XTR117: XTR117: 4-20mA to 0-3.3V converter

Luiz Catalan

Part Number: XTR117
Other Parts Discussed in Thread: TLV333, ESD321, ESDS311, INA219

Tool/software:

Dear E2E Team,

I'm trying to get a 4-20mA receiver circuit with analog output in the 0-3V3 range and I found a suggestion from 4 years ago, in a reply by Luis Chioye to the Ambimat Team.

It essentially comes down to the circuit shown in Luis's final reply and I found it very well suitable for what I want, but I have some questions:

- I intend to work with input voltage between 10 and 28V, what would be the recommended TVS?

- For better precision, wouldn't it be recommended to use resistors with an error better than 1% (or better than 0.1%)? If so, what would these resistors be?

- What error would be introduced with variations in the 3V3 power supply of the TLV333 amplifier?

Thank you very much for any answer!

Luiz

Luis's schematic is this:

over 1 year ago

0 Robert Clifton56 over 1 year ago

TI__Mastermind 31870 points

Hi Luiz,

I'll answer this in more detail next week but for the TVS diode, I'd recommend taking a look at the How to select a Surge Diode document for determining what would be a good fit.

Best Regards,

Robert Clifton

0 Clemens Ladisch over 1 year ago

Guru 327760 points

A current of 20 mA will drop 3.2 V over R_load, and much larger voltages are not allowed at the opamp's inputs. So you should use a 3.3 V diode, e.g., ESD321 or ESDS311.

The R1/R2 network attenuates errors in the 3.3 V supply to about 20 %.

What is the accuracy of the input current signal, and of the ADC? I.e., what accuracy do you actually need? Can you calibrate out all these errors?

0 Luiz Catalan over 1 year ago in reply to Robert Clifton56

Prodigy 20 points

Thank you very much, Robert, I will read the document. Sure will solve my doubt!

Best regards!
Luiz

0 Luiz Catalan over 1 year ago in reply to Clemens Ladisch

Prodigy 20 points

Hi Clemens,

Great to have your recommendation for the right diode!

About R1/R2 attenuating the error by 20%, the supply will certainly oscillate much less than this value.

The input signal has an error of around 1-3% (inherent to the sensor) and yes, eventual correction will be made through the software, as you suggested.

The AD to be used has 12-bit resolution and according to the specified precision will meet the customer's demand.

Thank you very much for the very enlightening answer!

Best Regards,

Luiz

0 Clemens Ladisch over 1 year ago in reply to Luiz Catalan

Guru 327760 points

What I meant to say, if the supply has an error of 100 mV, the output will shift by 20 mV.

0 Luiz Catalan over 1 year ago in reply to Clemens Ladisch

Prodigy 20 points

Hello Clemens, thanks for pointing this out to me, I hadn't noticed it.

Could you please suggest me some alternative to get a stable reading on the AD that would be less prone to errors? What should be done? I liked the simplicity of the circuit proposed by Luis Schoie, but I'm open to ideas and ideally (please, I'm not an expert!), I would use an integrated circuit like the INA219 to read volts and current and would read values from 4-20mA (or with a wider current range, say 0-24mA) without needing any calibration, just through calculations. Using a sufficiently accurate shunt (remembering that the consumption current is quite low), could I assume that the errors would not be greater than 1-2% o even less?

Am I correct or can this idea be discarded?

Once again, many thanks for your suggestions.

Best regards, Luiz

+1 Clemens Ladisch over 1 year ago in reply to Luiz Catalan

Guru 327760 points

If you configure the opamp as a voltage follower, then the output range is from 0.64 V to 3.2 V, i.e., you lose one fifth of the ADC range. But then the error depends only on the R_load accuracy, on the opamp's offset voltage (which is 15 µV, i.e., negligible for the TLV333), and on the ADC.