A smart transmitter has two places where its calibration can be off, and confusing them wastes a lot of bench time. The sensor side can misread the actual process, and the analog output side can put out a milliamp signal that does not match what the device thinks it is reading. Sensor trim fixes the first, output trim fixes the second, and they are genuinely independent adjustments that correct different errors. To make matters more confusing, neither of them is the same as re-ranging, which does not correct any error at all. This page separates the three cleanly, explains the order to do the trims in, and shows how mixing them up produces errors that stubbornly refuse to go away.
Sensor Trim vs Output Trim in one line: Sensor trim and output trim are the two independent calibration adjustments in a smart transmitter. Sensor trim aligns the transmitter's internal digital reading of the process variable to a known applied reference, correcting an error between what the process actually is and what the device digitally reports. Output trim, also called DAC trim, aligns the 4 to 20 mA analog output to that digital reading, correcting an error between the device's digital value and the current it actually drives. Re-ranging changes which values map to 4 and 20 mA and corrects no error, so it is not a substitute for either trim.
It helps to picture a smart transmitter as a chain with two conversion stages. First, the sensor and its electronics turn the real physical process, a pressure or a temperature, into an internal digital number, the device's own reading of the process variable. Second, a digital-to-analog converter turns that digital number into the 4 to 20 mA analog current that travels down the loop. Each of these stages can drift or be slightly off, and each has its own trim, because an error in one stage is a completely different problem from an error in the other, and correcting one does nothing for the other.
Sensor trim addresses the first stage. If you apply a known, accurate reference to the transmitter, a precise pressure from a deadweight tester or a precise temperature from a bath, and the device's digital reading of that input does not match the reference, the sensor conversion is off and needs a sensor trim. The trim tells the transmitter what the true value is at one or more points so it corrects its internal reading to agree with the reference. Sensor trim usually offers a lower and an upper point, letting you correct both the zero end and the span end of the sensor's digital reading against applied references. After a good sensor trim, the device's digital reading of the process variable matches reality.
Output trim, sometimes called DAC trim or 4 to 20 mA trim, addresses the second stage. Here you are not concerned with the process at all; you are checking whether the analog output current matches what the device thinks it is outputting. You command the transmitter to drive its output to the 4 mA and 20 mA endpoints and measure the actual current with an accurate meter, and if the measured current does not match, the output trim corrects the digital-to-analog conversion so the loop current agrees with the device's digital value. After a good output trim, the milliamp signal on the loop faithfully represents the device's internal reading, independent of whatever the sensor is doing.
Because the two trims correct different stages, the order matters, and the sensible sequence follows the signal chain. You generally do the output trim first so the analog output is faithful to the digital reading, then the sensor trim so the digital reading is faithful to the process, or you verify each stage independently. The logic is that you want a known-good output path before you judge anything by the loop current, and you want the sensor reading correct against a reference standard. If you try to correct a sensor error by fiddling with the output, or an output error by fiddling with the sensor, you will chase the error around the two stages and never cleanly resolve it, because you are adjusting the wrong conversion. Doing each trim against the right reference, an applied process reference for the sensor and a measured current for the output, keeps the two corrections clean and independent.
The crucial thing that neither trim does is change the range. Re-ranging is setting which process values correspond to 4 mA and 20 mA, for instance telling the transmitter that 0 to 100 units of pressure should map across the output. Re-ranging does not touch either conversion's accuracy; it just redefines the scaling. A transmitter can be re-ranged with a keypad entry and no reference standard at all, precisely because it is not correcting an error, it is choosing a mapping. This is the difference people most often blur: re-ranging changes what the numbers mean, while trimming corrects whether the device reads and outputs those numbers correctly. A newly re-ranged transmitter with a bad sensor calibration is still inaccurate, because re-ranging did nothing to fix the underlying reading.
This distinction has real consequences on the bench. If a transmitter reads a genuine, known input incorrectly, the fix is a sensor trim against a reference, not a re-range, and re-ranging it to make the number look right at one point just hides the error and leaves it wrong everywhere else. If the device reads correctly on its display but the control system sees a different value, the problem is in the output path or the loop wiring, and an output trim, not a sensor trim, is the tool. Knowing which symptom points to which action is the whole practical value of understanding that sensor trim, output trim, and re-ranging are three separate operations addressing three separate things.
Confusing the two trims produces errors that persist no matter how many times you adjust the device, because you keep correcting the stage that is not broken. Suppose the sensor conversion is reading a little high but the technician, seeing the loop current is off, keeps applying output trims. Each output trim makes the current match the digital reading, but the digital reading is still wrong, so the process value the control system sees stays wrong, and the technician is baffled because the output trim seems to keep drifting. The reverse happens too: sensor-trimming a device whose real problem is a bad output path gives a device that reads correctly on its display but still sends the wrong current to the loop. The signature of a trim mix-up is an error that will not stay fixed, because the tool being used never touches the stage that is actually off.
The way out is to separate the two stages during verification. Read the device's own digital value, through a communicator or the display, and compare it against the applied reference to judge the sensor; separately, measure the actual loop current and compare it against what the device reports it is outputting to judge the output path. If the digital reading is wrong against the reference, that is a sensor problem calling for a sensor trim. If the digital reading is right but the current is wrong, that is an output problem calling for an output trim. If both agree with each other and with reality but the control system still shows something different, the problem has moved past the transmitter into the loop wiring, a barrier, or the input card. Bracketing the error to a stage is what turns a frustrating chase into a directed fix.
This stage-by-stage thinking pays off directly when a plant runs its instruments under SCADA or a cloud monitoring layer, because the monitored value is the far end of exactly this chain. A transmitter can be locally correct on its display yet report a wrong value to the platform because of an output-path or loop error, or it can be reporting a value that the platform faithfully shows but that is wrong because the sensor was never trimmed against a reference. When a point looks suspect in a platform such as Merobix, understanding sensor trim versus output trim tells the technician where to look: whether the device's own reading is off, whether the analog output is off, or whether the discrepancy lives between the transmitter and the system. That structured diagnosis, sensor stage, output stage, then loop, keeps a monitored measurement honest end to end rather than leaving a persistent offset that repeated, misdirected trimming never resolves.
Sensor trim aligns the transmitter's internal digital reading of the process variable to a known applied reference, correcting an error between what the process actually is and what the device reports digitally. Output trim, or DAC trim, aligns the 4 to 20 mA analog output to that digital reading, correcting an error between the device's digital value and the current it actually drives on the loop. They are independent because they correct two different conversion stages, and fixing one does nothing for the other.
No. Re-ranging sets which process values correspond to 4 mA and 20 mA, redefining the scaling, and it corrects no error at all, which is why it can be done from a keypad with no reference standard. Trimming corrects whether the device reads and outputs values accurately. A transmitter that is re-ranged but has a bad sensor calibration is still inaccurate, because re-ranging changed only what the numbers mean, not whether the device reads them correctly.
A persistent error usually means you are trimming the wrong stage. If the sensor reading is off but you keep applying output trims, the current matches the digital reading each time but the reading itself stays wrong, so the process value never comes right. The fix is to separate the stages: compare the device's digital reading against an applied reference to judge the sensor, and separately compare the measured loop current against what the device reports to judge the output. Correcting the stage that is actually off makes the fix stick.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.