Geometric magnification M in microfocus radiography is the ratio of the focus-to-detector distance (FDD) to the focus-to-object distance (FOD), expressed as M = FDD / FOD, with magnification always greater than 1 because the X-ray cone is divergent from a near-point source [S2][S4][S5].
Choosing M is the central engineering decision in any microfocus inspection: M governs the pixel-per-millimetre scale on the detector, the effective unsharpness, and therefore the smallest flaw that can be resolved, while the focal-spot size F (from under 1 micron to 200 microns in commercial tubes) caps how far M can be pushed before the image blurs [S5].
The Core Formula and Three Distance Variables
The magnification relation M = FDD / FOD is the same expression written three ways across the literature: SID/SOD, FDD/FOD, and (a+b)/a, where a is the focus-to-object distance and b is the object-to-detector distance, so FDD = a + b [S2][S3][S4]. Source-to-image distance (SID) and source-to-detector distance (SDD) refer to the same physical separation, and the similar-triangles derivation is identical: image size divided by object size equals SID divided by SOD [S3].
Two corollaries follow directly. First, halving FOD at fixed FDD roughly doubles M, which is why operators of microfocus systems physically lift the part toward the source on a programmable stage rather than moving the detector back [S5]. Second, the same magnification can be reached by long-FDD/long-FOD or short-FDD/short-FOD pairs, but the resulting geometric unsharpness U = (M-1) x F is unchanged if F and M are matched, so the practical choice is driven by cabinet geometry and X-ray tube power loading rather than resolution [S5].
How Focal Spot Size Sets the Resolution Ceiling
Geometric unsharpness U = (M-1) x F is the half-shadow (penumbra) width cast by a focal spot of size F at magnification M, and it is the dominant resolution-killer in magnified radiography once U exceeds roughly half the detector pixel pitch [S5]. A 200 micron focal spot at M = 5x already produces U = 800 microns of penumbra, which is fine for inspecting a 1 mm solder bump but unacceptable for sub-100 micron features in a BGA or micro-CT voxel work [S5].
Microfocus tube families bracket this trade-off: a 200 micron spot typically delivers 1.5-10x M at production-relevant power for electronics inspection, while submicron-resolution tubes (spots under 1 micron) can be pushed past 50x but require shorter exposure windows because heat is dissipated from a small, stationary transmission target [S5]. Industrial computed-radiography setups that need both magnification and high tube output therefore sit at the upper end of the spot-size range, whereas wafer-level vacuum packaging and micro-CT work stay in the 1-5 micron spot regime.
Practical Magnification Ranges by Application
Electronics solder-joint inspection on PCB assemblies typically runs M between 2x and 10x with 5-50 micron focal spots, balancing throughput against the need to resolve ~25 micron lead-free solder voids [S5]. Die-level and semiconductor packaging inspection moves to 10x-50x with 1-5 micron spots, where the unsharpness budget U must stay below the detector pixel size (often 50-100 microns on flat-panel detectors, or 10-20 microns on a CMOS-based unit) [S5].
Micro-CT and scientific imaging extends further, with magnifications above 50x and focal spots at or below 1 micron; the trade-off is dose and exposure time, since a submicron target cannot dissipate the same power as a 200 micron spot without melting [S5]. The study reports that a 1-mm nominal focus paired with a 250-µm aperture produced ×2 magnification on a computed radiography receptor, improving limiting spatial resolution from 3.55 lp/mm (contact) to 5.6 lp/mm (magnified) [S1].
Apparent vs True Focal Spot, and the Effective Resolution
Effective focal spot size is not the same as nominal rating: it is influenced by electron energy, the scattering volume inside the transmission target, and the target material itself (commonly tungsten on a beryllium or diamond exit window for microfocus tubes), and a 1 micron nominal spot can degrade to several microns of effective size if the target is too thick or the kV is mis-set [S5]. Transmission target thickness therefore directly caps maximum achievable M at a fixed FDD: thinner target, smaller effective spot, higher usable M [S5].
Two setup habits keep the math honest. First, measure (or look up) the effective F at the operating kV, not the marketing number, because M x F is what sets U [S5]. Second, when CT reconstruction or geometric metrology is the goal, calibrate FDD and FOD with a calibrated stage or a reference phantom rather than trusting the software readout, since a 2% distance error at M = 50x becomes a 2% dimensional error in every voxel of the reconstructed volume. The same calibration discipline that governs stage size and optical comparator fit for tight-tolerance parts applies to microfocus stage travel: the smaller the feature, the less forgiveness in the geometric chain.
Microfocus vs Pinhole (Pseudo-Microfocus) Geometry
True microfocus geometry places the focal spot on a thin transmission target inside the tube, with the cone beam limited by external apertures, while pseudo-microfocus or pinhole geometry uses an external aperture downstream of a larger conventional spot to define a small effective source, at the cost of throwing away most of the tube output [S1]. The magnification formula is identical for both: M = FDD / FOD, and the same U = (M-1) x F relation applies with F now equal to the aperture diameter rather than the on-anode spot [S1].
The pseudo-microfocus route buys accessibility (any diagnostic tube can become a magnification source) at the price of tube loading, since the same 1.0 mm spot now has to be operated through a 250 micron aperture, dropping useful output by roughly (250/1000)^2, or about 6.25% of the original flux at the same kV/mA [S1]. For industrial work, the decision is whether the throughput penalty of the pinhole approach is worth avoiding the capital cost of a sealed microfocus tube, and for most production NDT it is not.
Quick Selection Rules for a Working Setup
Set M first from the smallest flaw size you must resolve (Ffeature) and the detector's effective pixel pitch (P): a usable rule is M such that (M-1) x F is less than P, which keeps geometric unsharpness within one pixel [S5]. Pick F from the available tube family, then solve FDD = M x FOD, where FOD must be large enough that the cone beam covers the part and the air-gap attenuation at your kV is acceptable (typical FOD range 5-50 mm for electronics, 50-500 mm for castings) [S5].
For a worked example: a 5 micron focal spot, a 100 micron pixel flat-panel detector, and a target M of 20x gives U = 19 x 5 microns = 95 microns, which is just under the pixel pitch, so the system is at its usable limit; dropping M to 15x gives U = 70 microns and a comfortable 30 micron resolution headroom for fusion 2026-style X-ray diagnostic work and similar high-resolution NDT tasks [S5].
The next signal to watch is microfocus tube power density: transmission target cooling is the practical limit on how high M can be pushed at production throughput, and any vendor datasheet revision that lowers the minimum effective focal spot below 1 micron at useful mA ratings will move the magnification ceiling for both semiconductor and oxy-fuel cutting-adjacent NDT inspection in the same direction.
Spec-level background on the components involved: pressure transmitter, flow meter, and industrial valve.