“A tech for a well-known audio manufacturing company posted to a forum that it’s not wise to power devices that have word clock off a circuit that’s protected by a UPS backup system. He mentioned something about stepped power. I would like to get a second opinion, please.”
You may get a third, fourth, and fifth opinion as well…and still no consensus. There are a variety of unspecified aspects to your question that make it difficult to precisely answer, but there are some points worth considering in general.
The first we’ll talk about is the stepped power issue. What is stepped power? The output of many (lower cost) UPS systems is not a true sine wave like you ideally get from the power company. It’s relatively difficult to build a good AC sine wave from a DC power source (i.e. a battery) so quite often to save costs UPS’s will be designed to output only an approximation of the sine wave. The resulting waveform resembles a sine wave, but instead of a smooth curve there are a number of distinct steps as the voltage goes up and down. This is known as a stepped sine wave or approximated sine wave. They get away with this because in most cases a true sine wave isn’t necessary to power the equipment we use. Most power supplies in the equipment we encounter can “adequately” filter out inconsistencies in the incoming AC waveform. In fact, some high speed switching type supplies (like those found in most modern computers) actually run more efficiently on stepped power.
The quotes are around the word “adequately” above because there are understandably different opinions of what adequate really is. In general you can assume that better (more expensive) pro audio equipment is equipped with better power supplies, and thus able to perform better under duress, or at least with less than perfect AC power. In fact, over the years we’ve observed that this is one of the more underrated and overlooked benefits of buying higher quality equipment. However, it doesn’t necessarily follow that even the best power supplies are totally impervious to even minor inconsistencies in the source of power. And the exact nature of the power anomaly may effect different supplies to different degrees. Opinions on this vary, but there is evidence to suggest that even subtle changes in the power we feed our equipment can cause “significant” (yes, there are those quotes again, for the same reason as before) changes in performance. And this is before we even begin to talk about the effects of noise and other problems that occur on our AC lines as a result of other equipment in use at a given location. One common symptom you will observe with audio gear not reacting well to the output waveform of a UPS is an increased level of hum.
You can now somewhat anticipate how this relates to your clocking question. Most of us “non engineer” types still have a lot to learn about the subtleties of word clock, jitter, and other phenomenon critical to digital audio. Just as old recording “engineers” (quotes for emphasis on the word engineer, which is what they really were way back when) had to really know about the properties of tape, magnetism, and tape heads, we need to know about things like jitter and clocks.
Many devices that accept external word clock signals do it as part of a process that involves a PLL (Phase Locked Loop) and VCXO (Voltage Controlled crystal Oscillator). Essentially this circuit compares the output (phase) of the VCXO to any incoming clock signal and reconciles them. It does this by using a voltage that is generated from any difference to “pull up” or “pull down” the frequency of the VCXO. (This explanation is drastically oversimplified, and not necessarily even true of all devices, but in a general sense it will serve our purposes.) Once we know that voltage differences can change the output frequency of the crystal in a device it’s not too much of a stretch to think that power supply fluctuations may also effect it regardless of whether we are actually synchronizing it to external clock signals. In fact, many audiophiles theorize that this is indeed the case, and is one reason why they are very concerned with the power supply used in their digital audio devices.
Now, whether all this implies you shouldn’t use a stepped AC output as your supply voltage for your digital devices is a matter for further discussion and research, but your inSync editor is inclined to think there is merit to this person’s point. And there is just as much, if not more, merit to the point that higher quality power supplies are something that separate a great deal of mediocre equipment from the really good stuff. While I don’t know of any objective tests comparing the behavior digital devices under different types of power ranging from very “clean” to very “dirty” I think it is likely that in general higher quality devices will outperform their lower quality counterparts by even more when given less than ideal power. Of course there are always exceptions, one of which could be switching power supplies. Many switching power supplies are thought to be of lower quality than their linear counterparts, and they certainly cost less, but under the right circumstances their performance may not degrade nearly as much when driving from a stepped power source such as a UPS. How do you tell if your unit uses a switching type power supply? Your best bet is to call the manufacturer, but they are usually much lighter than linear supplies and cost less. That should give you a clue.
One final point of significance. Most low cost UPS’s that are of the modified sine wave variety only output the modified sine wave when they are activated – meaning their power is removed and you are operating off of the battery. In normal day to day affairs the battery is not part of the circuit and the unit is just passing through the normal power that comes to it from wherever you have it plugged in. So the UPS itself is really nothing to worry about. Good power supplies in your gear, however, is something you certainly should be concerned about. They’re in use all the time.