There are lots of companies trying to sell you gear to “make your power better.” Here’s what you really need, and why.
By Scott Dorsey
Nothing in the recording field gathers as much misinformation and just plain nonsense as AC power issues. I’ve seen people spend huge amounts of money on power conditioning just because they thought they should—believing that they’d solve problems that, in fact, are totally unrelated to power issues.
Power conditioning is a broad topic. A lot of devices out there come under the name “power conditioners.” There are simple line filters that are just lowpass networks that remove power line harmonics. There are voltage regulators of various sorts—multitap transformers operating in steps, and ferroresonant transformers that give tighter regulation at the expense of weight and operating difficulty. There are UPS (Uninterruptible Power Supply) devices, which actually make harmonic problems worse but which are essential for workstations in environments where power outages are common. Then there are isolation transformers, balanced power transformers that provide isolation and also reduce noise problems for chassis leakage. And then, of course, there are various combinations of the above.
All of these devices solve individual problems, and most of them cause other problems as side effects. The one device you want to use, if any, depends on the particular line problems you might have. If you have frequent outages and a system that loses data in an outage, there is no way around getting a UPS. But if you have line noise problems and you get a UPS to solve it, you’ll probably find your problem getting worse rather than better. That’s why it’s important to find out what problems you might have, and look for a device to solve them. Here are some of the more common devices out there.
The surge suppressor is a common gadget that you’ll find all over the place.The general idea is that it acts on your power line, sort of the way a limiter acts on audio; it takes large spikes and clamps them down to a lower level. This is absolutely essential if you’re in an area where there are occasionally big spikes on the line, usually due to induced lightning noise or large industrial customers that aren’t very well protected. It’s possible to gets spikes as high as 6KV across the line before they arc across at the terminals in a typical breaker panel, but dealing with spikes below 6KV is left up to the surge suppressor.
Let me first say that if you are in an area where spikes are common, it is well worth the money to install a lightning arrestor at the service entry of the building. The power company may even be willing to do it for free, or you may have to get an electrician to do the work. Either way, the arrestor at the service entrance can prevent serious future problems. Make no mistake—a nearby lightning strike is in a different league than an “ordinary” surge, and can put a serious hurt on otherwise protected systems.
Once that is done, you can then get small surge suppressors for each piece of equipment. These come in three basic kinds: the sort with MOVs (metal oxide varistors), with MOVs and gas discharge tubes, and with active electronics.
The MOV types are the least expensive and they rely on two strips of zinc with a thin oxide coating between them. The oxide coating is insulating to lower voltages, but at around 200V or so, current abruptly arcs through the insulating layer and it becomes conductive. This clamps spikes above that point down very effectively, although it can take some microseconds for the layer to break down, and after a certain number of hits the layer will break down completely and the device will fail into a dead short and pop the breaker on the strip. This is actually a good thing—when it fails, you can tell that it is bad because the breaker won’t stay closed.
The gas discharge device is a little arc gap inside a sealed glass tube full of an inert gas. Unlike a typical arc device, it can spark over many times without damage, and it, like the MOV, is normally nonconducting but becomes abruptly conductive at one voltage level and clamps anything above that level. It has a sharper operating curve; that is, it becomes more conducting at a single specific point. It also costs more money. Usually gas tubes and MOVs will be mounted in parallel together, giving you the fast response of the gas tube and the high current sinkability of the MOV. Gas tubes do degrade, and if they aren’t paralleled with an MOV, they will degrade faster than just an MOV device.
Currently there are a lot of folks out there who will tell you that MOVs are bad because they fail by their very nature, and because you can’t tell when they are going to fail. These folks tend to be promoting solid state devices like the Trans-Zorb, or other Zener-diode type semiconductors, which turn into a short abruptly at a transition frequency. These comparatively new devices also fail. They are a good replacement for the MOV, and they are faster, but again they are best when used with a backup device like a gas tube or even an MOV, and they seem to be surrounded with a huge amount of hype in spite of being good devices.
There are also devices out there that use active electronics. Rather than short a spike to ground, these devices briefly disconnect the load from the line when a spike occurs. The manufacturers of these devices claim that they prevent ground corruption problems from when spikes are shunted to ground. This is a correct statement, but the devices I have tested are also less effective at preventing spikes from making it through to the protected equipment, and they can fail silently in a manner where they no longer protect the equipment but give no sign of a problem. For this reason I cannot recommend any of the ones I have seen, although I agree that the idea is a good one—in theory. Again, there is a huge amount of hype attached to some of these devices.
Don’t believe anyone who tells you that their device doesn’t ever fail. Everything fails sooner or later. The question is whether they fail in a safe way or an unsafe one.
Be aware that there are a lot of grossly undersized and cheap surge suppression devices out there that really don’t do a very good job, and which self-destruct. They don’t really do more than give you a false sense of security. There is also a lot of meaningless junk in many of the ads and the data sheets, like bragging about conforming to UL1449 (which basically means the surge suppressor won’t injure you), or being able to deal with the ANSI C62.41 standard surge waveform (which is something you would expect any surge suppressor to do).
And, a surge suppressor is only as good as the grounding system of the service that it’s attached to. Since the suppressor shunts spikes to ground, if your ground is poor (and it can be pretty poor and still pass the building code), it won’t do much good.
There are no real disadvantages to using surge suppression, other than the fact that good surge suppressors cost money, and you could sometimes spend the money somewhere else. For field work and studio work, especially in areas prone to thunderstorms or with poor line suppression on the utility lines, good surge suppressors certainly are a good idea. A suppressor attached directly to the service entry is worth every penny.
Noise filters prevent high-frequency garbage on the power line from making it to your equipment. They are rather like lowpass equalizers for the power line. Many inexpensive power strips have a simple noise filter made from one or two shunt capacitors across the line. These tend to have a very low filter slope, sort of like a parametric EQ with the Q set very wide. Because of this they are normally built with a very high turnover frequency so they don’t waste any of the 60 Hz power. This makes them fairly effective at RF frequencies but not very effective down in the kilohertz area. Even so, it’s sometimes enough.
You can get filters with more poles, and therefore a greater slope, from companies like Corcom, as modules intended to be built into equipment. Some vendors (IEPS comes to mind) also make similar devices in standalone boxes. If you have a noise problem on your power line, these can do an admirable job of reducing the problem.
In general, though, it’s more effective to fix noise problems at the source rather than to filter them, because when radio signals get into the power line, they then radiate out of the lines and leak into everything. You can filter them off of the power line, but they are still radiating in the air. After all, that’s why it’s called radio.
So, if you are getting mysterious buzzes on guitar pickups and mic lines, go to the circuit breaker panel; start popping breakers one at a time, trying to see if something in your own building is putting noise onto the power line. When you find it, put the noise filter right on the appliance that is producing the problem.
Unless, of course, it is a dimmer or a touch lamp. Dimmers and touch lamps are the most evil noise sources known to man—the only real fix is to eliminate them. Neon sign transformers in bars are also great sources of radio frequency noise that leaks into everything around, and often the only real solution is to shut the sign off.
A noise filter on the power line won’t really hurt anything, but it’s only effective if the noise is confined to the power line and not radiating through the air. For field recording work, a noise filter is a handy gadget that gives you an added measure of safety from possible line noise trouble. A lot of devices come with both noise filters and surge suppression in one convenient box.
An isolation transformer is a transformer whose primary voltage is the same as the secondary. There are two coils; power goes through one coil, which generates a magnetic field that is picked up by the second coil, which turns it into power again.
This does a couple of things. First of all, it acts as a natural lowpass filter. The frequency response of the transformer is fairly small, so high-frequency noise on the power line tends not to pass through the transformer very well; that makes the transformer a filter.
Second, since the secondary and the primary of the transformer are completely isolated, you can operate a device off of the isolation transformer safely with the power line safety ground lifted. While it is thoroughly, thoroughly hazardous to use “cheater cords” that disconnect the ground pin on the power line to solve hum problems—and musicians have been killed doing it—the isolation transformer allows you to lift the ground on a piece of equipment safely.
This can be very, very handy when folks bring in guitar amps or synth racks with serious internal problems, and lifting the safety ground is the only way to get the noise down. It’s not a good long-term fix for anything but it can be a very handy quick fix for problems like this when there is no time to fix them properly. I often carry a couple small ones around with me to remote recording jobs, since they can be lifesavers in dealing with temporary equipment and power troubles.
Multitap voltage regulators
There is currently a real craze for multitap voltage regulators, and there are a lot of them out there from companies like Furman. The heart of these boxes is an autotransformer with a whole bunch of taps on it, every five volts or so, and a stepping relay arrangement that adjusts the tap that the output is on, so that the output voltage is always within five volts of 120V.
In a place where the power line voltage is consistently low or high, this is a very handy thing to have. Because it’s an autotransformer (a transformer with only one winding that serves as primary and secondary), you get no isolation and no filtering, but you do get some amount of regulation. And these devices are very small and light since the autotransformer is small and efficient.
The bad news with these is that you only get regulation within a few volts, and in environments where the voltage is rapidly swinging up and down, they will constantly bounce from one tap to another and back, and the end result might not be all that much better than the power going into it.
Because a multitap voltage regulator is so light, it’s a good choice to have out in the field. But most voltage problems in the studio are problems that need to be fixed in the building’s electrical system or at the power company’s service point. Voltage regulation can only hide these problems, and it does a poor job of hiding them. Sometimes when you’re playing in a club with serious wiring problems, though, it can be enough. At other times it’s not enough at all, since the regulation is in fairly large steps.
I often see people buy these gadgets to fix particular problems with the power line voltage rising and falling as a large load in their building is switched on or off. But they find that multitap voltage regulators don’t really fix these problems because they just bounce from one tap to the other all day and the voltage still varies within five volts or so. And what’s worse, folks are just ignoring the real problem: the large load elsewhere in the building, which is probably an unsafe condition as well. On the other hand, I have been at festivals where power was supplied on a very long line from the nearest substation, and there was no way to compensate for the voltage drop on short notice, and the little regulator was a huge help in dealing with the resulting problems.
A ferroresonant transformer is an isolation transformer that has a third winding. This third winding is connected to a capacitor to make a resonant element at 60 Hz in such a way that the field created slightly bucks the field produced by the primary. As the input voltage is decreased, the bucking field is decreased which increases the output voltage. This gives you excellent voltage regulation over a fairly wide range of input voltages, with none of the instantaneous changes of a tap-switching transformer.
There are a few problems with the ferroresonant transformer, though. The first is that all hell breaks loose if the power line frequency changes, because the device is a resonant system. This isn’t a problem running off of the standard mains in the US where the line frequency is pretty tightly controlled by the grid, but it means if you’re running off a generator you can have some very serious problems. This makes it a bad choice for use in some field applications where you might need to be using local generator power.
Another real problem is that for proper regulation the output load needs to be greater than 50% of the maximum rating. If it’s underloaded, it won’t regulate as well. This means you can’t just buy a huge and oversized one and expect it to be effective—you need to size it fairly carefully for the job. And you’ll either need to replace it or add additional ones if you expect to expand greatly.
Depending on the unit you buy it also may improve the power line waveform, or it may degrade it. Sola makes some “harmonically neutralized” units which have less than 2% line distortion on the output, which is pretty good. Since it’s an isolation transformer, you also get all of the standard benefits of isolation (see above).
In 1996 the National Electrical code added a paragraph to Article 530 allowing “motion picture and television studios and similar installations” to use AC power with two hot legs, each 60 volts from ground and 180˚ apart in phase. This gives you 120V between the two legs, but it means that if there is internal leakage on equipment between the power lines and ground, at best the leakage between the two legs cancels out, and at worst the leakage is half what it would be normally (since the difference between each leg and ground is half what it would be normally).
In 2002 this was removed from Article 530 to Article 647, so that this sort of “balanced power” installation can be installed anywhere. These systems involve an isolation transformer with a center tap that is tied to ground rather than one leg of the output being tied to ground. Therefore, in addition to reducing problems with the internal leakage on equipment, they also give you all of the benefits of an isolation transformer.
While balanced power installations are extremely useful for dealing with leakage problems, the amount of unjustified hype surrounding them is phenomenal. I have seen a lot of people installing balanced power to solve totally unrelated problems, and being very upset at the results. Recently I even saw someone installing balanced power for a synth rack because the manufacturer (who—of all people—should know better) convinced him it would reduce his noise floor, even though the dominant noise was hiss from the thermal noise of the electronics in the synths!
I know a lot of folks who have installed a small balanced power transformer in order to deal with gear that has clear chassis leakage problems, like guitar amplifiers and synth racks, brought into the studio by customers. I think, though, that doing this sort of thing on your own equipment is a poor idea; it doesn’t fix the real leakage problem, it just hides it. With stuff brought into the studio from outside it’s often impossible to fix real problems in the time allotted, and hiding it is a lot better than doing nothing.
I also know people who have installed large balanced power systems for their entire studio, since they needed isolation from line noise, and the cost of a balanced power transformer was no higher than that of a standard isolation transformer.
If you intend to install isolated power, ask your electrician what’s available. He may be able to use a standard dry transformer with a center tap rather than a more expensive transformer specifically designed for balanced power; while these transformers usually don’t have as good matching between the two legs they are often quite close. You should also be aware that, as far as I can tell, balanced power is not legal for installed services in Canada.
Online UPS systems
An online UPS system is a battery charger run off the AC line, which charges a large battery, and that battery runs an inverter which generates AC power. The output of the UPS is always drawn from the inverter, at all times, whether or not the power line is operating.
It used to be that even the best inverters produced a fairly poor-looking waveform, because the better the waveform, the less efficient the conversion is. Even the expensive higher-grade Litton online UPS systems produced a waveform that was worse than that of the typical power line, and which had more harmonics than the typical power line.
This meant that anything plugged into an online UPS was apt to have more noise problems and radio noise issues, and some online UPS systems might also dump noise onto the power line that they were plugged into. But the advantage was (and is) that you get constant power from the output at all times. This is very good for a computer system, but generally not good for other gear in the studio.
Today things are a lot better if you are willing to spend the money. There are online UPS systems out there that are “true sine wave devices” and have very low distortion specifications. They are substantially more expensive than the older and cheaper “modified sine wave” devices, but if you have to run analog audio gear off a UPS, they are a great solution.
The modified sine wave devices tend to be a source of more power problems than they actually fix, but the higher-grade units can do a very fine job of cleaning up power under almost all circumstances. The problem here is that these units cost substantially more money. They are almost certainly the most expensive possible power protection system—but they can also fix a wide variety of different problems at once.
Standby UPS systems
A standby UPS is a different creature altogether. It too has a battery charger and a battery and an inverter, but the inverter only runs when the power goes out. This means that you may have a period of several milliseconds between the actual power going out and the inverter firing up. This makes the standby system less effective for blackout protection, and some touchy computer systems may not like this.
But it also means that you don’t have any noise problems generated by the UPS except when the system is running on batteries. And when the system is running on batteries you probably aren’t recording—you’re probably just saving your files and shutting things down properly (at least you should be). The standby UPS gives you a chance to shut down and not lose any work if you are using a computer system or recorder that would lose data during power outages. The noise problem is generally a non-issue unless you absolutely have to run analog gear during the outage.
Most of the standby systems have a fairly poor sine wave output, with a good bit of distortion rather than a true sine wave, but in most cases this is not a real issue since you are operating off of the inverter only when the power is out.
In general, the standby UPS system will not do anything to improve the quality of the power line; it won’t clean up RF noise and it won’t regulate a line with unstable voltage. In fact, some of the standby UPS systems will have real problems with brownouts, where they begin switching back and forth rapidly between the inverter and line.
The strong exceptions to this are some standby UPS systems built with ferroresonant transformers in them, like the Best Ferrups unit. The ferroresonant transformer continues providing power for a few cycles after the power goes out, just from the residual magnetism in the core, and those few cycles are enough for the inverter to pick up the slack smoothly without any dropout. The ferroresonant transformer also regulates incoming power over a wide range, filters line noise, and smooths out the noisy signal coming from the inverter so that the output waveform has less distortion than with a conventional standby UPS system.
However, these units have the same disadvantages of other ferroresonant transformers; they still can add additional distortion if the power line is very clean, and they are still very sensitive to loading. Also, these UPS systems tend to be very expensive, although they tend to show up on the used-surplus market when your local internet startup runs out of venture capital and goes bust.
Don’t buy power line conditioning products unless you either have a problem now, or you will be working in a place where you foresee a future problem. Just adding more stuff to your power line at random is about as effective as adding stuff to your signal path at random; it just wastes money and adds clutter. But if you do have a problem, get a product that is designed to solve that particular problem—there are a lot of fine products out there intended to fix them.
Scott Dorsey spends a lot of his time clearing up misconceptions about everything from AC power to tube circuitry to analog tape formulations. He often does this for Recording, which we think is pretty cool.