Audio Amplifiers.

The FTC and Amplifier Power.

If you go to a nationally known electronics chain store and buy a 100 watt amplifier, are you really getting an amplifier that will deliver 100 watts of honest power? From about 1965 until the early 70s, I don't remember exactly when the FTC power rules were imposed, the answer would be no. From then until the early 80s, the Reagan era of deregulation, the answer would be yes. From the 80s up to the present the answer is once again no. Why? Go Figure, I shouldn't have to tell you.

Honest Power.

To an electrical engineer or a physicist power is power. 1 watt is 1 joule per second. Case closed. But starting in the 60s deceptive advertising so clouded the picture that it became necessary to refer to something called RMS power. To an electrical engineer this is totally redundant, like light daylight, or dark darkness. In AC, power is the RMS voltage times the RMS current but there really isn't any such thing as RMS power. In this discussion then, I will simply refer to power in watts. When I'm talking about dishonest power I will make it clear what I mean.

Dishonest Power.

Let's start out with a very low power amplifier, say, 15 watts per channel. Now the very first thing the writers of advertising copy are going to do is add the two channels together and call it a 30 watt amplifier. OK, so most buyers are going to ask the salesman "that's 15 watts per channel, right?" Thus cornered, he has no choice but to answer "yes".

Now the fun begins. The ad writers quizzed the EEs about power. "Is there anything higher that's still real?" The EEs being rather naive, mentioned peak power revealing that it is twice the real power. "OK, now our amplifier is a 60 watt power house.

But there is more. Unlike tube amplifiers there's no such thing as impedance matching in a transistor amp, and that's what we are talking about. The amplifier is basically a power supply that will deliver what ever amount of current the load asks for. So if you connect a 4 ohm load instead of the standard 8 ohms you can double the power again. In reality the power won't double because of sag in the DC power supply but nobody is going to worry about such a trivial matter. So now our original 15 watt per channel amplifier is being advertised as a 120 watt blaster.

Now the engineers realized that their salaries were tied to the financial well being of the company and became willing parties to the deception. "If we substitute a laboratory bench power supply for the stock ones in the amplifier we can get even more power." The marketing guys replied, "I don't understand that, just do it and tell me what you get." In most cases a power increase of about 30 percent can be obtained by substituting regulated power supplies and by boosting the voltage to the maximum the transistors will stand, as much as a 50 percent increase may be possible. "Possible or not, let's go with it." So what have we got? A 180 watt earth quake monster.

Enter the Federal Trade Commission.

With things so totally out of hand that nothing you read about amplifier power could be trusted the government stepped in. Some say this was good and others say it was bad. But good or bad, it was done.

  1. NO MORE ADDING CHANNELS. The rules said that specifications had to be stated in watts per channel.

  2. NO MORE PEAK POWER. Because the confusion about power was so pervasive the FTC felt compelled to make up something called RMS power. Meaning the RMS voltage across the load multiplied by the RMS current through the load. The stated specifications had to use the phrase "RMS power".

  3. LOAD IMPEDANCE MUST BE STATED. The load impedance at which the power was measured had to be clearly stated along with the power. If the power was specified at 4 ohms, or even 2 ohms, it had to be stated and the specified load had to be used when the measurement was made.

  4. NO MORE FIDDLING WITH THE POWER SUPPLY. The amplifier had to be tested in the same state it would be delivered to a customer. Everything had to be strictly stock, suping up the test version would no longer be allowed.

There were other tricks that were banned. For example the test model would most likely be removed from its cabinet and fans set up to blow a lot of air over the heat sinks. That would keep the poor little overworked transistors from overheating and burning out during the test. Sometimes they would even supplement the heat sinks to further increase the power output.

A very early trick was to test with only one channel driven. FTC rules said "We'll have no more of that. The test had to be made with both channels driven to the specified power.

The heat sinks had to be adequate to the task. Before the test, the amplifier had to be operated at 1/3 power, both channels, for 30 minutes. This is the highest dissipation condition for a class B amplifier. A hot amplifier delivers less power than a cold one so this rule ensured that the amplifier was tested under worst case conditions.

Frequency response had to be specified honestly as well. An amplifier that would deliver 60 watts at 1,000 hertz but 30 watts at 20,000 hertz could only be advertised as a 30 watt amplifier.

No More Rules.

These rules were withdrawn while "you know who" was in the Whitehouse. The forces of greed have won out and the old cautionary rule, "let the buyer beware" must once again be strictly followed. Just to show how bad things are now, not long ago I saw an ad for a pair of computer speakers. They weren't anything special. They were your standard computer speakers, about 8 inches tall, in plastic cases, and powered by a wall wart. They were advertised in large black bold letters as 350 watt speakers. So, you can't believe anything you read, or anything you are told in a high-fi store.

Buy By the Pound.

About the only way you can get a feel for how two amplifiers compare is to heft them. It is an immutable law of physics that you can't get high power through a small power transformer. If the heft isn't there, the power isn't there.

May God forbid two things. Melted down led sinkers poured into the bottom of the chassis, and switching power supplies.


This page last updated October 27, 2006.