Your incoming inspection passed it. Your AOI passed it. Your functional test passed it. Your burn-in passed it.
It shipped. And three months later, a customer in Munich is on the phone because it failed in a way your test team has never seen before — intermittent, temperature-dependent, and impossible to reproduce in the lab.
That failure was always there. It was a latent defect — a hairline solder crack too small for AOI to catch, a component with a marginal bond wire that functional test never stressed, a press-fit connector seated 0.2mm shy of full engagement. None of those defects caused failure at room temperature on a bench. They caused failure under the mechanical and thermal loads of real deployment, after enough time had passed to accumulate the fatigue damage that turned a marginal joint into an open circuit.
HASS — Highly Accelerated Stress Screening — is the production test designed to find those defects before they ship.
---
The relationship between HALT and HASS
HASS cannot exist without HALT. That dependency is absolute and worth understanding before anything else.
HALT testing, covered in the previous post, characterises a product's fundamental stress limits during development — the temperatures and vibration levels at which it begins to malfunction and eventually fails permanently. Those limits define two boundaries: the Operating Limit (the stress level beyond which the product malfunctions but recovers) and the Destruct Limit (the stress level beyond which damage is permanent).
HASS uses those limits to construct a production screen. The HASS profile applies stresses between the operating and destruct limits — severe enough to precipitate latent defects into detectable failures, but not severe enough to damage a good unit. That corridor between "hard enough to find real defects" and "gentle enough not to create new ones" is called the Screen Strength, and finding it is the engineering work that makes a HASS program valid.
A HASS screen applied without prior HALT data is guesswork. You don't know where the limits are, so you don't know whether your screen is strong enough to find anything or aggressive enough to damage good product. Either outcome is expensive. The screen that's too gentle ships latent defects. The screen that's too harsh creates defects in product that was fine before it went in.
---
What the HASS screen looks like
A typical HASS profile runs for 30 to 90 minutes per unit — fast enough to be economically viable in production, aggressive enough to precipitate real defects. It combines two simultaneous stresses.
Thermal cycling at ramp rates of 40–60°C/min — far faster than conventional burn-in, which typically ramps at 1–5°C/min. The rapid temperature transitions apply thermal fatigue stress to solder joints, bond wires, and interface materials in a way that slow burn-in cannot. A hairline solder crack that survives 24 hours of burn-in at 70°C will often propagate to an open circuit within two or three fast thermal cycles.
Six-degree-of-freedom vibration at 10–20 Grms — well above real-world vibration environments, but below the vibration destruct limit established in HALT. The broadband random vibration stresses every mechanical interface simultaneously: solder joints, connector seating, press-fit contacts, fastener torque retention, and component-to-board adhesion. A connector seated fractionally short of full engagement will rattle loose under six-DOF vibration in minutes. In the field, it would have taken months of road vibration.
Both stresses are applied simultaneously, with the product powered and under functional monitoring. Every unit is tested electrically throughout the screen — not just before and after. A joint that fails during a thermal ramp and recovers when temperature stabilises will show up as an intermittent failure in continuous monitoring. After-screen-only electrical testing misses every intermittent defect, which are precisely the defects that cause field failures.
Thermotron's HALT/HASS technical resources and ESPEC's reliability engineering documentation both detail the hardware requirements for combined-stress HASS — specifically the pneumatic vibration table integrated into the chamber floor that makes simultaneous thermal and vibration screening possible.
---
HASS vs. burn-in: the comparison every production engineer should make
Burn-in has been the dominant production screen for electronics since the 1970s. It's simple: power the product, hold it at elevated temperature (typically 70–85°C) for 24–48 hours, and check it electrically at the end. The rationale is that early-life failures — the "infant mortality" portion of the bathtub reliability curve — will manifest during the burn-in period rather than in the field.
The problem is that burn-in is slow, consumes energy, and is poorly matched to the actual failure mechanisms it's trying to precipitate.
Most latent defects are mechanical — marginal solder joints, incompletely seated connectors, cold welds at wire terminations. Thermal stress alone does not efficiently stress mechanical interfaces. What does stress them is thermal cycling — the expansion and contraction differential between joined materials — and vibration. A conventional burn-in oven applies neither.
A study published by JEDEC comparing traditional burn-in to HASS for a representative electronics assembly found that HASS screens at equivalent defect escape rates required roughly one-tenth the throughput time of equivalent burn-in programs. The HASS screen ran 45 minutes. The burn-in ran eight hours. The defect escape rate for HASS was lower.
That comparison matters in production economics. Eight hours of oven time per unit, at scale, is a significant capital and energy cost. Forty-five minutes of HASS chamber time per unit, for better coverage, frequently justifies the capital investment in HASS equipment within two to three years at volume production rates above a few thousand units per year.
---
The screen survival test: how you know HASS isn't damaging good product
This is the validation step that separates a proper HASS program from an aggressive burn-in with a different name.
Before deploying a HASS screen in production, a sample of known-good units — units that have been fully characterised and verified — is run through the proposed screen profile multiple times. JEDEC's guidelines and standard industry practice recommend 10 to 20 consecutive screen cycles on the same units as a minimum validation.
If the known-good units show no degradation after repeated screening — no change in electrical parameters, no evidence of physical damage under microscopic inspection, no reduction in HALT-determined operating limits — the screen profile is safe. It is strong enough to precipitate defects without consuming life from good product.
If the known-good units show degradation, the screen is too aggressive. The ramp rate, the vibration level, or the temperature range is consuming product life with every cycle. That consumed life comes out of the product's field reliability budget — the exact opposite of what HASS is supposed to achieve.
Validating the screen before production deployment, and revalidating it whenever the product design or manufacturing process changes significantly, is not optional. A HASS program running on an invalidated profile is a reliability liability, not an asset.
---
What HASS finds that other screens don't
The defect types that HASS is uniquely effective at finding — and that conventional screening consistently misses:
Marginal solder joints. The joint that passed AOI and X-ray because the void fraction was below the rejection threshold, but which has insufficient mechanical integrity to survive thermal fatigue. Fast thermal cycling propagates the crack to failure in minutes. Slow burn-in or functional test at room temperature never stresses it enough to find it.
Cold welds and incomplete wire terminations. A wire termination that appears mechanically sound but has insufficient metallurgical bond at the interface. The joint passes pull test at room temperature. Under combined vibration and thermal cycling, the insufficient bond propagates to an open. HASS finds it in the first screen cycle.
Incompletely seated connectors. Press-fit connectors seated 90% of depth instead of 100%. Board-to-board connectors with one row of contacts not fully latched. These pass functional test because electrical contact is made at bench conditions. Under vibration at field conditions, the incomplete seating allows micro-motion that generates fretting corrosion and eventually an open circuit. Six-DOF vibration in HASS exposes incomplete seating reliably.
Component parametric shifts. Components whose electrical parameters are within specification at room temperature but shift out of specification under thermal stress — capacitors with marginal temperature coefficients, resistors with high TCR, oscillators that drift frequency at temperature extremes. Continuous electrical monitoring during HASS catches these as functional failures during thermal excursions.
---
The economic argument for HASS
The cost of a latent defect escaping to the field is not the cost of the component that failed. It's the cost of the failure analysis, the service call, the warranty replacement, the customer relationship damage, the potential regulatory notification if the product is safety-critical, and the reputational consequence if the failure rate is high enough to generate visible pattern complaints.
Against that cost, the HASS investment has three components: the capital cost of the chamber (a HASS-capable combined environment system runs $150,000–$500,000 depending on configuration and workspace size), the validation cost of the screen development program, and the per-unit throughput cost of running each production unit through the screen.
The break-even calculation depends on production volume, field failure rate, and field failure cost — numbers that vary by product and market. But as a general reference, manufacturers who have implemented HASS programs consistently report field return rate reductions of 30–70% in the product lines screened, with payback periods on the capital investment of 18 months to three years at moderate production volumes.
Angelantoni Test Technologies publishes case studies on HASS implementation economics from their European customer base. Cincinnati Sub-Zero's technical library includes application notes on HASS screen development that are worth reading before scoping a program.
---
The programme that HASS doesn't replace
HASS screens production units. It does not replace the development reliability program.
A product with a fundamental design weakness — a component operating outside its rated temperature range, a PCB stackup with inadequate via barrel thickness for thermal cycling, a connector with insufficient mating cycles for the intended service life — will fail HASS screens consistently. The right response is to fix the design, not to soften the screen.
HASS is a manufacturing quality tool. It finds defects introduced by the manufacturing process — assembly errors, component lot variation, workmanship issues — not design deficiencies that exist in every unit regardless of how well it was assembled. HALT finds design deficiencies. HASS finds manufacturing defects. Both are necessary. Neither replaces the other.
The full reliability picture — from design characterisation through production screening to field monitoring — is what the environmental testing standards post addresses at the programme level. The types of environmental test chambers post covers the hardware differences between HASS-capable combined environment systems and conventional burn-in equipment in more detail.
---
The question worth asking before the next production run ships
How many units in your current production run have a latent defect that passed every inspection and test you ran?
You don't know. Nobody does. That's what makes latent defects expensive — they're invisible until they're not.
HASS doesn't make them visible at inspection. It makes them fail in a chamber instead of in a customer's hands. That shift in where the failure occurs — from field to factory — is the entire value of the program.
---
Next in this series: Thermal Shock Testing: Why Slow Ramps Miss the Failures That Matter · Temperature Cycling Testing · HALT Testing
Related: The Top 10 Environmental Test Chamber Manufacturers · Environmental Testing Standards Explained
---