Eddy-current proximity probes are specified first by tip diameter — 5 mm FFv, 8 mm, 11 mm, and 25 mm are the four mainstream sizes used for radial vibration, axial position, and case-expansion measurement on rotating machinery [S2]. A 5 mm FFv tip is the correct starting point for small-diameter shafts or for retrofits where lateral clearance is physically constrained; an 8 mm or 11 mm tip covers the bulk of API 670 turbine and compressor applications; a 25 mm tip is reserved for large slow-speed machinery where the longer linear range and stronger drive field outweigh the spatial cost [S2][S3].
Probe, cable, and driver (or transmitter) form a matched set and are not mixed across series such as 3300, 3300 XL, or 7200 — extending the cable or substituting a different driver shifts the impedance match and degrades the gap-voltage calibration [S3]. Selecting a proximity probe therefore means locking the whole channel — tip + extension cable + driver — to one calibration curve, then verifying sensitivity with a portable vibration calibrator on the actual target material [S1].
Tip Diameter and Linear Range: Matching Probe to Shaft
Linear range scales with tip diameter, and the 5/8/11/25 mm family directly maps to a typical linear-range envelope that grows with physical size [S2]. For a small-diameter shaft or any application where the engineer needs to limit the lateral field of view — for instance to prevent cross-talk between two probes on a narrow bearing journal — the 5 mm FFv tip is the practical default [S2][S3]. An 8 mm tip is the workhorse for most industrial turbomachinery, balancing tip mass, sensitivity, and the radial clearance inside a standard bearing housing.
Linear range is not a single number; it is a function of the target material, the driver electronics, and the gap voltage setpoint, and vendors publish a calibrated curve per target — AL7075-T6 aluminum, 360 brass, 4140 steel, C1045 steel, 17-4, 304, 316L, and 420 stainless are the common calibration materials [S2]. Specifying a probe without confirming the target material on the calibration chart is the single most common cause of a probe that "looks right" on the bench but reads wrong in the machine. For more on the broader sensor landscape that proximity probes sit inside, see the proximity sensor reference page.
Target Material, Gap Voltage, and Calibration Discipline
Eddy-current probes are not absolute devices; their output is calibrated against a specific target alloy, and re-calibrating against the actual shaft material is a documented requirement before commissioning [S1][S2]. Gap voltage alone is incapable of detecting faults such as incorrect impedance matching, and that is the dominant failure mode in the field — the probe shows a believable mid-range gap voltage, but the sensitivity is off, so vibration alarms trip either too early or, worse, never [S1].
Portable vibration calibrators are the standard tool to validate probe sensitivity in the loop and catch these faults before the machine is released for service [S1]. A complete selection therefore includes a sensitivity-validation step against the real target — a step that is easy to skip on a fast-track shutdown and that engineers should not skip. The same gap-voltage discipline that governs a pressure transmitter loop applies here: the loop as a whole, not the individual sensor, defines the measurement.
Mounting Method: Internal, Through, or Reverse-Mount
Three mounting options dominate, and the choice is dictated by the bearing-housing geometry rather than by preference [S2]. Internal mounting places the probe entirely inside the machine or bearing housing on brackets such as the PRO Line DM902 series; it gives the most precise shaft-position and vibration reading and minimises installation cost on new builds [S2]. Through mounting, also called internal/external, uses an adapter bushing such as the PRO Line DM901 or DM903 and counter-bores the casing so the probe tip sits inside the bearing housing while the body is externally accessible for service [S2].
Reverse-mount housings (for example the PRO Line DM91000) keep the probe external and are used only when no internal option is feasible; the trade-off is that exposed shaft areas can scuff, scratch, or rust, which degrades the measurement [S2]. The mechanical environment — not the electronics — usually picks the mounting. A practical tip: in reverse-mount applications, surface finish, mechanical guarding of the probe, and cable routing deserve the same engineering attention as the probe itself.
Driver Outputs, Enclosures, and Hazardous-Area Ratings
Drivers come in two output flavours: a BNC isolated output for direct connection to a vibration monitor or analyser, and a 4-20 mA or voltage terminal-block output for direct tie-in to a DCS or PLC [S2]. For API 670 vibration channels, the BNC path is conventional; for a plant that wants a single twisted pair back to the control room, a 4-20 mA proximity transmitter is a valid alternative, but it loses raw waveform data and should be specified consciously. Intrinsically safe driver options exist for hazardous-area duty, and AISI 316L stainless-steel probe cases with IP68 sealing are the mainstream ruggedised spec [S2].
Driver calibration options spanning aluminum, brass, several carbon and stainless steels cover the majority of turbomachinery shafts, but exotic alloys — Inconel, Monel, titanium — require vendor confirmation that the published linear range and sensitivity still hold [S2]. Where the loop ends in a flow computer or a flow meter skid, a 4-20 mA proximity transmitter can be wired in parallel for machine health without disturbing the primary measurement.
Selection Criteria Comparison: Tip Sizes Side by Side
Lining the four mainstream tip sizes up against the criteria that actually drive a purchase order makes the trade-off visible. The 5 mm FFv tip wins on small-shaft fit and minimal lateral view, loses on linear range; the 8 mm tip is the balanced default; the 11 mm tip gains linear range and signal-to-noise at the cost of lateral clearance; the 25 mm tip maximises linear range for large, slow-speed shafts but cannot fit a standard bearing housing. All four are available in AISI 316L stainless with IP68 sealing and in intrinsically safe driver variants [S2].
Material calibration coverage is a separate axis: vendors publish full curves for AL7075-T6, 360 brass, 4140, C1045, 17-4, 304, 316L, and 420 stainless [S2]. Warranty terms differ sharply between premium lines (lifetime on materials and workmanship) and mainstream lines (three-year limited), and that gap is a useful proxy for case construction, sealing, and driver MTBF [S2]. A useful adjacent reference is the inductive sensor selection map, which covers the lower-frequency cousin of the eddy-current probe and clarifies where inductive and eddy-current technologies diverge.
Who Should Not Pick the Default 8 mm Tip
The 8 mm workhorse is the wrong choice in three common cases. First, on a small-diameter shaft where the tip diameter itself constrains the lateral field of view, the 5 mm FFv probe is the correct pick and prevents cross-talk between adjacent probes [S2][S3]. Second, on large slow-speed machinery where the linear range of an 8 mm tip is exhausted before the full thermal growth of the shaft is captured, an 11 mm or 25 mm tip is mandatory. Third, on a reverse-mount installation where the exposed shaft can scuff or rust, no probe diameter compensates for the lost surface quality — the engineer should redesign the mounting before picking a different probe [S2].
Engineers specifying replacement probes in a hurry should also avoid mixing series — 3300, 3300 XL, and 7200 cables and drivers are not interchangeable, even when the tips look identical [S3]. The matched-set rule is non-negotiable. For a complementary view on the analyser side of the same measurement chain, the vibration analyzer buying guide covers the downstream signal-conditioning and FFT math that turn a raw proximity-probe voltage into a usable spectrum.
Limits, Failure Modes, and Pre-Commissioning Checks
Two failure modes dominate the field. The first is incorrect impedance matching, which produces false alarms or, in the worst case, missed trips on a critical machine — gap voltage alone will not catch it, only a sensitivity check with a portable calibrator will [S1]. The second is a target-material mismatch, where a probe calibrated for 4140 steel is bolted up against a 17-4 stainless shaft and the published linear range no longer applies [S2]. Both are caught by a documented pre-commissioning loop check.
Reverse-mount installations add a third failure mode: surface degradation of the exposed shaft, which gradually biases the reading and is invisible on a single gap-voltage check [S2]. Periodic visual inspection of the shaft, not just the probe, is part of the maintenance plan. The full pressure sensor and industrial valve loops in the same skid share the same sensitivity-validation discipline and the same trap of trusting a single "looks healthy" reading.
Track these signals over the next procurement cycle: vendor disclosures of new target-material calibration curves (especially for Inconel and titanium shafts), updates to IP68 and IS certifications on driver enclosures, and any movement on the API 670 guidance for proximity-probe channel testing. A shortlist of two or three vendors — one premium line, one mainstream line, one reverse-mount specialist — gives engineering the negotiating room to lock the channel spec without locking the plant into a single supplier.