Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

4-20 mA ADC IC needed

Status
Not open for further replies.

gtslabs

New Member
I need to read the pressure from a transducer that puts out 4-20 mA. I
was planning on putting an ADC across a precision 250 Ohm resistor to
read the voltages from 1v to 5v. I need at least 16 bits of
resolution. I have been searching data pages for a good 16bit ADC but
most ratiometric ones are limited to 2.5v. I use an AD7730 24 bit for
some projects but it is limited to 80 mv I believe. Do I need to use
a Voltage divider to get the voltages low enough or are there any 16
bit ADC that can handle the 1-5V? I dont need very fast output, say
less than 20 Hz.
 
Are you sure you need 16 bit?, most of the resolution will be lost in noise anyway, and to achieve anything like 16 bits requires extremely good PCB layout. I'm also dubious that 20mA current loops can handle anything like 16 bits in the first place?.
 
As Nigel has pointed out 16 bits will be a stretch on a 4-20 mA current loop. But if you are bound and determined then use an opamp as a unity gain follower. If you need to adjust the levels for the A-D converter then you might want to consider two stages since a non-inverting stage cannot have a gain of less than 1. So in two stages you can convert the 4-20 mA signal to a voltage in the range 0-2.5 volts.

Now go back to your 250 Ohm resistor and calculate the inherent noise in that device according to the bandwidth of interest. Then figure out how many of the sixteen bits of your A/D conversion will be swampped by the noise in the resistor. You might just find that a 10-bit converter will be more than sufficient. You might find the following link helpful.

http://www.ecircuitcenter.com/Circuits/Noise/Noise_Analysis/res_noise.htm

BTW 2.5/65536 ≈ 38 μV
 
Last edited:
hi,
Are you limited on how the ADC data is output?
eg; 16 bit 12/4 bit parallel or will a serial string be OK.
How do you plan to interface to the ADC output?

As Nigel has pointed out 16bit resolution from a 4 to 20mA is not going to give you what you are expecting. IIRC thats about 76uV/bit!!!

I use the MCP3202 dual 12bit with SPI control, using a PIC, the MCP handles 0thru +5V input.
 
Papabravo said:
As Nigel has pointed out 16 bits will be a stretch on a 4-20 mA current loop. But if you are bound and determined then use an opamp as a unity gain follower. If you need to adjust the levels for the A-D converter then you might want to consider two stages since a non-inverting stage cannot have a gain of less than 1. So in two stages you can convert the 4-20 mA signal to a voltage in the range 0-2.5 volts.
[/url]

hi papabravo,
One way I have used in the past, is to use two 125R's connected as a divider, the junction of the divider going to the non-inv input of the opa, this gives a buffered output of 0 to +2.5Vout.
 
In commercial process control equipment I've never seen more then a 12 bit A/D used to process 4-20ma instrumentation loops, usually measuring a 1-5vdc drop across a 250 ohm resistor.

Keep in mind that your pressure transducer will most likely set the accuracy and resolution limits of the measurement loop. Check it's accuracy specifications and then reevaluate your A/D requirements.

PS: If one did wish to measure a 4-20ma loop and was limited to a 0-2.5v A/D range one would simply use a 125ohm precision resistor rather then a 250ohm resistor.

Lefty
 
Last edited:
Thanks for all the comments.

I need to measure the force of a hydraulic piston via a pressure transducer in the line. My piston is 8" diameter and my pump is 2000 psi. So my max force would be 100,480 lbs. I am using a "Red Lion" PAXS readout on another machine that is set up for 10,000 psi on a 6" piston. The lowest resolution is 10 lbs force. I can live with 10 lbs. But to get that I need to divide the force into 10,048 steps. If I used a 16 bit ADC then the resolution would be 1.5 lbs. So I can have some noise if I only need 10 lbs.

The sensor I am using is a 3000 psi Omega PX315-3KGI which has an accuracy of 0.25% BFSL so at 2000 psi I would have an error of 5 psi or 250 lbs. My ALLOWABLE system error is 1% (of Reading) or 1,000 lbs.

So am I overkill with 16 bits? I still need the 10 lb resolution.
I am using a pic.
Thanks
Steve
 
gtslabs said:
Thanks for all the comments.

I need to measure the force of a hydraulic piston via a pressure transducer in the line. My piston is 8" diameter and my pump is 2000 psi. So my max force would be 100,480 lbs. I am using a "Red Lion" PAXS readout on another machine that is set up for 10,000 psi on a 6" piston. The lowest resolution is 10 lbs force. I can live with 10 lbs. But to get that I need to divide the force into 10,048 steps. If I used a 16 bit ADC then the resolution would be 1.5 lbs. So I can have some noise if I only need 10 lbs.

The sensor I am using is a 3000 psi Omega PX315-3KGI which has an accuracy of 0.25% BFSL so at 2000 psi I would have an error of 5 psi or 250 lbs. My ALLOWABLE system error is 1% (of Reading) or 1,000 lbs.

So am I overkill with 16 bits? I still need the 10 lb resolution.
I am using a pic.
Thanks
Steve

hi,
In that case, do as Lefty suggests, use a 125R load in the 4 to 20mA line, two 250R in parallel will give you +0.5 to +2.5v for the ADC.:)
 
gtslabs said:
So am I overkill with 16 bits? I still need the 10 lb resolution.

Yes, with only 0.25% accuracy having 16 bits wouldn't be of much value?. Using the PIC's internal 10 bit A2D would give you 0.1% resolution, and will make your project pretty trivial to make.
 
There are PICs with 12 bit A2Ds. Some of the PIC24 series are under $3.00 in single quantities.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top