In my experience, Vicor is known primarily among engineers who work with systems that must be able to withstand stressful environmental conditions. The Vicor website indicates that they supply products to a wide variety of industries—telecom, automotive, industrial, etc.—but the first item in the list is “aerospace and defense”. In any event, I recall thinking that their products were overall quite solid and, in this article, we’ll take a look at a converter series that may be a good option for any design that calls for above-average performance or reliability.
The specs for the ZVS series make me wonder: “How high can switching-regulator efficiency actually go?” This product page from Mouser indicates that at least one specific configuration can achieve “over 98%” efficiency.
I don’t read DC/DC converter datasheets in my spare time but, still, I’ve never seen a number that high. Pretty soon we’ll be reading about converters that actually add power to your system—i.e., you put in 10 W from the supply, you get 10 W for your load, and you also get an extra watt or two for charging a backup battery. If you think this violates the laws of physics . . . well, Newtonian mechanics turned out to be not quite right, and maybe we need to do some tweaking with thermodynamics as well.
ZVS products are available in LGA-style and (hooray!) non-LGA-style packages. Images taken from ZVS series datasheets (click here and here).
Seriously, though, efficiency is a critical parameter. In casual engineering, we can often ignore it but, when it comes time to deliver products, efficiency may be far from negligible.
A more obvious effect is battery life. It makes intuitive sense that a battery-powered product will need to be recharged more frequently if power is wasted in generating heat that is needed by neither the user nor the device.
But in this article we’re focusing more on high-performance systems that generally are not limited by batteries—industrial systems are plugged into the grid, airplanes have generators (or something equivalent, I’ve never designed an airplane), and so forth. Why are we concerned about a little bit of efficiency when we have a relatively abundant supply of power for our circuit?
The issue is temperature. Lower efficiency means that more heat will be generated for a given amount of power delivered to the load. This extra heat raises the temperature of the converter itself, and it also raises the temperature of nearby components, especially if everything is boxed up in an insulating enclosure.
For those of us who are accustomed to working at room temperature, it’s easy to forget about the restrictions imposed by harsh environments. Some systems need to remain functional at very high temperatures, and in this context low efficiency is an obstacle, because it limits the maximum ambient temperature at which a device’s internal temperature can be maintained within the acceptable range.
One thing to keep in mind is that input voltage affects efficiency. In the case of the PI34xx ZVS series, the difference is fairly small:
Plot taken from a ZVS series datasheet.
The ZVS series comes with reasonable requirements for external components:
Diagram taken from a ZVS series datasheet.
As you can see, few are required and, from the physical diagram on the left, we see that they can be arranged in an orderly fashion. That diagram also provides a clear reminder of (one reason) why there is a bit of animosity between modern electronics and inductors—you could fit six capacitors in the space required for that one inductor!
The Eval Board
You don’t have to design and assemble a custom board to confirm that a ZVS device will provide adequate performance for your application. An evaluation board is available, and it looks like you might even be able to get one for free.
You can extract some interesting information from the eval board if you also happen to own an infrared camera.
Image taken from a ZVS datasheet.
The idea is to vary the operational conditions—input voltage, load current, ambient temperature—and then use the camera to acquire precise data regarding the temperature of the converter itself and of nearby components. I could imagine a fairly enjoyable Friday afternoon in the lab with a variable power supply, a ZVS eval board, a thermal imager, and a bottle of homebrew.
Do you have any experience with Vicor products? Do you consider it unwise to consume alcohol while operating a thermal imaging device? Let us know in the comments.