Galvanic isolation refers to the situation in which portions of a system are electrically separated. If these subsystems don’t need to exchange information—well, the isolation is more or less a non-issue.
But things get a bit complicated when data of some kind must move between these isolated circuits. Typically we move data via conductors carrying voltages, but we can’t have a conductive path between isolated subsystems. Fortunately, there are other ways to communicate.
A familiar example is electromagnetic radiation: our smartphones, Bluetooth devices, and so forth are accomplishing galvanically isolated data transfer. However, we don’t want to build a radio when all we’re trying to is, for example, enable SPI communication between ICs in isolated sections of a control system. Two common solutions for circuit-level isolation are optical coupling and magnetic coupling; as usual, both can be implemented with the convenience provided by integrated-circuit technology.
Opto-couplers are not exactly new, and they may not represent the state of the art in IC isolation. But apparently Fairchild (now ON Semiconductor) still considers them a worthwhile investment, because they recently introduced a new line of opto-coupler products, namely, the FODM217 series. Despite the fact that the FODM217 comes in a small surface-mount package, it can withstand some serious voltage differentials between the input and output terminals: the datasheet specifies a peak working insulation voltage of 565 V, with the ability to survive 3750 VAC (RMS) for 1 minute.
The high-frequency performance doesn’t seem overly impressive to me.
Image courtesy of Fairchild Semiconductor
Both tON and tOFF are given as 3 µs typical, so it seems to me that you can’t expect to operate at frequencies much higher than 100 kHz. I suppose this must be adequate for the part’s intended uses; the datasheet says that the FODM217 is “primarily suited” for isolation in DC/DC converter circuits, though it also mentions communications, consumer, and industrial applications.
The ADuM250N, along with various other products from ADI, takes an entirely different approach to galvanic isolation. They call their technology iCoupler, and it employs magnetic fields rather than light to transfer data while maintaining electrical separation.
The ADuM250N is specifically designed for digital signals, and this is reflected in its high-frequency performance (which, in all fairness, makes the Fairchild part look painfully slow). The propagation delay is 6.8 ns typical at VDD = 3.3 V, and the datasheet indicates a “guaranteed data rate” of 150 Mbps.
The isolation capabilities of the ADuM250N are a little better than those of the FODM217: 5000 VAC (RMS) for 1 minute and a maximum working insulation voltage of 849 V. However, those of us working with low voltages will be more interested in the various other features mentioned on page 1 of the datasheet, including high immunity to transients and noise, low power consumption, and fail-safe functionality.
Another new digital-signal isolation product is the MAX12931. I’m guessing that this part is also based on magnetic coupling, but it appears that Maxim is not eager to reveal details about their isolation techniques. The part description merely refers to Maxim’s “proprietary process technology,” and I couldn’t find further detail elsewhere in the datasheet.
Fortunately, you can still use this part without knowing how exactly it works. The maximum data rate provided by the MAX12931 is similar to that of the ADuM250N; however, the MAX12931’s power consumption is significantly lower, so you might want to consider a Maxim isolator for use in battery-powered applications. Also, note that there is a wider package option and a narrower package option; opt for the wider package if you need higher isolation-voltage capabilities.
Do you have a preferred type of isolator? Feel free to share your experiences in the comments.