Shielding effectiveness testing: what a dB figure means and which fixture produced it

What shielding effectiveness is, and its units

A gasket fixture measurement is stated as 20 times the base-ten logarithm of the ratio of the field without the gasket to the field with it, in decibels [S22]. In the power-ratio wording one design guide uses, 10 dB is a tenfold reduction, 20 dB a hundredfold and 30 dB a thousandfold [S39]. Manufacturer tables routinely give five points, a 200 kHz magnetic-field value, 100 MHz and 500 MHz electric-field values and 2 GHz and 10 GHz plane-wave values, with a 40 GHz plane-wave point on newer sheets [S15][S11]. Parker®'s current guide prints the caution under its tables: a value measured in the MIL-DTL-83528 fixture cannot be assumed for a real flange, because tolerances, stiffness, fastener location and size can raise or lower it, and the fixture number is for comparing materials [S15]. The handbook's version is that the only way to know a real enclosure's shielding is to test the enclosure [S9].

Field regions: why the field type rides with the number

An electromagnetic wave carries an electric field and a magnetic field, and their ratio is the wave impedance: close to a source that is mostly current the ratio is small and the field is called magnetic, close to a high-voltage source it is high and the field is called electric, and far from any source it settles to 377 ohms, the plane-wave impedance of free space [S9][S42]. Reflection at a metal surface depends on the mismatch between the wave impedance and the surface impedance of the metal, which sits at milliohms, so a high-impedance electric field is mostly reflected while a low-impedance magnetic field is reflected less and must be absorbed instead [S9][S42]. That is why low-frequency magnetic fields are the hard case for any shield and why datasheets quote the 200 kHz magnetic-field value separately [S9]. The handbook writes the shielding of a solid barrier as a reflection term plus an absorption term plus a re-reflection correction that is practically always negative, only matters when absorption is under about 6 dB, and is a thin-wall effect roughly below 20 kHz; thickness is only decisive at low frequency, and at high frequency even foil is an effective barrier [S9].

Why a seam leaks, and what the two gasket numbers measure

Shield current must cross the joint; if the gasket has lower conductivity than the wall, current decays more slowly in the gasket and more of it reaches the far side, and if an air gap exists at the gasket-to-flange interface the current is diverted to the points that do touch, so a high-resistance joint behaves much like an open seam [S9]. A slot becomes a radiator when its longest dimension approaches half a wavelength; the rule of thumb in the trade press is an aperture no larger than one twentieth of a wavelength, and the handbook gives the attenuation through a slot with depth as approximately 54.5 times the depth divided by the cutoff wavelength, in decibels [S42][S9]. At 10 GHz the free-space wavelength is 30 mm, so a half-wave slot is 15 mm, our arithmetic from the speed of light. Shield currents flow parallel to the electric-field component of the wave, so where the polarization is unknown the joint is designed and tested for the worst case, current parallel to the seam [S9].

Transfer impedance is the second gasket number. Current crossing a seam develops a voltage across it equal to the current times the impedance of the seam, and the seam then radiates like a slot antenna with a field that follows that voltage [S24]. The SAE ARP1705 coaxial method drives a known current through a gasketed joint and measures the voltage drop across the gasket; the result is stated as resistance per unit length or as attenuation in decibels, and lower impedance means better shielding [S21]. Which number a designer should trust is disputed: Bal Seal writes that only the radiated MIL-G-83528 measurement can be used directly in shielding calculations and that there is no consensus translation from transfer impedance, while Kunkel writes that the radiated fixture produces highly inflated levels and that transfer impedance converts to fairly accurate shielding figures [S21][S23]. Both agree on the mechanics and that the transfer impedance method is the more repeatable [S21][S23].

The methods in one table

Method, what it measures, the frequency range each source states, the fixture and the standard's status. Ranges are as published by the source in the last column, not read from any standard's text.

The MIL-DTL-83528 arrangement is still described by manufacturers as a modified MIL-STD-285 setup even though MIL-STD-285 itself was cancelled in 1997 [S9][S21][S7]. Check with IEEE for any replacement of IEEE 299-2006 and 299.1-2013.

The MIL-DTL-83528 fixture in detail

A shielded enclosure has a square opening in one wall; a plate is bolted over the opening with the gasket specimen sealing the interface; a transmitting antenna sits outside and a receiving antenna inside, each one meter from the opening, and the ratio of the signal through the open aperture to the signal through the gasketed plate is the shielding effectiveness [S21]. Bal Seal gives the opening as two feet per edge and Brull as 610 by 610 mm; the handbook says its data come from a MIL-G-83528B test with a 24 by 24 in aperture at about 100 psi on the gasket; Kunkel describes a 26 by 26 in, 3/8 in thick aluminum plate over that opening with the antennas aimed at its center [S21][S22][S9][S23]. Bal Seal notes that the enclosure and opening support resonant modes above roughly 175 MHz, so results are read with resonance in mind [S21]. Brull reports nominal repeatability of plus or minus 6 to 10 dB, degrading to 20 dB with antenna positioning, and a growing offset from real-application values as frequency rises; Kunkel's critique is stronger, with about 30 dB of usable dynamic range and inflated levels that grade gaskets only on a relative scale [S22][S23]. What the fixture does well is yield a dB number that drops straight into a shielding budget, which is why the type minimums of MIL-DTL-83528 are defined in it, and why Parker uses a modified version of it, with its own test-procedure designation, for typical application values [S21][S15][S9].

SAE ARP1705 transfer impedance in detail

A coaxial housing of specified dimensions holds a center conductor, a fixed impedance and a plate in series; the gasket is clamped between the plate and the housing with fasteners; a signal injected at the input flows through the impedance and the gasket and returns through the fixture walls, the current is known from input power and impedance, and the voltage across the gasket is measured [S21]. Spira runs the method on its own fixture with a spectrum analyzer [S24]. Kunkel states that the fixtures resolve gasket conductivity to about plus or minus 2 dB and discriminate between gaskets over more than 120 dB; Brull gives repeatability of plus or minus 3 to 6 dB [S23][S22]. The method is repeatable because the path is rigid cable and connector hardware rather than free space, and it tests the gasket against any structural material and finish, as a function of clamping force, and after moisture or salt fog exposure [S21][S23]. Its designed frequency range is the disputed item in the methods table on this page, because it changed with the fixture revision [S21][S22][S23][S27].

The material tests that ride with the shielding number

A MIL-DTL-83528 datasheet carries more than the fixture value; the rows below are the tests manufacturers name on their sheets, with the paragraph numbers they cite (4.6.x in the MIL-G era, 4.5.x from Revision C onward) [S9][S10][S15]. Parker marks each test (Q) qualification, (C) quality-conformance or (Q/C) both, which tells a buyer which properties are checked on every lot and which only at qualification [S15].

What a test report contains

Read from the fixtures and sheets above, a gasket shielding report states the items below; a report that omits the fixture, the pressure or the field type cannot be compared with another.

1. The method and fixture, with the manufacturer's test-procedure designation where a modified method is used [S9][S15].
2. The aperture size and plate, the gasket cross-section and material, and the compression or pressure applied: about 100 psi in the handbook's data, a minimum 10 percent deflection on the fabric-over-foam sheet, 40 percent in Nolato's example [S9][S16].
3. The field type at each frequency (magnetic, electric, plane wave), the frequency list, the open-aperture reference level and the dynamic range [S9][S21].
4. The result in dB per frequency, with the statement that fixture data do not transfer to a flange [S15].
5. For transfer impedance: the fixture and its revision, the joint materials and finishes, the clamping force, the frequency sweep, the impedance or dB attenuation versus frequency, and any environmental precondition such as moisture or salt fog [S21][S23].
6. For a material conformance report: the sheet's (C) properties with the lot's measured values against the ceilings or minima [S15].

Qualification versus conformance

Qualification is proof, once, that a manufacturer's material meets the whole test group of the specification, after which the material is listed on the qualified products list; Parker states that manufacturers receive material approval for listing on the QPL, and the DLA database showed 32,911 part designations for QPL-83528 when read, last updated 7 May 2026 [S15][S2]. Nine manufacturers held QPL-83528 qualification on the DLA Qualified Products Database on 2026-09-14 [QPD-1]. The database's governing entry read MIL-DTL-83528C with Supplement 1A (2001) while DLA ASSIST lists Revision J (2023) as current; the two records disagree, and ASSIST is the record of the current issue [S2][S1]. Conformance inspection is the per-lot subset, the (C) items, that shows a delivered lot still meets the qualified values [S15]. Qualification is a statement about the material and its maker; it does not certify a specific delivered gasket, and it says nothing about the flange it will meet [S15][S21].

First article inspection under AS9102 documents, on a new part or after a change, that every drawing and specification requirement is understood, verified and recorded, and it is separate from material qualification [S44]. A certificate of conformance is the supplier's signed declaration that the delivered lot meets the order, naming the specification, revision, type letter, sheet and dash number and the material lot, and where the order asks, the conformance values and the manufacturer's QPL status. The practical package for a MIL-DTL-83528 gasket is therefore the manufacturer's datasheet with the type letter and the Q/C marks, the lot conformance values for the (C) items, the certificate tying lot to specification and revision, and a first article report when the part is new to the program. Where a distributor waterjet-cuts sheet to a drawing, the certificate chains to the material manufacturer's lot certification, and the cutting adds dimensional inspection without changing the material qualification. The MIL-DTL-83528 guide explains the type letters and the QPL; the conductive elastomers page carries the shielding figures by frequency, and the design guide the groove that turns a fixture value into a flange value.

Fixture descriptions and ranges on this page come from manufacturer and trade sources, never from a standard's text; standards are cited by title and clause.