends to dominate the market. There's also "feature creep". Once you have a digital link for data, it becomes very tempting to use that link for fault detection, firmware updates, two-way comms, etc. All of which require management packets as well as data to travel over the wire.
Relax, we don't make process instruments and that I why we see thing differently. We do control valves, mostly hydraulic valves and sometime pneumatic valves that are an control for a bigger valve controlling something else. Many valves now require + or - 10 volts and must have a amplifier to provide enough force to move the spool.
It dependend on the valve some require 4 amps at 24 volts.
From a 24 volt power supply
It
Making the instrument is easy with the new low power devices available. What about the controlling output card? I don't think you can do that simple DC to DC power supply and get the power from the back plane of a PLC.
Some times 100 meters as in a steel mill. Usually much less. In these cases the controller is often mounted in a J box with the valve amplifier and the main controller sends commands to the motion controller over a field bus. ( Are you reading this Gene?)
The valve needs power. More that 48 mw.
It means that AC power must be available for the 24 vdc power supply.
Fifteen years ago must servo valves could be driven with + or - 50 ma. The problem with this is that our controller had to generate this current which required an external power supply connected to the controller. When customers mis wired the power supply the isolated part of the motion card was toast. A lot of motion controller cards met an early end this way. Then there was the hassle of figuring who was at fault. We had customers that would send back controllers that were black from arcing and they wanted a warrentee repair. In the late 80's the hydraulic companies starting making proportional valves that required only + or - 10 volts. This allowed the motion controller to use only a DC to DC power supply to generate + or - 10 volts. In some cases the motion controller was in a PLC rack where the card can only draw .5 amp at 5 volts or less. We jumped on the +10 to -10 outputs so fast because now our controller wouldn't be destroyed because external power is not connected to our controllers.
I agree that in your world the 4-20 ma makes sense for devices. However, even in your world the control signal must be generated by a controller that generates the 4-20 ma signal. Most PLCs will not allow enough power from the pack plane to drive these 4-20 ma signals. Therefore they too have the same problem where an external power supply must be connected to the controling cards. I bet they have the same type of wiring problems and destroyed control cards.
You are looking at things from the device side. I am look at it from the controller side. I hope this clarifies things.
Excellent, no spelling errors! Everyone must use a spelling checker!
From experience I have found 4-20mA often poorly installed and of poor installation can make it frustrating to servive, calibrate maintain and repair unless you have good engineering standards.
What t needs special terminal blocks with test jumpers/ links to interrupt the current and allow test point access to divert it through a multimeter or calibrator. Also the terminal block system should include links to allow for the supply common.
These don't sound like a big ask but you know the way the world is.
Use the right terminal blocks that can be unlinked easily to test the current. Phoenix and Klippon spring to mind as terminal block suppliers with this type of block.
WorldFIP has demonstrated 1 usec accuracy, although it requires a master timebase (they claim that slave clock error is a larger contributor to timing error than the bus itself).
Delta Tau's MACRO was explicitly designed for motion control.
TTTech's Time Triggered Protocol (not exactly a fieldbus, but designed for determinism).
Once again we are discussing across industries that use the same words for different things. I'm talking process control and when I hear "fieldbus" I understand "Foundation Fieldbus". In our field we would not attempt to control the position of a valve using an FF signal. The FF signal would tell the valve where to go and the valve's internal controller (a part of the FF system) would position the valve where is was requested. It would also report this position and the required air pressure back to the central control system.
This can be a confusing newsgroup since it deals with such different industries which work on such different scales. In the process industries
48 mWatt (4-20 mA) is sufficient to operate the great majority of instruments. External power is only used for very complex devices such as chromatographs. Loop power comes directly from the DCS. 4-20 mA also comes from the DCS and is sufficient to power even the largest valves. I mean 24" diameter, 1000 psi throttling valves. But that is because the real motive power is in the form of compressed air at 100 psi. We almost never use direct motor operated throttling valves.
Our wire runs are typically in the hundreds to thousands of feet. We avoid field located electronics as much as absolutely possible. This is because:
a) We deal with potentially explosive atmospheres and maintenance absolutely hates opening X-proof boxes. They also hate closing them which is another problem.
b) The boxes are outdoors in temperatures form + to - 40 deg C. It's very difficult getting -40C rated electronics. It's even harder to get maintenance out there on days like that. They'd rather wait until spring.
c) Supplying DC power is another nuisance we don't need. We would get serious voltage drops. Local DC supplies have serious reliability problems. Our central power supply is fully battery backed and redundant with auto transfer.
d) Lightening does weird things.
e) Ground loops do even weirder things.
f) Etc.
So each of our preferences have their reasons.
Walter.
PS. I still don't like 0-10 V. I prefer the live zero for even more reasons. ;-)
The CAN bus can be coerced into deterministic timing -- the automotive guys have an extension to it that competes with the time-triggered protocol. But just saying "CAN" doesn't give the whole story, and the entire system has to be designed with deterministic timing in mind.
Are you refering to transmitter manufacturers or control system (receiving end) manufacturers? Having worked for a DCS manufacturer, the input range depends upon what part of the hardware you are looking at. The actual circuit board with the A/D converter may take a 0-10v (or other zero based voltage input) because that is what the A/D circuit takes. However, the terminal block is looking for a 4-20 ma signal. Somewhere between the terminal block and the circuit board is a high precision resistor to convert current to voltage. The A/D circuit converts voltage to counts (an integer number, perhaps from 0 to 4095), the software then looks at the counts that would come from 4 ma as 0% and 20ma as 100%. (eg. 750 to 3750). So the system is expecting a 4-20 ma signal, but there is 0-10 volt hardware present. This method will also handle signals slightly below 0% (4ma) or above
100% (20ma) due to slight calibration errors in the transmitter.
To avoid error due to voltage drop, the resistor is located only a few feet from the A/D, perhaps in the same or a nearby cabinet.
The many very goog reasons for using 4-20 ma rather than zero based volts as the signal from the field have been fully discussed in earlier posts.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.