19

I bought a powerful tweeter (or so I thought) to experiment with the ultrasound. The device is quite heavy, and it is rated at "350 watts RMS/700 watts peak": 2'' Titanium Super Tweeter PDBT78

Its nominal impedance is "4-8 ohms" with a frequency response of "2-25 kHz".

To make the long story short, the thing burned on me with a sine 20-kHz signal at average (real) power of 9.6 W. How is that possible?! The specs promise 350 W "RMS" (I think they mean average power), yet it burns at under 10 W?! Please review my setup below. Maybe I calculated the power incorrectly and came to a wrong conclusion. This is how I measured the voltage and current:

Measurement Diagram

The scope displayed waveforms for the voltage and the current and showed their numerical RMS values on the screen. When the speaker burned, the scope showed \$V_{RMS}=40\:\textrm{V}\$ and \$I_{RMS}=0.24\:\textrm{A}\$. Thus, I concluded the power was less than 9.6 W. — I don't know how much exactly, because I did not get to measure the phase shift (:

$$P=V_{RMS}I_{RMS}\cos{\phi}$$

Thank you for your interest in solving this puzzle.

PS: I have received a response from the manufacturer:

We test tweeter power using signal EIA-426-B, frequency range 3.5k-20kHZ, PDBT78 power can be 40W ... it has no problem for testing over 8 hours... The 350W RMS is data on giftbox, it is not true data.

This confirms the community answers.

SlowBot
  • 191
  • 1
  • 6
  • Comments are not for extended discussion; this conversation has been [moved to chat](https://chat.stackexchange.com/rooms/119591/discussion-on-question-by-slowbot-why-has-my-tweeter-speaker-burned-up). – Voltage Spike Feb 10 '21 at 16:49
  • 2
    With a setup like this, you can burn up your ears as well. Near-audible ultrasound is just as bad for ears as an audible sound with similar power. It may be even worse because loud audible sounds are unpleasasnt and make you plan your next experiments better. – fraxinus Feb 12 '21 at 09:22

5 Answers5

35

Your linked description shows that the tweeter is intended to be used with a woofer for music reproduction. The description doesn't cover use of the tweeter as a specialist ultrasonic transducer on its own. The power rating indicates that the tweeter is intended to be used with a 350 W average/700 W peak audio amplifier driving both tweeter and woofer. For normal music or speech the average level is 10 dB below the peak level, bringing the peak level down to 70 W, much of which is feeding the woofer. In order to ensure the bulk of the power actually does go to the woofer the tweeter impedance is, as you found, much higher than 8 Ω. The 4-8 Ω rating in the specification indicates that the tweeter is intended to be used with a 4-8 Ω woofer, not that the tweeter itself has a 4-8 Ω impedance.

The Pyle description is written to be easy to understand for someone with modest technical knowledge putting together a car audio or other speaker system. The description tells them that if they have a 350 W amplifier and a 4-8 Ω woofer then this tweeter is suitable to use with them. It doesn't mean that you can put 350 W into just the tweeter.

Graham Nye
  • 3,560
  • 1
  • 11
  • 20
  • 9
    It turned out you can't even put 10 W in the device. I think this is an obvious case of false advertisement:. The manufacturer or seller never mention that woofer has to be connected in the description of the product.. This make the "power" number very poorly defined. Thank you for the explanation, though!
    What can be used as inexpensive US radiator?     – SlowBot Feb 10 '21 at 00:02
  • 4
    I did not expect 300 W, but 10 was disappointing. I feel the average person who is buying these products is misled too. Look at the number of posts about failed speakers, replacement coils, etc. – SlowBot Feb 10 '21 at 00:20
  • 21
    The first paragraph of your linked description states this tweeter "will complete your current car woofer enclosure system for full-range stereo sound reproduction". The advert mentions "car" six times. Whilst I'm sorry you've burnt out your tweeter the advert seems entirely clear that it is talking about car audio use and not standalone transducer use. If you ask Pyle I expect they can give you technical data on standalone use. There's no point putting that into an advert aimed at car audio users as it would mean nothing to them. – Graham Nye Feb 10 '21 at 00:30
  • 5
    @SlowBot Is is still misleading when the numbers are so clearly ridiculous? There's no way even 40 W steady state could be pushed into such a device without it being thoroughly cooled (either with water or a big heatsink and a fan). Where would all that power go? Think of a 40 W lightbulb and how hot that thing gets how quickly. The coil of the speaker would have done the same. – TooTea Feb 10 '21 at 09:24
  • 2
    @GrahamNye I'm fairly confident that advertising a component this way would entitle SlowBot to a refund in Australia, since it wasn't Fit For Purpose. – nick012000 Feb 10 '21 at 12:01
  • 1
    @nick012000 it wouldn't entitle him to refund in Poland, since he wasn't using it for the purpose it was advertised for. It probably is fit for the advertised purpose of completing a car audio system. – Mołot Feb 10 '21 at 13:32
  • 8
    @SlowBot Putting 350W continuously at 20kHz through a tweeter *inside a car* would leave you clinically deaf within a few minutes (assuming you could withstand the pain level). At the rated sensitivity, that would be like standing close to a military jet engine with no ear protection. As the answer says, this is meant to be used as part of 350W *total* audio power. – alephzero Feb 10 '21 at 13:54
  • @SlowBot 10W is a lot even. For the smallest noticable sound difference you need twice the power. 10W or 40W isn't much of a difference. – Swedgin Feb 10 '21 at 14:08
  • 1
    @nick012000 I can't find a Pyle Audio Australian website. However amazon.com.au lists tweeters from a range of brands which all seem to be stating system power ratings rather than component ratings. None of them specify they are system ratings. Presumably Amazon.au is operating in line with Australian law. I assume the tweeter is suitable for the car audio use shown in the OP's Pyle Audio link. – Graham Nye Feb 10 '21 at 15:04
21

It appears that the problem lies with the power ratings of loudspeaker components being the power rating of the system and not that of the individual component.

For example, in the 1970s, Philips were specifying the ratings of individual speaker components.

enter image description here

In later years, it may have been felt that specifying the system power would make it easy to select individual components while configuring a system.

The statement in the following specifications confirms that.

'System powers tabulated below are for complete two or three-way systems making use of the listed loudspeakers; corresponding cross-over networks and recommended enclosure volumes are listed on pages 4 and 5.'

enter image description here

A pair of loudspeakers, configured by Philips years ago, had the following components:

Woofer - AD8061W - 30W

Mid-range - AD0210 - 20W

Tweeter - AD01610/T - 4W

It shows that the tweeter power is indeed quite low when compared to the system power.

vu2nan
  • 15,929
  • 1
  • 14
  • 42
11

I'm afraid that you are another Amazon victim. I note that you have linked to an Amazon ad rather than a manufacturer's datasheet which would explain how the device is rated and expected operating conditions. Our motto here is, "No datasheet? No sale!".

I expect that a similar device from a reputable manufacturer might give the amplifier rating at several hundred watts but when used with a passive speaker crossover setup which might filter out a lot of power.

"The specs promise 350 W "RMS"." That's usually a big clue that we're dealing with bad marketing hype. Power measurements don't use RMS as the "square" term is already built in to the power equation P = V2 / R.

Transistor
  • 168,990
  • 12
  • 186
  • 385
  • :I thought Pyle was a reputable company. I've written them, but so far no reply. But to "exaggerate" the max. power by 30 times? It is possible that the power rating is for a certain lower frequency in the range.

    Should I buy their "600 Watts" speaker, or that would be stupid? What do you think failed there, a wire in the coil?

     – SlowBot Feb 09 '21 at 23:09
  • 2
    @SlowBot They may be reputable, but they still need to operate in a world where all their competitors advertise hundreds of watts, so they wouldn't sell anything if they stuck to the physically correct ratings of like 4 W continuous. – TooTea Feb 10 '21 at 09:26
  • 7
    A correction to your last paragraph. "Power" certainly does need "RMS" specified, because it does not inherently cover the "root mean" part, and your statement for power calculation is only true for a resistor which a speaker isn't. "Instantaneous power" is very much a thing. Also in the audio world it's important to be clear what's RMS (hence actual power) and what's a dodgy marketing number like PMPO. Seeing RMS is actually the sign of a number you ***can*** reasonably expect to achieve. (Or as the OP has found, a repeatable reference value.) – Graham Feb 10 '21 at 18:49
  • Thanks, @Graham. Can you provide an example RMS power calculation? I don't have a problem with the concept of instantaneous power; it would be \$V \times I\$ at any instant. – Transistor Feb 10 '21 at 19:25
  • @Transistor: for instance, suppose a 10 ohm resistor is plugged into 110V power mains. What is the power dissipation? Because that 110V is a RMS voltage, we can calculate that easily using \$V^2/R\$ as 1210W. RMS power means power obtained from a RMS voltage or RMS current. – Kaz Feb 10 '21 at 21:04
  • @Transistor The peak voltage of that 110V is around 156V, right? But the votage swings positive and negative: the average voltage is therefore zero. We cannot use that average voltage to calculate a power of zero; it makes no sense, because the resistor doesn't care about polarity; it dissipates either way. So what we do is we square the voltage. This gives us a displaced sine wave at twice the frequency: \$sin^2(x) = 1/2 - 1/2cos(2x)\$. The average of this squared sine wave ("mean square") is the constant displacement term \$1/2\$. The root of this ("root mean square") is 0.707. – Kaz Feb 10 '21 at 21:08
  • @Transistor the final square gives us something back that is in voltage units again, which we can use for power calculations. Those power values are then called RMS, after the RMS voltage. – Kaz Feb 10 '21 at 21:11
  • @Kaz, thanks for the comment. I'm an electrical engineer for many decades so I'm familiar with the theory. My point is that we **don't** use the term RMS power. Have you ever seen a light bulb, toaster, electric drill or computer power supply with a rating quoted in "W RMS"? No, because power is measured and quoted in watts. If we need to differentiate between average power (watts) and the instantaneous maximum then we would use the term "peak power" or "peak instantaneous power" and the units are still watts. – Transistor Feb 10 '21 at 21:23
  • 4
    @Graham: But "RMS" is not the correct way to average power. You want simply "mean power". For linear circuits, there is a relationship between *mean* (not RMS) power and *RMS* voltage and *RMS* current. – Ben Voigt Feb 10 '21 at 22:01
  • 1
    Strictly speaking the term "RMS Power" is devoid of physical meaning. What quantity do you get from multiplying power by power? (Of course it can be calculated from the instantaneous power, to provide the RMS statistic.) – SlowBot Feb 11 '21 at 00:00
  • @Transistor In audio, RMS power is used a fair bit; it's not unusual at all. This is just informal jargon to make it clear that a proper figure is being cited, and not "peak power" or "musical power" or some other nonsense, whereby some 20W amplifier is made to look like a 100W one or whatever. Amps are not toasters; they are tricky. For short durations of time, an amplifier can produce a lot more than its continuous power with a steady reference signal. That is not the same thing as instantaneous power. RMS may not be the right way to describe honest, continuous power, but it's used that way. – Kaz Feb 11 '21 at 08:06
  • @Kaz, thanks again. "*RMS may not be the right way to describe honest, continuous power, but it's used that way.*" Correct (and the use is incorrect). – Transistor Feb 11 '21 at 08:37
  • You can think the American FTC for "RMS Power", they should have used the phrase "Average Power" or something of the sort. It was an attempt back in the 1970s to get hifi marketing power claims under control, trouble is they got the terminology very wrong. Trouble is it stuck, if you read it as "long term average power" you will not be too wide of the mark. – Dan Mills Feb 11 '21 at 12:13
  • 1
    @SlowBot *":I thought Pyle was a reputable company. I've written them, but so far no reply."* I'm an audio guy, mostly with pro-level stuff. Pyle is a cheap consumer brand and thinks/operates that way, sometimes with 19" rack ears glued on afterwards. They have their place, and I've used them before, but you often get what you pay for. – AaronD Feb 11 '21 at 19:43
  • "RMS Power" is computed from RMS voltage. That is, you feed a (typically 1 KHz) signal into an amplifier, and connect the output of the amplifier to a resistor with known resistance. You then measure the voltage the amplifier produces across that resistor, and compute the power as P = E^2 / R. For an RMS power rating, the voltage you feed into that is the RMS voltage. – Jerry Coffin Feb 12 '21 at 15:53
  • @JerryCoffin: No, that's how you calculate average or mean power. That's my point. "RMS power" is a makey-up term to mean continuous (average / mean) power in a music system. It shouldn't be used in electrical engineering. Strictly speaking the power reading should be the average / mean of the *VI* product to take into account any reactance in the load. – Transistor Feb 12 '21 at 16:02
  • @Transistor: You can use that for measuring a long term average or you can use it to measure a fraction of a wave. I'm not saying it's a term that really makes sense, only pointing out how they use it, and that their usage is unrelated to anything like sustained or continuous output as opposed to a short-term output. – Jerry Coffin Feb 12 '21 at 16:08
11

The power ratings for audio equipment are often completely bogus. Check out the amount of copper in a 350W motor (and note that a speaker is also essentially a motor, only the motion is linear instead of rotational):

enter image description here

There's nothing in the photo for size comparison, so I'll just note the diameter is about 5" (13 cm).

Guessing a power rating from a photo is akin to palm reading, but your speaker indeed looks like it could handle a couple of Watts, certainly not hundreds. I assume the 2" in the description refer to the size of the magnet, so the coil diameter is probably less than 1".

PS. Try to get your money back from the seller. There's nothing dishonest in claiming the speaker was defective, after all, you didn't exceed its maximum ratings, did you?

Dmitry Grigoryev
  • 25,576
  • 5
  • 45
  • 106
  • It might be an uphill battle to get a refund, since it wasn't in the manufacturer's intended use. But it WAS within the STATED ratings. So that might be an advantage. I wonder what would happen if lots of people did that? Effectively called out the "easy to match", pre-engineered specs that are spectacularly wrong in isolation. – AaronD Feb 11 '21 at 19:53
  • And +1 for consumer-accessible audio specs being almost useless. They can show you the math and the test methods, and that all lines up, but it's all designed *not* to be an indicator of actual use, but to make a bigger number that they can "truthfully" advertise. – AaronD Feb 11 '21 at 19:57
  • @AaronD The batlle might easier that it seems. Refusing to reimburse a defective product opens the seller/manufacturer to a lawsuit, and using an audio speaker for arbitrary sound generation (as opposed to playing music) falls under the category of "reasonably foreseeable use", unless the product manual specifically lists product use cases (check out your laptop charger for an example of such a warning). Unless the shop is flooded with money-back claims, they wouldn't risk it because they know their product description won't stand up in court. – Dmitry Grigoryev Feb 12 '21 at 10:48
  • But IS it defective? It's designed for a *total system* with X specs, and that's what it's advertised for. If it's not used within that context, they can claim that it's outside of the intended use, and so "the user exceeded the spec." For an even lower example of seller integrity, I actually had a high-rep eBay seller argue with me about their 8AWG jumper cables (I measured them) with super-thick insulation in a package labelled "4AWG" and sold as such. It went all the way to eBay's arbitration, which immediately reimbursed me, despite multiple protests from the seller to not go that far. – AaronD Feb 12 '21 at 14:22
  • @AaronD If the usage instruction states that the tweeter "is designed for a total system with X specs", then it's the user's fault for not reading it (they should have returned it under "product not as advertised" clause, but without breaking it first). If it simply says "350 watts RMS/700 watts peak", it is reasonable to assume the device is designed to withstand such power. Either the seller takes it back as defective, or they admit false advertisement. – Dmitry Grigoryev Feb 12 '21 at 15:53
  • @DmitryGrigoryev: I think even if the advertising copy could be read as claiming the tweeter itself was rated for 350W, this would fall into the same area as Pepsi's commercial years ago that said you could get a Harrier fighter jet for N million Pepsi Points. They got sued over that, and the judge decided that it was obvious to any reasonable person that the advertisement wasn't intended to be taken seriously/literally, so the fact that they wouldn't deliver it was *not* false advertising. – Jerry Coffin Feb 12 '21 at 16:47
  • Likewise here. In normal use, something like 1% of the overall power of the system goes to the tweeter. So, in normal use 350 watts going to the tweeter would mean you were feeding something like 35 kilowatts to the system as a whole. A typical home in the US has a 100 to 150 amp breaker panel, allowing the entire home to consume a maximum of about 18 kilowatts. So roughly speaking, you need all the power for two homes to drive a speaker to go with this mythical tweeter. – Jerry Coffin Feb 12 '21 at 16:59
  • @JerryCoffin Yes, a knowledgeable person would understand all of that, but a knowledgeable person would probably not buy this particular product either. :-) Likewise for Dmitry's comment about returning "not as advertised" without breaking it first. The knowledge to recognize that would also cause the person to buy something else anyway. – AaronD Feb 12 '21 at 17:51
5

Impedance

I'm going to start by addressing only one (admittedly rather minor) point raised in the question--that of the impedance.

The impedance of a loudspeaker isn't a single fixed number. It's generally a curve that will look at least vaguely like this (hand drawn, so don't try to read too much into the details):

enter image description here

That tall, narrow spike is at the speaker's free-air resonance (fs).

The rated impedance is then at the first minimum above the free air resonance (roughly where I've drawn in the red, sort-of vertical line).

You usually won't drive a tweeter at anything close to its free air resonance frequency though. Its free air resonance might be at, say, a couple hundred Hertz, but the crossover from midrange to tweeter might happen at, say, 3 KHz.

As such, it's not only reasonable, but entirely expected that for a tweeter, its impedance throughout its entire normal range of operating frequencies will be (often substantially) higher than its rated impedance.

Power

Getting back to the power situation, I'd agree with those who've pointed out that a normal tweeter would never be expected to handle a power input of tens of watts, not to mention hundreds of watts.

I'd add, however, that the amount of power a tweeter is expected to handle is rarely a function of the sounds its really intended to reproduce. Rather, it's almost entirely a matter of how much abuse the manufacturer expects the tweeter to be subjected to. When/if you drive a bipolar amplifier beyond its capabilities, it will clip. This will typically result in drastically increased power at higher frequencies. A FET- or tube-based amplifier will also clip, and when it does, the power and higher frequencies increases as well--but this happens much more...violently with a bipolar amplifier. Exceeding its power capabilities even a little bit can increase harmonic distortion from a bipolar amplifier a lot, whereas with a tubed amplifier you have a fairly broad range in which you can increase power output with gradually increasing distortion.

Reproducing normal music requires very little power at higher frequencies, so the power handling capability of a tweeter primarily reflects the manufacturer's estimation of how likely you are to drive your amplifier into clipping (and/or how much clipping they think you'll tolerate).

Jerry Coffin
  • 3,336
  • 17
  • 18
  • Sure, a product is not expected to have a good performance at its maximum power rating, but it is expected to withstand it. – Dmitry Grigoryev Feb 12 '21 at 10:59
  • @DmitryGrigoryev: Yes, but the question is whether that rating is for the tweeter itself, or a complete system in which the tweeter would be used. To anybody who starts out with even a general notion of what they're looking at, it's obviously the latter. The only part open to question is whether the advertising he read actually mis-characterized the rating, or he simply misread what was there. – Jerry Coffin Feb 12 '21 at 15:58
  • I suppose only the OP can check what is actually written in the manual/instructions/whatever document they got with the speaker. If it mentions what the power rating stands for, then sure. – Dmitry Grigoryev Feb 12 '21 at 16:07