Once you understand the types of stress and types of failure that make up a HALT test, you can begin to understand how those stresses are combined into a series of tests that offer informative results.
Order of Tests
HALT is controlled, systematic, and rigorous. Unique testing setups are created for each product to ensure that appropriate functions are evaluated–for example, a screen may be tested to ensure that it displays correctly at the various stress points. But the types of stress, and the order in which they are presented, are fairly standard.
Classic HALT testing consists of five phases:
- Cold Thermal Step Stress
- Hot Thermal Step Stress
- Rapid Thermal Stress
- Vibration Step Stress
- Combined Thermal and Vibration Stress
“Step Stress” means that stresses are increased in increments (e.g. 10 degrees C at a time) and then the product is left to “soak” at that stress level before the next increment. This helps identify failures that might occur after a certain amount of time with a given stress and prevents both failures that could be caused by unnecessarily fast ramping and false measurements that could come from lack of soak time before incrementing stress.
The Cold and Hot Thermal Step Stress tests measure the product’s tolerance to extreme temperatures. Generally, 10 degree C increments are used, with ten minutes of soak time once each temperature is reached. At the end of each soak, the functions specific to your product could be tested–for example, power cycles, input and load variations, and performance. The increments could be decreased to 5 degrees C as the test nears the product’s limits. A typical temperature range covered by HALT is -100C to +200C. Practically speaking, that means HALT gets colder than the coldest recorded natural temperature, and gets as hot as your oven does while baking bread.
Rapid Thermal Stress is not a step test. It tests the product’s tolerance to rapidly changing temperature, ranging to within 10 degrees of the product’s destruct limits (established in the previous Cold and Hot Thermal Step Stress tests). Typically, the temperature is changed as quickly as the chamber will allow. A typical speed of temperature change in HALT is 70C per minute.
The fourth phase is Vibration Step Stress testing. Like the initial thermal tests, this is a step stress test, with the rate of vibration increased incrementally and then held steady for specific soak periods. Typical HALT chambers produce random vibration in the 0-10kHz range. Testing often starts at 5 GRMS (a measure of the total power of vibration across frequencies) and steps up by 5 GRMS at a time until failure. This will allow you to understand the vibration response of the product.
The final phase is the most stressful for the product. This phase combines Rapid Thermal Stress with Vibration, showing how the product acts when suffering different types of stress simultaneously.
The goal of the HALT is to find the product’s operating limits, and destruct limits, for each of the five steps. This becomes the product’s baseline. Any design changes can be measured and compared to that baseline, helping to develop a stronger and more rugged product.
For example, testing one of our media players revealed a weakness in the display. It would develop lag, and finally stop, at extreme temperatures. That was the unit’s operational limit. But we wanted to understand our destruct limit, too, so we knew whether the unit would survive in extreme environments. So we isolated the display and continued testing. Finally, a processor burnt up. That was our destruct limit. The display was our operational limiting factor, and that one processor was our destructive limiting factor. Replacing or shielding those components would result in a more rugged product.
We’ve been using HALT on our own products for a decade. It’s helped us diagnose design problems ranging from the complicated to the mundane, and then test that our solutions to those problems will work.
As a rigorous way to evaluate a product–in a short time span–HALT can’t be beat.