A magnetostrictive level transmitter's waveguide is a ferromagnetic wire that runs the full length of the measurement range and, on rigid or flexible insertion probes, sits inside a sealed stainless steel tube that is fully wetted by the process fluid [S5][S8].
The waveguide is not a contactless sensing element, and the contact-or-not answer depends entirely on the mechanical configuration: rigid insertion, flexible insertion, or externally clamped to a magnetic level indicator (MLI) chamber [S3][S4][S6].
What "waveguide" actually is and where it sits
The waveguide is a magnetostrictive wire that conducts a torsional strain pulse back to a pickup head, and it is physically housed inside the probe tube together with the interrogation electronics and return-pulse sensor [S1][S2]. ABB's LMT-series white paper describes the waveguide as a "ferromagnetic sensing element usually referred to as a waveguide" that "serves as a conductor of the torsional wave to the pulse converter" [S2]. The interaction of two magnetic fields, one from a current pulse on the waveguide and one from a permanent-magnet float outside the tube, generates the strain pulse that is timed to determine float position [S1][S2][S5]. Because the waveguide is enclosed inside the probe tube, anything that contacts the probe tube also contacts the waveguide assembly, so process-side wetted parts are limited to the tube, end plug, float, and process connection [S3][S5].
Two architectures, two different contact answers
Insertion-type units, both rigid and flexible, place the waveguide directly into the process stream: the Jogler ILT-6000, for example, is "designed for direct insertion into process media such as sumps, tanks, bridles, and stilling wells," with a low-profile waveguide that enters the tank through a process connection [S8]. Cross Company likewise notes that magnetostrictive probes "can be used alongside a magnetic level gauge chamber or even inserted directly into the process media," giving the same conclusion in different wording [S6].
Externally mounted units, by contrast, clamp onto the outside of an MLI chamber and read the magnetic field of the chamber's internal float; SOR's April 2026 technical article states explicitly that the transmitter "mounts externally to a magnetic level indicator (MLI) chamber and reads the magnetic field of the float" with "no vessel penetrations required. No process contact" [S3]. In that architecture the waveguide, electronics, and pickup are all outside the pressure boundary, and the only wetted parts belong to the MLI chamber and its float [S3]. Engineers reading vendor cut sheets must therefore distinguish between "insertion" probe SKUs and "external" or "MLI-coupled" SKUs before assuming any non-contact behaviour [S3][S8].
Why process contact does not hurt accuracy or life
The SenTec product reference highlights two consequences of the contact architecture: the measuring rod and waveguide wire are sealed inside a stainless steel tube, so although the assembly is wetted, there are no moving mechanical parts in contact with the fluid, no friction, and no wear; the converter is "not in contact with the measurement medium" in the sense of moving seals, but the tube wall and waveguide enclosure are fully immersed [S5]. The same article notes that displacement is determined by measuring the time of the start pulse and the end pulse, which is why these devices are routinely quoted at sub-millimetre resolution [S2][S5]. The waveguide sensing principle is therefore classified as a "contact-type" level technology in the Sino-Inst April 2026 comparison guide, alongside capacitance and differential-pressure probes, and is contrasted with non-contact radar [S4].
For a working perspective on how contact and non-contact level technologies trade off against each other in real spec sheets, see the engineering overview of magnetostrictive level transmitters, the broader level transmitter taxonomy, and the competing capacitance level transmitter probe family, all of which sit on the wetted side of the pressure boundary.
Selection rules: when contact is fine and when it is not
For clean liquids, water, condensate, fuels, and most chemicals compatible with 316L stainless steel, insertion-style magnetostrictive probes are widely used because the wetted materials are limited to a small set of alloys, the float is the only moving part, and there are no seals sliding against the waveguide [S5][S6][S8]. Cross Company lists direct insertion into sumps, tanks, bridles, and stilling wells as the standard mounting, which is consistent with the ILT-6000 datasheet approach [S6][S8]. For process fluids that foul, coat, crystallize, or attack stainless, the contact architecture becomes a maintenance liability, and the same vendors therefore offer the external MLI-clamp variant, in which the waveguide never sees the process [S3][S6]. Hazardous, toxic, or high-pressure service is another case where users prefer the external mount, because the process stays fully contained inside the MLI chamber and operators read level without exposure [S3].
Comparison against the main competing level technologies
The Sino-Inst April 2026 selection guide lines up the principal contact and non-contact level technologies across four decision criteria: physical contact, typical accuracy, susceptibility to foam or vapor, and typical service [S4]. On contact, magnetostrictive and capacitance probes are both wetted, while radar and ultrasonic are non-contact; on accuracy, magnetostrictive leads the group at the millimetre level; on foam and vapor immunity, non-contact radar and ultrasonic hold the advantage; on service, magnetostrictive is preferred for interface measurement, interface and total level in a single probe, and high-resolution inventory [S4]. Engineers weighing these trade-offs should also note that magnetostrictive transmitters share the waveguide principle with a class of proximity probes used for shaft position and vibration, where the same time-of-flight torsional-wave physics applies on a much shorter stroke.
Limits, failure modes, and what to verify on the datasheet
Three failure modes follow directly from the contact architecture. First, coating or buildup on the outside of the probe tube can magnetically decouple the float from the waveguide, producing a sticky or lost reading; in heavy fouling service, specify a stilling well or external-mount configuration [S3][S6]. Second, internal failure of the waveguide wire or its termination, often from vibration, mechanical fatigue on a flexible probe, or lightning-induced surge on the current pulse line, shows up as a total loss of return pulse, which is why surge protection and rigid probe selection matter in tanks with agitation [S1][S2]. Third, material compatibility of the float, tube, and end plug with the specific process fluid, including NACE MR0175 compliance for sour service where applicable, must be checked on the vendor datasheet, because the waveguide enclosure inherits the wetted material's corrosion envelope even though the wire itself is sealed inside [S5][S8].
Before specifying, confirm three items on the cut sheet: insertion versus external-mount SKU, full wetted-material list including float magnet encapsulation, and any minimum/maximum process temperature, pressure, and density window the float must track [S3][S4][S8]. For broader process context on how probe-style contact devices compare with non-contact radar on accuracy and lead time, the engineering reference on displacement transducer accuracy classes is a useful adjacent read.