4-20 ma driver

Hi, anyone have schematic for a 4-20 ma driver with power supply, change in ma range with a pot. i'd like to send in the neighborhood of 24 v into a regulated power supply and have an output signal between 4-20ma depending on position of a pot.meter. Thanks, Frank

See attached... [and, actually, that could be an LM317L (instead of an LM317)]

And in case you don't know this, potentiometers can have tolerances as bad as 20 (or even 30) % so you'll, likely, need to add some trimmers in there to get the right values.

Attached is a version that is not so sensitive to potentiometer tolerance or even resistance value (will work with a wide range of pot values, so things like current drain and noise immunity will drive component selection.

Also, this circuit is actually a voltage to current converter, so if you have an application that needs to monitor a voltage and convert it into a 4-20ma signal, then this circuit will fit the bill.

To figure out the values of the resistors, R3 is what converts the input voltage into a current. The op-amp will do it's best to keep the voltage across R3 the same as the voltage on it's non-inverting '+' input -- lets call that Vin+. So, the current in the send loop is given by: I = Vin+/R3. So, just choose R3, then choose values for R4,R5,R6 that will produce a voltage range that will result in a 4-20ma range for 'I'. The trick for choosing R3 is to have Vin+ not go too low, such that it will be in conflict with noise (environment and circuit layout will determine that), and not too high that it will too much limit the range of voltage in the current send loop. OR if monitoring a voltage, then it's range will influence the selection of R3 (and some voltage translation might be necessary).

Also, use a high beta transistor, as the current in the loop, IC, is actually IE - IB. It's up to you to determine how critical this error is. A good high beta transistor is the 2N5089 (or even the 2N5088). Or you could even use a Darlington (with a slight loss in current loop voltage range), but in most cases that would probably be overkill.

i get very easily lost in the jungle of components.
if i see a schematic that calls for an op-amp or transistor, i don't have the knowlegde to pick out the right one.

one of the uses for this would be to give a 4-20ma signal to an analogue input on a plc.
in that case i want connect a pot(or some sort of variable resistor)to the circuit so i 'send' 4ma to the input when pot at one end and 20ma when pot turned to the other end.

and i always have a 24v power supply to feed the circuit with.
previously i have achieved this with 0-10v signal with just calculating resistors, but this signal is too 'jumpy' most times and 0-10v is more prone to interference than a 4-20ma signal(or so i hear).

For the Op-amp try:
* LM358AP (Digi-key: 296-9554-5-ND)

and for the transistor, I suggest the 2N5088, but just about any NPN small signal transistor will do -- make sure the Vceo is above 24 volts (and probably best to make it 30 volts or above). You could even use a PN2222A, or a 2N3904

Hi,

I prefer shopping with RS because my company has an account there, i could'nt find that op-amp there but would MC33078N be an ok replacement?

Thank you for helping me out.

Fraknk

Yeah, it looks like the MC33078N will work (can't be sure, though, 'cuz I'm never worked with that part, but based on the datasheet it appears to be OK).

Im not sure if i'm calculating this correctly but if r3=1k, r5=1k,r4=5k and r6=1,2k is the ma range
4-20,6?

Yup, you got it right, but with R3 set to 1K you will only have about 4 volt headroom at 20ma because R3 will be dropping 20V at 20ma. If your 4-20ma loop will never be long enough for it to drop more than 4V, then this will work (but also consider tolerance for contact corrosion over time, noise immunity and stuff like that). The lower you set the value of R3, the greater the loop voltage headroom. But, with the MC33078N you are limited by the minimum Common Mode Voltage Range which will only allow a minimum voltage at R3 of 2 volts, so 500ohm is as low as you can go on R3 and still have regulation.

With the LM358, the Op-Amp input can go to 0 volts, so with that device you can set R3 very low. Then the only consideration is noise, so I would keep the R3 voltage at .5V or above.

Thanks for all your help with this,
could i step up the voltage with a simple circuit to maybe 35v? it would need to be big enough to handle 4 current loops.
I think i'm going to make up a couple circuit boards with 4channel 4-20 current loops. for tuning this i suppose i should have a trimpot in here somewhere, any pointers on how to incorporate that?
should maybe r5&6 both just be trimpots?

edit:
never mind the stepup, if i can find an op-amp that can handle o,7v i can with r3=200 r4=500 r5=300 r6=120 get 3,9ma to 20,5ma with a maximum voltage drop of 4,1v?

Frank

It's hard to answer on such vague information (such as details about the application -- will each loop be controlled by a variable resistor, and other such things that I'm probably not thinking of because I have no clear vision of how this is going to be used).

But, generally, just do 4 versions of the same circuit. No need to step up the voltage. But, if they are all to be powered by the same supply, then it needs to handle, worse case, 4x20ma+4x[whatever quiescent current is involved].

Also consider using the LM324. It's a quad op-amp designed for use with a single power supply and has inputs and outputs that go all the way to ground.

what i need this for now is simply to send ma signal to an analogue input on a plc, most times using a potentiometer.

if i go for cheap pot's with +-20% accuracy, is that inaccuracy for rated ohm, meaning that are they accurate in terms of repeatability?

Tolerance on a potentiometer [or 'pot'] is like tolerance on a fixed resistor. If a pot has a tolerance of 20% that means that the overall value of the pot can vary as much as 20% above or below the nominal value (the value stamped on the pot). So if you had a vat full of 20% 10K pots and you started measuring them, you could get values as high as 10K * 1.20 or 12K, and you could get values as low as 10K * 0.80 or 8K.

Temperature stability is yet another parameter, and some pots are more stable than others. Then there is number of rotations, or endurance.

As for feeding PLC inputs, just use one circuit per input. They can all be powered from the same supply or different supplies.

If i have an unstable input voltage, could i do it like the attachment?
the regulator circuit is copied from a 12v regulator i found, how will r1 be for 24v (20v zener)?

I would just use either an LM7818 or if 18V is too low, an LM317 programmed to 20V. The 2N3055/zener circuit is not very consistent and has poor regulation compared to the IC regulators, and the IC regulators require fewer parts. You don't need all that much current, after all, and, in fact, the low power versions should serve you well (e.g. LM78L18 or LM317L)

so all i need for the lm317 regulator to work is 2 resistors, or 1resistor and 1pot.?

Yep

please tell me the no. of opamp and transistor