0-10v, 4-20mA design

[blackjames]

my dad is an electrician, and would like a 0-10v supply with 4-20mA output aswell to test AC drives and similar stuff. last night i started searching on rs for an adjustable switching regulator. i found the ltc3631. it has a 100mA max output from 0.8-45v with 4.5-45 volt input. it has soft start, current limiting, and other features. it comes in a msop package wich is quite small but it costs $12 from rs and $6.13 from digikey. is there a suitable replacement with so few smd components?

[Jon Chandler]

Am LM317 voltage regulator and a pot will provide the same functionality for a buck or two.  See an LM317 data sheet for the configuration as a constant current source.

[IanB]

4-20 mA is usually used as a signal transmission standard for measurements and instrumentation. I suspect there is dedicated test gear for this purpose. You probably would want fine control and good precision so you can dial in specific signal levels. Noise from switching regulators might be a problem. I don't know that this is necessarily a case for a home brew solution.

[Psi]

Yeah, i believe it's current loop control. 4mA = zero   20mA = fullscale

It's a very common standard so there will be IC's for doing just this

[pickle9000]

The technical name for the instrument a V/I Source. Dedicated instruments would be called millivolt source and milliamp source.

...mike

[IanB]

Here's one document of many on the background to current loop control:

http://www.bapihvac.com/CatalogPDFs/I_App_Notes/Understanding_Current_Loops.pdf

[blackjames]

A lm317 wastes lots of power. My plan is to.connect a 500ohm resistor across the output. Apparently it gives you something like 0-20mA. Those professional units cost two or three hundred dollars and even if it costs 40 dollars to make one, im still going to be a happy teen.

[Zero999]

Am LM317 voltage regulator and a pot will provide the same functionality for a buck or two.  See an LM317 data sheet for the configuration as a constant current source.

Agreed, I don't think a buck is appropriate here.

Unfortunately the LM317 has a minimum output voltage of 1.25V and a minimum load current of 10mA so isn't suitable.

The LM317L will do for the constant current source, as it can reliably work down to 4mA.

There is a workaround for the minimum voltage but I doubt it's worth it in this case.

A lm317 wastes lots of power.

How do you plan to power this circuit?

Unless it's a small battery I can't see power dissipation being a problem. A linear regulator may waste power but even at 24V 20mA is only 480mW which is nothing. Not only is a switching regulator more complicated it's also more noisy. I think linear is the best way to go here.

My plan is to.connect a 500ohm resistor across the output. Apparently it gives you something like 0-20mA. Those professional units cost two or three hundred dollars and even if it costs 40 dollars to make one, im still going to be a happy teen.

The problem with that is the current will vary depending on the load resistance. Change the cable length or device at the end and the current will also change.

What's the maximum load impedance for the 4mA to 20mA current source? Is that it 500 Ohm plus the cable run.

What's the maximum current requirement for the 0 to 10V voltage source? Assuming it's low current (i.e 20mA). An op-amp buffered voltage reference is probably a better option."

[IanJ]

my dad is an electrician, and would like a 0-10v supply with 4-20mA output aswell to test AC drives and similar stuff. last night i started searching on rs for an adjustable switching regulator. i found the ltc3631. it has a 100mA max output from 0.8-45v with 4.5-45 volt input. it has soft start, current limiting, and other features. it comes in a msop package wich is quite small but it costs $12 from rs and $6.13 from digikey. is there a suitable replacement with so few smd components?

Hi,

4-20mA sink or source.....or both? Probably you need source, but it's handy to have sink for simulating/testing some instrumentation sensors etc.
Look up Burr Brown XTR115 for a couple of quid (and others in the series). Your design MUST be load independant for true 4-20mA usage.

Saying that i've been playing with my own 0-10vdc/4-20mA source design. Battery powered design using simple op-amps/transistors etc. Schematic is strictly beta/prototype, has too many pots, and a wee issue with the DC-DC......but might give you some ideas though.

http://www.ianjohnston.com/content/images/stories/IanJ/VoltageCurrentSource/VI_Source.pdf



Ian.

[IanB]

Where is the 0-10 V used, out of interest?

For a long time in my industry segment I have seen 1-5 V as a signalling standard (with 1 V being 0% and 5 V being 100%), and I've seen occasional references to 2-10 V in Google. Is that what 0-10 V means?

[IanJ]

Where is the 0-10 V used, out of interest?

For a long time in my industry segment I have seen 1-5 V as a signalling standard (with 1 V being 0% and 5 V being 100%), and I've seen occasional references to 2-10 V in Google. Is that what 0-10 V means?

1-5vdc is common in offshore E&I as it's simply the voltage across a 250ohm load when 4-20mA is applied.
You might just see 4-20mA, 0-20mA & 0-10vdc as just common signal ranges for transmitters etc. in many industries.

Ian.

[Jeff1946]

Keep it simple.  Two 9 volt batteries thru LM317 to give 10 Volts.  Also use the batteries to power op amps.  Put across a potentiometer to give 0-10V.  Feed to a op amp (single side type such as LM324 quad) as a voltage follower.  Now you have 0-10 volts.   Feed this voltage to the (-) input of an op amp with a 1% 500 ohm resistor.  Ground + input. Use output of the op amp and the (-)input as your 4-20 ma source (this is feedback of the op amp) the output should have a capacitor across it as could be driving a long twist pair of wires.  (2V = 4 ma and 10V = 20ma.  Make a suitably calibrated dial for the pot.  Put a green led with dropping resistor between input and output of the LM317.  If there is less than 2+ volts across the LM317, then time to change batteries.

You could add an LCD voltmeter for more accuracy if you need it.   

[IanB]

the output should have a capacitor across it as could be driving a long twist pair of wires

Isn't there a danger that putting a capacitor on the output of an op amp could make it go unstable?

[IanJ]

Keep it simple.  Two 9 volt batteries thru LM317 to give 10 Volts.  Also use the batteries to power op amps.  Put across a potentiometer to give 0-10V.  Feed to a op amp (single side type such as LM324 quad) as a voltage follower.  Now you have 0-10 volts.   Feed this voltage to the (-) input of an op amp with a 1% 500 ohm resistor.  Ground + input. Use output of the op amp and the (-)input as your 4-20 ma source (this is feedback of the op amp) the output should have a capacitor across it as could be driving a long twist pair of wires.  (2V = 4 ma and 10V = 20ma.  Make a suitably calibrated dial for the pot.  Put a green led with dropping resistor between input and output of the LM317.  If there is less than 2+ volts across the LM317, then time to change batteries.

You could add an LCD voltmeter for more accuracy if you need it.

That would work, but be aware a good 4-20mA loop driver will operate at higher voltages in order to drive larger loop resistances. Typically you might have a loop powered indicator, a zener barrier and a PLC input all in series and contributing to the overall end to end resistance, hence to obtain 20mA you need that higher voltage overhead. 24vdc or 28vdc are common, but can be anywhere between 10vdc & 40vdc. Probably why my old 4-20mA had 3off PP3 batteries in it.
If I remember correctly a good loop powered indicator should only drop 1vdc across it, but I have seen some which are much, much higher.
Saying all this, you'll probably get away with 10vdc.... .....just depends on the hardware/system you are tying into.

Ian.