XTR117: XTR117 Bypass Capacitor

Hi Reilly,

The capacitor is implemented primarily for noise filtering purposes, but also helps with stability. It does indeed act as an LPF, to reduce unwanted EMI noise that can often be present in industrial environments. The XTR117 datasheet does not discuss this in detail but it mentioned in the datasheet of the related part XTR106, on pages 9 and 14. Different part, but same principle. The relevant excerpts from page 9 is below -

"A 0.01µF to 0.03µF supply bypass capacitor connected between V+ and IO is recommended. For applications where fault and/or overload conditions might saturate the inputs, a 0.03µF capacitor is recommended."

The next excerpt is from the "RADIO FREQUENCY INTERFERENCE" section, not all of it is applicable to the XTR117 so I've crossed some of it out below (placing a cap directly between IIN and IRET on the XTR117 will almost certainly cause stability issues).

"The long wire lengths of current loops invite radio frequency interference. RF can be rectified by the sensitive input circuitry of the XTR106 causing errors. This generally appears as an unstable output current that varies with the position of loop supply or input wiring. If the bridge sensor is remotely located, the interference may enter at the input terminals. For integrated transmitter assemblies with short connection to the sensor, the interference more likely comes from the current loop connections. Bypass capacitors on the input reduce or eliminate this input interference. Connect these bypass capacitors to the IRET terminal as shown in Figure 6. Although the dc voltage at the IRET terminal is not equal to 0V (at the loop supply, VPS) this circuit point can be considered the transmitter’s “ground.” The 0.01µF capacitor connected between V+ and IO may help minimize output interference."

The capacitor also plays a role to stabilize the part, affecting the stability/AC response of the transmitter overall. Without it, you might observe settling issues with IRET and thus get unexpected behavior from IO. It would be TI's recommendation to keep the 10nF cap present if possible. Are you experiencing issues with the LPF caused by the cap cutting off your desired input signal? You could experiment with decreasing its size if necessary.

Cheers,

Jon