ICT Fixture Probe Pressure: Spring Force, Wear and Contact Resistance
An in circuit test result is only as good as the electrical contact that produced it. Probe pressure, tip geometry and fixture flatness together decide whether the measurement describes the board or the interface, and a marginal contact is the most common cause of a test that fails a good panel.
What Probe Pressure Has to Achieve at the Test Point
A spring probe has to accomplish three things at once: penetrate the surface film on the test pad, maintain a stable contact area for the duration of the measurement, and do so without damaging the pad or the solder beneath it, a requirement written into most solder joint acceptance criteria. Penetration is achieved by force, and the force has to exceed the yield strength of the oxide or residue layer. On a clean tin or tin lead pad that takes very little force; on a pad that has been coated with a thin conformal film or that carries flux residue, the tip has to break through before any current can flow.
The measured contact resistance is the sum of the probe’s own resistance, the constriction resistance at the tip, and the film resistance at the pad. Only the third term is affected by pressure in a meaningful way, and it falls steeply as force rises until the film is fully displaced. Above that point, additional force buys almost nothing and simply accelerates probe wear and pad damage. The useful operating range is therefore narrow, and it is defined by force rather than by travel.
Spring Force, Travel and the Working Range
A typical ICT probe has a rated spring force of 2 to 4 newtons at the recommended working travel, which is usually two thirds of the maximum travel. A probe with 6.4 mm of total travel is normally operated at 4.2 mm, and the force at that point is the figure quoted on the data sheet. Running the probe at its maximum travel gives more force but leaves no margin for board thickness variation and shortens spring life by compressing the spring into its solid region.
Board thickness tolerance and fixture stack up together decide how much of that travel is actually used. A panel that measures 1.55 mm instead of 1.60 mm, plus a fixture plate that is 0.05 mm low in one corner, removes 0.10 mm of travel from every probe in that region. Compute the worst case before the fixture is built, and choose probes whose working force band covers the whole stack up tolerance. A fixture that is correct at nominal thickness but marginal at the extremes will produce intermittent failures that move from board to board.
Tip Geometry for Different Test Points
Tip selection is driven by the feature being contacted. A crown tip with four or eight points concentrates force and cuts through oxide, and it is the default for round pads and vias. A spear or chisel tip is used where the pad is small enough that a crown would overhang, and a flat tip is reserved for gold plated castellated features where marking is not acceptable. Matching the tip to the pad also sets the cleaning interval, because a crown tip picks up solder faster than a flat one.
Incorrect tip geometry is a frequent cause of a test that passes at the fixture qualification and fails in production. A crown tip that is too large for the pad contacts the solder mask on both sides, so the effective contact area is limited to the ridge and the measured resistance rises. A tip that is too small for the pad concentrates force to the point of deforming the copper, and the deformation is permanent. Measure the actual impression left on a rejected board before deciding that the electrical problem is in the assembly, and confirm it with a microsection when the impression depth is in question.
Contact Resistance Limits and How to Measure Them
Set an explicit limit for contact resistance and measure it as part of fixture qualification. For a standard spring probe on a tin finished pad, less than 100 milliohms is a reasonable target, and anything above 500 milliohms should be treated as a marginal contact rather than as a board defect. The limit is a property of the fixture and the surface finish, so it should be established on a known good board before the fixture is released.
Measurement is best done with a four wire arrangement on a test coupon in the fixture itself. Drive a known current through two probes and measure the voltage across two others, so that the probe and wiring resistance is excluded and only the interface is measured. Repeat the measurement at several points across the fixture, and repeat it after 10,000 cycles. A contact resistance that has doubled from its initial value indicates contamination or wear, even if the absolute figure is still below the limit.
Probe Wear, Contamination and Cleaning
Probe wear shows up as a flattened or blunted tip, a shortened travel, or a spring force that has dropped below the working band. The most common mechanism is not mechanical abrasion but accumulation: solder, flux residue and conformal coating build up on the tip and around the barrel, and the build up both increases resistance and restricts travel. A probe that has 4.2 mm of travel when new may have only 3.5 mm after 50,000 cycles because the barrel is packed with debris.
Cleaning should be scheduled by cycle count rather than by symptom. A dry cleaning with a soft brush at the end of each shift, and a deeper clean in isopropyl alcohol at defined intervals, keeps the barrels free. Never lubricate a probe barrel with oil or grease; the lubricant attracts dust and becomes an insulator. When a probe’s force measures outside its band, replace it rather than adjusting the fixture, because the spring has taken a set and will not recover.
Fixture Mechanics: Plate Flatness, Vacuum and Alignment
The probe plate, the board support and the vacuum seal form one mechanical system, and any one of them being out of tolerance changes the force at the probe tips. Plate flatness should be within 0.05 mm across the working area, and the board support should be stiff enough that the panel does not bow under the applied force. A support that flexes by 0.1 mm at the centre reduces the travel at the centre probes and increases it at the edges.
Vacuum fixtures add a further variable. With insufficient vacuum the board lifts on the side away from the seal, and the probes there see reduced travel and force. Check vacuum with a gauge at the fixture rather than at the pump, and inspect the seal for compression set at each maintenance interval. Where the fixture is located by the board rather than by tooling pins, the fiducials set the position and any wear in the locating pins shifts every probe at once. Where the fixture uses clamps instead of vacuum, verify that the clamping force is applied symmetrically, because an offset clamp will tilt the panel and produce a force gradient across it.
Programming Around Marginal Probes
Test programs are sometimes written to tolerate a marginal probe by widening a tolerance or by retrying a failing measurement. That converts a mechanical fault into a permanent reduction in test coverage. If a measurement fails intermittently and the probe’s contact resistance is above half the limit, fix the probe rather than the program. A retry that succeeds on the second attempt is evidence that the first attempt was made through a poor contact.
The same applies to guard measurements and to any test that depends on a shared reference. A reference probe with elevated resistance shifts every measurement that uses it, and the shift appears as a small, consistent error rather than as an obvious failure. Include the reference and guard probes in the routine contact resistance check, and verify them against a known resistor at the start of each shift so that drift is detected before it reaches the boards.
Receptacle and Wiring Faults That Look Like Probe Faults
Before a probe is replaced, check the receptacle it sits in. A receptacle that has lost its retention allows the probe to seat deeper than intended, which increases travel and force at that point and reduces it at the tip. A receptacle contaminated with flux, or one that has been soldered with a poor fillet, adds resistance that is indistinguishable from a worn tip when measured at the fixture wiring.
Wiring faults follow the same logic. A wire that has broken inside its insulation, or a crimp that has loosened, produces a high resistance that changes with temperature and with the position of the cable bundle. Measure from the probe tip to the connector pin rather than across the whole harness, and flex the cable while watching the reading. An intermittent fault that moves when the harness is moved belongs to the harness, not to the probe.
Fixture Maintenance Schedule and Records
A fixture maintenance schedule should state the interval, the tasks and the acceptance criteria for each task. A workable baseline is a daily visual check of tip condition, a monthly contact resistance measurement across a sample of probes, a quarterly vacuum and seal inspection, and an annual full fixture requalification with a known good board and a full set of measurements across all probes.
Record the probe type, the installation date, the cycle count and the measured force for every probe position. When a batch of failures is traced to one region of the fixture, the record shows whether that region has been replaced recently or has been running since the fixture was built. Fixture quality data used this way turns a recurring test problem into a scheduled maintenance item, which is the only way the intermittent failures ever stop.
FAQ
What probe force should an ICT fixture use? Most spring probes are rated at 2 to 4 newtons at two thirds of maximum travel. Operate within that band, and confirm with a worst case stack up calculation that the whole board thickness tolerance still lands inside it. More force accelerates wear without improving contact.
What contact resistance is acceptable? Below 100 milliohms is a reasonable target on a tin finished pad, and above 500 milliohms should be treated as a marginal contact. Measure with a four wire arrangement so that probe and wiring resistance are excluded, and compare against the value recorded when the fixture was new.
Why does a probe fail intermittently on some boards but not others? Intermittent failure usually points to travel rather than to the probe itself. Board thickness variation, fixture plate flatness and support stiffness all change the travel at individual probes, so measure the impression left on the pad and check the stack up before replacing the probe.
