09.11.12 | Autor / Redakteur: David Krakauer * / Kristin Rinortner
Digital isolators have made an already complicated puzzle of safety standards even more confusing because not all standards address the requirements for digital isolators.
Designers do not add galvanic isolation to their systems because they want to – they do it because they are required to meet domestic or international safety regulations. The downside is that isolation is placed directly in a data path, introducing delays and slowing down system performance. Adding isolation also increases power consumption, size and cost. These are unfortunate trade-offs.
For years, designers used optocouplers and grudgingly managed the trade-offs, but a new breed of galvanic isolators, digital isolators, have come to market and reduced those penalties. They enable smaller, more energy efficient and cost effective designs capable of higher levels of performance.
However, safety standards have not kept pace, creating confusion and uncertainty about whether digital isolators can achieve the one reason designers use galvanic isolation: do they meet safety regulations?
The answer is yes, digital isolators can provide the same safety required by domestic and international standards. However, unlike optocouplers from most suppliers which have similar structures, digital isolators are designed and manufactured in different ways that affect isolation capability, particularly when compared to the stalwart isolation capability of optocouplers.
Therefore, not all digital isolator technologies and implementations provide the same level of safety.
Consider four key isolator elements:
There are different options for each element, and the resulting combination defines an isolator’s capabilities. We will focus on the insulating material which is a key differentiator for safety. Optocouplers use a variety of polymer materials, including the epoxy molding compound of the package. Digital isolators use a similar polymer, or polyimide, material or they may use silicon dioxide.
The materials and manufacturing process lead to differences in both the life time of the insulation and the ability to withstand high voltage surges. Let us first consider safety standards and how they relate to different types of isolators.
System-level standards address differences between environmental conditions and system usage. Requirements for household appliances, for example differ from patient monitors used in hospitals or motor drives in factories. They often address isolation safety by calling out component-level standards specific to galvanic isolators. There are three such standards of note:
While each has a similar objective – ensure user, operator and equipment safety – they take different approaches. IEC 60747 includes distinctions between classes of isolation (e.g., “basic” vs. “reinforced” insulation) while UL 1577 emphasizes the capability of isolators to withstand certain voltage levels over a defined period of time, typically one minute.
It is common for system designers to rely on certification from more than one of these component-level standards in order to cover all possible uses and conditions.
The rise of digital isolators has complicated matters because many of these standards were written at a time when designers were stuck with optocouplers. The standards address weaknesses associated with optocouplers and provide means for guaranteeing safety.
These methods work well for optocouplers; however, they may not apply to digital isolators. Consider the case of certified working voltage, which is the continuously applied voltage across an isolation barrier. The expectation is that an isolator with a certified working voltage should withstand that voltage over its life.
IEC 60747 requires a production partial discharge test to validate optocoupler working voltages. Standards bodies have determined that partial discharge inception and deception voltages correlate with optocoupler working voltages. The manufacturing process uses an injection molding which is prone to creating voids within the plastic material.
These voids can experience higher electric fields under stress and result in partial discharge induced degradation. Using a partial discharge test at high voltages detects the presence of voids and can be used to reject parts that would otherwise fail in the field.
This partial discharge approach is not fully applicable to digital isolators. Digital isolators do use similar package materials which must be tested for defects using partial discharge, but there are other aging mechanisms related to the insulating materials. The main isolation materials used for isolation elements deposited through well-controlled wafer-level processes and are less prone to voids and thus partial discharge; however, other aging mechanisms start to dominate. When a digital isolator claims a certain working voltage, usually denoted as VIORM, based on IEC 60747, that may be misleading as it reflects only the ability to pass a partial discharge test at a given voltage
Because partial discharge is an incomplete test for digital isolator working voltage, additional testing and characterization are required. Future standards from IEC will address this and incorporate these new methods. In the interim, it is incumbent on digital isolator suppliers to show how they guarantee lifetime operation at rated working voltages.
Dieser Beitrag ist urheberrechtlich geschützt. Sie wollen ihn für Ihre Zwecke verwenden? Infos finden Sie unter www.mycontentfactory.de (ID: 36003410 / English)