LMP7721 Electrometer Amplifier: 7 Checks for Guarding and Femtoamp Error Budgets
LMP7721 Electrometer Amplifier: 7 Checks for Guarding and Femtoamp Error Budgets
Plan LMP7721 high-impedance measurements with temperature-qualified bias limits, driven guards, leakage control and compensated feedback. This is a documentation-based planning guide, not a report of independent hands-on testing.
1. Interpret the femtoamp specification with its conditions
The LMP7721 has a typical three-femtoamp input bias current, but a typical value is not a complete worst-case measurement budget. TI specifies limits of plus or minus twenty femtoamps at 25 degrees Celsius, 900 femtoamps through the stated 85-degree range, and five picoamps through the stated 125-degree range, with the documented common-mode and supply conditions. Use the relevant table for the intended environment. A sensor interface qualified at room temperature cannot automatically retain the same current-error bound after warming. Manufacturer documentation.
2. Translate tiny currents into the actual voltage error
Source resistance determines how consequential input and leakage currents become. For an illustrative one-gigohm source, twenty femtoamps produces twenty microvolts, while 900 femtoamps produces 0.9 millivolt. Those products describe one error contribution, not the complete circuit accuracy. Add amplifier offset, source imbalance, resistor tolerance, noise and external leakage. State the required input-referred uncertainty before selecting gain or display digits. An impressive bias-current number cannot compensate for an uncharacterized cable, contaminated board or unsuitable protection component connected to the same node. Manufacturer documentation.
3. Use the special pinout rather than a generic footprint
The package separates sensitive inputs from supply and output pins using isolation positions. TI allows the two internally unconnected isolation pins to be tied to a suitable guard trace. This is not a standard eight-pin amplifier pinout that can be inferred from another device. Verify the top-view drawing, land pattern and every connection before assembly. Keep input routing short and physically separated from noisy outputs and supply transitions. A correct schematic symbol does not prevent leakage caused by a footprint or assembly assumption. Manufacturer documentation.
4. Drive the guard at the protected node potential
A guard is a low-impedance conductor surrounding a sensitive node and held near that node potential, minimizing voltage across leakage paths. It is not automatically a grounded shield. The appropriate guard source depends on whether the circuit is a buffer, inverting amplifier or transimpedance configuration. TI recommends a buffer when guarding extends beyond the immediate device area, to avoid loading the sensitive feedback node with extra capacitance. Review guard-driver stability and current demand instead of connecting a long cable shield directly to a delicate summing node. Manufacturer documentation.
Newest in the store
5. Control board, cable and fixture leakage
Surface contamination, moisture, connectors, sockets and test instruments can introduce currents much larger than the amplifier input bias. Choose construction and cleaning methods suitable for the intended impedance, then allow the board to dry and thermally settle. Include the sensor cable and protection network in the leakage budget. A solderless breadboard demonstration is not a credible femtoamp qualification fixture. Document measurement loading and any guarding arrangement, because connecting a conventional probe can change the behavior being measured rather than simply reveal it. Manufacturer documentation.
6. Check common-mode range and feedback compensation
The input range includes the negative rail, but rail-to-rail output terminology does not imply rail-to-rail input operation. Consult the supply-specific common-mode and output-swing limits. Large feedback resistances interact with sensor, amplifier and stray capacitance, potentially causing peaking or oscillation; use the documented compensation approach for the actual topology. Back-to-back input protection diodes can conduct during excessive differential input separation, so do not repurpose the amplifier as an open-loop comparator. High impedance and wide bandwidth require deliberate loop-stability design. Manufacturer documentation.
7. Retain a guarded measurement and uncertainty record
Begin with bounded low-voltage sources, current-limited supplies and an appropriately constructed guarded fixture. Record temperature, source impedance, gain network, compensation, cable, shielding and measurement bandwidth. Distinguish zero-input drift, settling, noise and response to a known stimulus rather than collapsing them into one offset reading. Keep calculated expectations separate from actual observations. The final evidence should support a stated uncertainty under known conditions, not claim femtoamp accuracy solely because the selected amplifier has an exceptionally small typical bias current. Manufacturer documentation.
Related reading: OPA387 Zero-Drift Op Amp: 7 Checks for Microvolt Errors and Switching Artifacts.
Catalog note: The exact-part search did not return a matching Epro listing during preparation. Check the current catalog; this guide makes no price, availability or included-accessory claim.