Toshiba Triple-Gate IGBT Achieves 40% Reduction in Power Loss

The Significance of a 40% Loss Reduction Figure

Toshiba has announced a new IGBT structure, which they call "Triple-Gate IGBT," boasting a maximum loss reduction of 40% compared to conventional models. This 40% figure represents more than just a generational upgrade in specifications. In the domains where IGBTs are predominant—industrial equipment, railway systems, and large inverters—a 1% improvement in loss directly translates to reduced cooling costs and annual electricity expenses. A 40% reduction, therefore, signifies a fundamental shift in design prerequisites.

IGBTs are silicon (Si)-based power switching devices that control conduction and cutoff via gate signals. While sometimes viewed as an "older technology" compared to next-generation materials like SiC and GaN, IGBTs maintain a strong cost-performance advantage in high-current, high-voltage applications, continuing to play a central role in industrial inverters and railway traction systems. Toshiba's announcement is an endeavor to significantly advance IGBTs while remaining within the silicon paradigm.

What is Triple-Gate and Why Does it Reduce Losses?

Conventional IGBTs are fundamentally structured with a single gate that controls the channel through which current flows. Triple-Gate, as the name suggests, incorporates three gates. This structure effectively expands the channel width, allowing more carriers to flow within the same die area. The objective is to simultaneously reduce the on-voltage (VCE(sat)), which corresponds to on-resistance, and switching losses.

IGBT losses can be broadly categorized into two types: conduction losses, which occur while current is flowing, and switching losses, which arise during the on-off transitions. Traditionally, efforts to reduce conduction losses often led to an increase in switching losses, presenting a trade-off that designers had to prioritize based on the application. The Triple-Gate structure is expected to mitigate this very trade-off. With three gates, the freedom to control carriers increases, making it easier to suppress both types of losses simultaneously.

However, it is important to confirm the conditions under which the "40% reduction" figure was measured, and whether it applies to conduction losses, switching losses, or total losses when making selections. The apparent performance of power semiconductor specifications can vary significantly depending on measurement conditions.

Why Advance IGBTs in an Era Dominated by SiC?

This question can be readily addressed by considering the industry's structural reasons. While SiC devices offer high efficiency, they remain expensive compared to silicon IGBTs for equivalent voltage and current classes. Although justifiable for high-value applications like EV main inverters, a complete transition to SiC is still not practical for many cost-sensitive large industrial equipment and railway traction systems that require high-current modules.

SiC devices present another challenge. Due to their smaller die size and higher current density, they experience faster temperature rises during short circuits compared to silicon. This imposes stricter demands on the response speed of protection circuits, significantly impacting gate drivers and system design. IGBTs offer greater compatibility with existing systems and are easier to leverage existing design assets. Toshiba's investment in IGBT advancements stems from the recognition that there is a substantial market where the "replacement cost" cannot be ignored.

How are Competitors Responding?

Toshiba is not alone in pursuing significant loss reduction in IGBTs. Fuji Electric continues to develop low-loss IGBT modules, and Mitsubishi Electric is enhancing short-circuit withstand capability by introducing a p-type protective layer into the trench structure of SiC MOSFETs. These efforts demonstrate a common theme across companies, transcending material differences: the pursuit of "balancing loss reduction and robustness."

While Infineon and onsemi are focusing on SiC, Infineon remains actively developing IGBTs, particularly for the high-current module market serving railway and wind power applications, where competition persists. Onsemi has established a SiC portfolio covering voltages from 650V to 1700V, aiming to penetrate voltage ranges traditionally dominated by IGBTs.

Toshiba's Triple-Gate IGBT signifies not a binary choice between "SiC or IGBT," but rather a landscape where each company is offering its unique answer to the question of "How far can IGBTs be advanced to maintain competitiveness in which markets and for how long?"

This graph illustrates the wide range of voltage bands where IGBTs are dominant. SiC currently primarily operates below 1700V, with limited options available in the 3300V and above range. If the technology for a 40% loss reduction can be extended to these medium and high-voltage bands, the pressure for replacement will be considerable.

Translating the 40% Loss Reduction into Design and Procurement Decisions

Even with technically superior figures, they are not sufficient for decision-making without an understanding of their impact on the overall system. When a 40% loss reduction translates to system efficiency, it may alter the design parameters of the cooling system (heatsink size, airflow volume, need for liquid cooling). This could manifest as a reduction in board area or system volume, or cost savings, but the actual effect depends on the specific design implementation.

From a procurement perspective, the ramp-up schedule and supply stability of new structure devices are key decision factors. New technologies often face challenges with specification variations in initial lots and production yield. How the "40% loss reduction on paper" is reproduced in mass-produced items will be confirmed through comparative evaluation of samples and production lots.

The technological achievement of a 40% loss reduction has the potential to change the perception of IGBTs as a "mature technology." In markets where the cost of transitioning to SiC is difficult to justify, these devices may emerge as a concrete option for design selection in the coming years. Further updates on the technology roadmap and mass production schedule will provide valuable information for more precise decision-making.