Scientific laser users have long relied on state-of-the-art performance (e.g., higher peak power, shorter pulse duration, wider wavelength tuning) to achieve groundbreaking results. Unfortunately, this high performance has often been emphasized at the expense of ease-of-use and reliability. Recently, however, this paradigm has dramatically changed, and some of the latest scientific lasers — including complex ultrafast amplifiers — now deliver both cutting-edge performance and exceptional reliability. This advance is sometimes referred to as “The Industrial Revolution in Ultrafast Science.”

Designing Reliability – Back to the Future

Figure 1. Conceptual representation of vibration and temperature profiles for HALT/HASS test in relation to operating and non-operating environmental condition for a given product.

While 100% reliability is the goal of many manufacturers, in reality, some failures will occur with any new product. There are two primary reasons for this. First are inherent weaknesses in the way the product and/or its key components are designed and engineered. Second are manufacturing errors. These can be systematic problems, or just a combination of random statistical deviations in various production processes. Engineers have long recognized these problems and sought methods to address them. In some industries, like aviation or medical equipment, in particular, the focus has been to identify and eliminate potential failures, before they occur.

The most direct approach to identifying failure sources is simply to put products out in the field and wait for them to fail. Needless to say, this isn’t a particularly effective tactic for companies concerned about their customers or their own reputations.

Another — and just as impractical — approach would be to manufacture products in their final form, and rather than delivering them to customers, continue to operate them under standard operating conditions and track their failures — a process that could take years before acquiring any statistical significance.

Figure 2. To enable high-density HALT/HASS testing, Coherent invested in a large (Qualmark Typhoon) chamber capable of accommodating even a complete ultrafast amplifier weighing hundreds of pounds.

But what if there were a way to put products out in the field long enough for all potential design and production weaknesses to reveal themselves, and then to travel back in time and redesign the product so that these weaknesses are eliminated beforehand? While actual time travel isn’t possible, in essence, it’s the underlying concept behind a set of protocols called Highly Accelerated Life Testing and Highly Accelerated Stress Screening (HALT/HASS).

HALT/HASS have proven themselves extremely effective in numerous other industries where failure is simply unacceptable. These accelerated testing methods serve to effectively compress time by orders of magnitude. This makes them particularly useful in cases where waiting for customer feedback is not a practical option because of low production volumes or other factors.

HALT – Highly Accelerated Lifetime Testing

HALT is a method where simultaneous stressors, typically but not solely temperature and vibration, are applied during the product design phase to greatly accelerate normal aging. To provide ample design margin, the HALT conditions should be more extreme than the specified operating and non-operating limits. HALT deliberately drives a product to failure in order to identify, and then eliminate, its weak points. Iterative cycles of HALT testing, re-design, and further HALT testing serve to eliminate any weakness or potential failure mechanism in the final design. The resulting final product will then reliably withstand stresses beyond its specified level.

HASS – Highly Accelerated Stress Screening

Figure 3. Some HALT results. Comparison of an earlier version (top) of optical mounts with the final version (bottom) used in a femtosecond laser manufactured at Coherent. The final version is much more resistant to externally induced vibration and temperature change cycles.

HASS is used to screen production units for any manufacturing weaknesses or errors. Once the overall design for the product has been “frozen,” but before starting the standard production cycle, an appropriate HASS test protocol is defined and implemented. These stresses are not as high as during the HALT phase because the goal is to reveal workmanship and material issues, without compromising the lifetime and performance of each tested unit (Figure 1).

At first, HASS is performed on each manufactured unit, verifying that there are only negligible variations in performances before and after the HASS test cycle. The analysis of failed units determines the weakness and provides paths to correct them, be it a material quality issue or a workmanship error. Once a sustainable, high pass rate is achieved, HASS is then performed on a statistically significant sample of production units throughout the lifetime of the product.

Practical HALT/HASS

Implementation and Immersion Because HALT/HASS had never previously been used in the laser industry, we had to first determine how best to implement HALT/HASS at Coherent. We partnered with Qualmark, a recognized expert in HALT/HASS, with a long history of successfully providing solutions in closely related industries, such as consumer electronics, medical, and defense. Moreover, Qualmark is the chosen HALT/HASS resource for several leading companies in civil avionics, an industry with understandably high-reliability targets.

Very early on it became clear that Coherent would have to invest in HALT/HASS equipment in-house, for two reasons. First, experience from multiple industries confirms that HALT/HASS is most successful when it is an integral part of the design and manufacturing process. Occasional, out-sourced testing at third-party facilities will not deliver the same results as high-density testing and screening. Outsourcing also adds time and cost to any iteration design cycles.

Second, for maximum effectiveness, HALT/HASS has to be fully embraced, from the component level up to the complete product. Coherent is a vertically integrated company where all critical components and sub-systems are manufactured in-house. Therefore, full implementation was clearly going to require significant, ongoing HALT/HASS activity — far more than could ever be realistically outsourced.

Figure 4. Vitara-S life test run at constant pump power.

Years of experience with industrial lasers had taught us that optical alignment issues were the leading cause of laser unreliability, followed by optics failure, and then electronics failure. To screen for these failure mechanisms, Qualmark recommended the purchase of one of its high-performance Typhoon testing chambers. These are capable of accommodating even a complete ultrafast amplifier, and can subject it to extraordinary thermal swings accompanied by digitally controlled vibrations in three dimensions (Figure 2). With a temperature range from minus 100°C to plus 200°C, the entire chamber (even when filled with lasers) can be driven through a 100°C oscillation in less than two minutes. Additionally, the vibration platform can handle lasers or components totaling hundreds of pounds in weight. This platform is supported by 24 separate pneumatic actuators, enabling programmable and/or random three-dimensional vibrational sequences with accelerations up to 70 g.

Sample HALT/HASS Results

The following examples demonstrate just a few of the results that have been achieved by applying HALT/HASS protocols from the component through the system level.

Figure 5. The HASS test protocol currently used for every single unit of a new type of short-pulse (< 35 fs) ultrafast amplifier called the Coherent Astrella.

High-Stability Optical Mounts. Figure 3 illustrates the successful use of HALT methodology in the design of optics mounts for a femtosecond laser. In the first design iteration, about 60 mounts were tested, and a significant population showed misalignment after extreme temperature and vibration testing. Moreover, the performance of this 60-unit sample covers a wide statistical range. As a result of these experimental findings, the mounts were redesigned. Repeated testing after each design cycle quantified the stability statistically. Figure 3 also shows the much improved stability and consistency from HALT tests of the final design iteration.

One-Box Ultrafast Oscillators. There has always been an accepted trade-off between short pulse width and reliability in ultrafast lasers. In fact, users report that ultrafast lasers from some manufacturers need optics cleaning and adjustment on a weekly interval or less. Figure 4 shows the end result on the first generation of broadband ultrafast oscillators, designed from the ground up using the HALT/HASS approach. This is a 12,500- hour run on the model Vitara-S, operated at constant diode pump power (i.e., without any light feedback loop), which indicates the absence of any degradation in the optical losses of the resonator components.

Integrated Femtosecond Amplifiers. Surely no commercial laser system is more challenging, in terms of stability and reliability, than a femtosecond amplifier. Moreover, today, many applications prefer a turnkey, one-box system with simple on/off functionality and reliability to pump various types of nonlinear processes (like parametric amplification or terahertz generation).

Figure 5 shows the HASS protocol used to test this first generation of ultra-fast amplifiers designed and manufactured completely using HALT. As can be seen in the figure, this protocol combines different intensity bursts of vibration, together with extreme temperature transients of 50 °C in 2-3 minutes. The system must show no significant change in output parameters at the end of these tests or it is rejected for shipping.

The end result for the laser user? As described in a recent case study, university researchers are successfully using one of these lasers to conduct two-dimensional spectroscopy in experiments requiring up to 48 hours of uninterrupted data acquisition.


The quality, consistency, and throughput of experimental data are essential contributors to a team’s or individual investigator’s success. Laser performance and reliability directly impact these, for example, by delaying the acquisition of data needed for a conference paper or time-critical grant proposal. In the case of tenure track, this data can even make or break a scientist’s academic career. While scientific laser users have long accepted a tradeoff in performance versus reliability, the implementation of design techniques already used in industrial lasers, together with HALT/HASS protocols, have now eliminated this compromise.

This article was written by Joseph Henrich, Allan Ashmead, Steve Butcher, and Marco Arrigoni of Coherent, Inc. (Santa Clara, CA). For more information, contact Mr. Henrich at This email address is being protected from spambots. You need JavaScript enabled to view it. or visit .

Leveraging Industrial Laser Experience

Coherent’s success with implementing HALT/HASS with scientific lasers was also facilitated by our experience in manufacturing industrial lasers, where the economic impact of downtime and higher product volumes had already served to drive high reliability. In fact, while HALT/HASS provides a method for improving product design and manufacturing methods, it also needs a good starting point. Our scientific laser team was able to draw on the expertise of engineers working with Coherent industrial lasers, already well-versed in good engineering design practices and knowledgeable in materials and production methods needed for long, maintenance-free lifetimes.

A typical example is understanding and avoiding cavity contamination to ensure long optics lifetimes, without frequent cleaning. For example, some of the degradation effects of materials subject to intense nanosecond UV laser pulses with high average power are similar to the ones resulting from femtosecond near-IR pulses where multi-photon effects play a role. Although a nanosecond UV industrial laser may be quite different from a femtosecond laser for a spectroscopy lab, the challenges in controlling degradation mechanisms may be surprisingly similar.