Is there a (IEEE, EIA, whatever) specification for voltage levels in a
4-20mA signaling loop? Or is there some common way of specifying the
amount of power available to a loop-powered device? If it isn't
specified, is there some common usage?
I'm curious as to how much freedom one has to power ones device when one
designs some gizmo that flaps in the breeze on the end of a current loop.
But I'm not designing anything right now.
That's really a system spec. The limit is the power supply voltage. 40V
is about the most you'll find, and 24V most likely. The drop across
other devices that may be in seried tells you how much you have left. At
20ma. At 4 ma, you have 1/5th of that. Google will turn up a lot of
Engineering is the art of making what you want from things you can get.
I'm curious about a "standard" as well. I _think_ the device
electronics must simply consume less than 4mA, so that "zero signal
input" can cause a draw of exactly 4mA??
dates back 8 years, in answer to an original post by Spehro.
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
Loop powered systems usually have a worst-case low end of about 3.6mA
+/-. They have a minimum and maximum voltage so you can determine the
maximum supply voltage and calculate the maximum load resistance. They
can go over the 20mA, usually by 10% or more, sometimes as bad as 30mA
total (for example, as a result of sensor failure) so you should allow
for that. The electronics can use more than 4mA but you'd have to use
a SMPS and eat more minimum voltage. Most will work to well under 10V,
so maybe you could get 6 or 8 mA at 3.3V. Often you need to run at
least one power supply with galvanic isolation plus a microcontroller
and perhaps a display off of that power, plus whatever signal
conditioning, sensor excitation, ADC converters and other stuff is
required to get a signal into the micro, meaning the power supply
can't be wasteful (nor are commercial modules usable).
The maximum load resistance limitation usually comes into play when
several loads are connected in series (and perhaps a lot of wire in
the middle). For example, a field-mounted loop-powered indicator
connected in series with the SCADA system input.
Four wire (non-loop-powered) systems don't have the lower end
limitation, but if you want to tell the difference between railed at
the low end and a broken loop you need to avoid going right to zero.
There's a maximum load resistance specification. If you're going
4-wire it's best to galvanically isolate things as much as possible as
the number of ways to f*** things up increases exponentially.
All the systems I've seen don't mind zero resistance load.
I don't know of any standards, but the above are examples of common
usage. You might be able to find some more stringent 4-20mA specs as
the analog part of the HART (hybrid analog/digital) standard.
Likewise I have never come across a specific standard for the 4-20mA loop
systems but I have seen a range of excitation voltages used in circuits that
feature them. Most I dealt with were the teleprinters running on 60VDC
supplied 4-20mA loops.
Wikipedia has some useful info.
Paul E. Bennett...............<email://Paul firstname.lastname@example.org>
You'd have to go by the compliance range given in the datasheets:
The standard would be ANSI/ISA S50.1 which I don't have but parts are
As Lasse hinted you'd have to obtain the compliance ranges of all the
devices in the system. After all, it won't do you any good if the whole
thing complies to the standard but doesn't work reliably.
Tim, I may be corrected but I'm not aware of anything. I've always consulted
the system specs, which generally specify in terms of maximum allowable loop
resistance. It's pretty common for equipment these days to be powered from
24V, so that gives one an indication, there's some voltage loss across the
current source electronics, plus the loop resistance voltage, plus whatever
you use for current sensing. AFAIAA, 250 ohms, or 1-5 volts, for input
circuits is still typical (but in practice it can be far lower).
On Wed, 7 Oct 2009 12:06:28 +0800, the renowned "Bruce Varley"
Yes, the 24V (or 32V or 40V or whatever) gives you an idea of
worst-case power dissipation, but there can be long wires on these
things and they need protection against relatively high transient
voltages (think lightning strike at the old refinery), which usually
means discrete components, not ICs, will take the brunt it.
"it's the network..." "The Journey is the reward"
email@example.com Info for manufacturers: http://www.trexon.com
Thanks for reminding me, if your loop is category Ex-I (intrinsically safe)
then the loop has to support the voltage drop of the safety barrier if one
is used. Barriers can have significant voltage drop.
I would say an excellent question; another way is to ask what is the
If one wanted to be really bizarre and have a large voltage
compliance, then the open circuit voltage could be up to 100KV which
would allow a small break in the circuit to be "ignored".
This would require a quite large loop current also.
While the high open loop voltage may allow sparks to occur across air
gaps, the created ionization does not last long, unless sufficient
power is supplied to compensate for the radiation losses.
Of course with long wires and large stray capacitances that would
create a resonant circuit, so this would be a nice spark gap
transmitter and as such, could be used for wireless communication :-).
Some of the transmitters that are 4 to 20 ma loop powered require
about 2 ma to operate. Typically, the receiver device has a 250
resistor to provide a 1 to 5 volt signal to it. The power supply is
typically 24 vdc. So if you allow 2 ma for operation, then for a zero
signal input to the transmitter you get 4 ma output and likewise an
extra 18 ma is added to get a 20 ma output at the 100 percent input to
the tramsmitter. The Motorola Veritrak Diff. Press and Press
transmitters did this. You could adjust the transmitters to use any 25
percent of the range for the 0 percent to 100 percent outputs. Foxboro
used a 10 to 50 ma system but used a 100 ohm resistor to get the 1 to
5 volt signal to the receiver device or controller.
Any way, you have to live with between 0 ma up to 4 4 ma electrical
output zero for your device. The transmitter amps are just fancy
electronic rheostats that vary from an effect 23K ohms to get 4 ma at
a 0 percent signal to the transmitter down to 950 ohms to get 20 ma at
the 100 percent signal to the trasmitter. That could be some thing
like 0 to 400 PSIG or 0 to 100 inches of mercury differential
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.