15
\$\begingroup\$

I know a similar question title has been asked, but I believe that does not answer my question (and I couldn't think of a better way to phrase the question).

I'm a bit confused by how exactly an amp can overload a speaker, and vice versa.

Many guitar amplifier speakers are 8Ω impedance.

If I understand correctly, the output amplifier (should) outputs a fixed voltage signal output regardless of what load is placed on it. If this step is wrong, please correct me.

So if there is a fixed voltage signal, (say, +-15V, i.e. 30V of swing) and if the speakers' impedance is ~8Ω (I understand it will vary with frequency but say it is around this figure), then how is the wattage varying with different amp combos even though the impedance is roughly the same? Is it that the voltage increases with higher wattage amplifier/speaker combinations.

For example a 10W combo with an 8Ω speaker vs a 100W amplifier connected to a 4 speaker cabinet wired for 8Ω impedance (parallel connect 2 series pairs of 8Ω speakers), the 100W is obviously louder. Is it that the output voltages of the 100W amplifier is more? Else how can you get the wattage to increase if you are keeping voltage and impedance constant?

What would happen if you connected the 10W amplifier directly to a 4 speaker cabinet? Would it overload the amplifier? Or just play it quieter? Theoretically, if the voltage is the same and the impedance is still 8ohms, the wattage should be the same, i.e. 10W through the 100W rated speakers.

If so, is this true then: when we say 10W 8ohm speaker, we mean its able to handle maximum peak voltages of (P=V^2/R, V=sqrt(PR)) ~9V. Whereas for a 100W 8ohm speaker, it is able to handle peak voltages of ~28V?

In what situation can you harm a speaker? By connecting a too powerful an amplifier to it? But then isn't that something that is recommended by many people? (Amplifier output of atleast 2x the speaker rating). If so then the voltage output of an amplifier is not fixed? It varies according to which speaker is connected to it? (even though the impedance is the same?)

In what situation can you harm an amplifier? By connecting too high a wattage speaker to it? Then why do I see so many people posting videos on youtube of 1/2W guitar amplifier builds connected to big wattage rating 4 speaker stack speakers or atleast even 2 speaker combinations?

\$\endgroup\$
1
  • \$\begingroup\$ You are wrong at step 1. The voltage across the speaker is proportional to the voltage presented at the input jack by the guitar. It is merely amplified several times. It isn't fixed. As the speaker is a low impedance, the amplifier also has to be able to source the corresponding amount of current given by V/Z, and this is why it is technically a power amplifier. \$\endgroup\$
    – user207421
    Mar 10, 2016 at 21:02

5 Answers 5

15
\$\begingroup\$

You have many questions but I think you can understand it better with a single explanation. See that there are many myths around this subject. But it is also a matter of analog electronics.

Speakers are a Z load in your circuit that may vary its impedance in terms of frequency. Note that a speaker main goal is to maintain a stable and almost constant impedance in the frequency range that it was build to work on. This impedance is almost equal to the coil resistance. So, when your speaker is working in a system well designed, your Z load can be viewed as a almost pure resistive load (8, 6 or 4 ohms in most cases).

With that said, we should have ways to supply power to the speaker so it can reproduce sound waves. Note that the magnetic part of the speaker is directly related to the current that pass through it. So we can say that the speaker is a kind of resistive load that deals with current variations to produce sound (simple way to understand). So the way we can vary current in a resistive load is by swinging a voltage across it.

If you connect a speaker or a simple resistor into a amplifier's output and also plug an oscilloscope probe across the load, you will see the voltage variations just as your music is varying (sound waves). It's not a constant voltage in the output. Otherwise you cannot produce sound waves since you need current variations to produce magnetic variations and forces by Lorentz formula.

Besides that, wattage is the power consumed by your system. Instantaneous power is calculated by P = UI or P = ZI². So the greater the current passing through your speaker, more power it will dissipate (and also more power consumption since part of it will be transformed to sound waves).

Also, you have to consider the volume control. Those examples you gave can only be applied if your amplifiers are always working at full amplification (0 dB). This way, a more powerful amplifier should produce higher voltages in the output compared to a less powerful amplifier (both in 0dB). Since instantaneous power is also calculated by P = U²/Z, then you cannot increase power with voltage and impedance being the same.

When you make connections (amplifier + speaker) you should care about some details:

  • Amplifier power output: it will tell you how much power it can deliver to your speaker in a determined impedance. This is the maximum power it can produce. Note that if you turn it on with 20% of volume, it will not be delivering its full power. Note also that even in 0dB it probably wont be producing full power all the time because music varies its amplitude waves so you should calculate the average power by the integral of all the signal.

  • Amplifier minimum impedance: This will tell you what is the lowest impedance you can connect to its output. It does not matter if you connect higher impedances there. You will just not be able to obtain too loud sound in your speaker system. Generally speaking, when connecting higher impedance speakers you can have a cleaner sound (less distortion) but a lower sound volume. In the other hand, if you want a louder system, you should connect the lowest impedance allowed but you probably will have more distortion. Note that what may harm any part of your system is excess heat. And heat is produced by Joule effect which relates to power directly. So, it's also possible to connect lower impedances than allowed since you do not increase volume more than a certain point. This way, even with lower impedances you produce same power as a higher impedance in full volume. You can see that by connecting a 2Ohms speaker to a 4Ohms-minimum amplifier but in a very low volume. It will work and it will not harm anything.

  • Speaker impedance: as already said, its the nominal impedance that a manufacturer try to reach and maintain stable in the frequency range that the speaker is designed to work.

  • Speaker power: this is the highest power the speaker is built to tolerate. Of course there are always questions about the ways people use to measure that and indeed there are misconceptions about the terms like RMS POWER. A common way to do that is to connect the speaker to some signal which has a AVERAGE power P and see if it can tolerate that for a long period of time. The greatest P value you can reach doing that is your nominal average power (again, it's a simple way to explain).

So if you are connecting a speaker to an amplifier, you should watch those variables to see if you will harm anything. Generally, you can harm a speaker when connecting a too powerful amplifier to it. Let's say you have a 300 W/8 ohms speaker and you connect a 800 W/8 ohms amplifier. As I said before, it also depends on the volume dial. Whenever this system is at low volume, nothing will harm. But when you reach a specific point of volume that the average power in the output will go over 300W, you probably will start to harm your speaker. People also sometimes say that a very powerful speaker could harm a non-powerful amplifier. Or that a non-powerful amplifier cannot drive a powerful speaker. What happens is that you can have now a 20W/4 ohms amplifier with a 800W/4 ohms speaker. Note that you can connect them and it will work normally. This will be just like connecting a more powerful amplifier with low volume to it. The problems are: you will probably want to reach full volume to have some sound. THIS could harm your amplifier since full volume many times means more than 0dB (plus distortion). The excess heat in the amplifier may damage its output. Another common problem is that this distortion at full volume may damage your speaker. This happens because the speaker is build to work in movement. Many speakers have holes to dissipate heat and obtain air flow to refrigerate. Whenever distortion occurs, the mobile part of the speaker may stop moving for a little while. It starts to overheat the coil.

In short, any combination of amplifier and speaker should be possible. You just have to take care of the volume. If you don't want any possible trouble, get a amplifier which is a little less powerful than your speaker in the same impedance, and never exceed something like 70%~80% of the volume control. If your volume dial has a dB scale, try using in 0dB at most.

I hope this has cleared your questions. Sorry for bad English.

\$\endgroup\$
3
  • \$\begingroup\$ I understood everything you said clearly, so don't say sorry! When I said amplifier, actually I was talking about it from the designers perspective, so when you say 0dB, what level is that? I've seen schematics for various simple amplifiers and they usually have a potentiometer at the final output for the volume control. Where is 0dB on this potentiometer? Is there a better way to control volume than this? \$\endgroup\$ Nov 10, 2013 at 21:02
  • \$\begingroup\$ When I say 0dB I mean 100% amplification. See that a "pure amplificator" will amplify your signal X times and that's it. But we don't want to hear sound at full volume all the time so they add a volume control in the final stages so you can attenuate the amplification. When your dial has a dB scale, it means how much you are attenuating your signal compared to full amplification. So, at 0dB you are using its full capacity of power. This would be like a potentiometer in its minimum scale (zero ohms). \$\endgroup\$ Nov 10, 2013 at 21:07
  • \$\begingroup\$ Also, it depends how you use your potentiometer to control volume. It can be added just in the final stages as a variable resistor or it can control the amplification process (which is better). The first method is the simplest way but it's easy to understand that part of the power generated by the amplifier will be consumed by the potentiometer (wasted). So you kind of amplify and then "disamplify" your signal instead of just reduce the amplifying gain at first. \$\endgroup\$ Nov 10, 2013 at 21:09
5
\$\begingroup\$

Matching impedances can be a problem with both solid state and tube amplifiers.

In the case of tube amps, the tubes cannot directly drive the speakers; they have to drive the speakers through an impedance matching transformer. It's pretty hard to damage the tubes, but the transformer or speakers can be damaged if the impedance is not matched. In a tube amp, the tubes are good at driving large voltages (100s of volts) but not good at driving large currents. So in order to drive 8 ohm or 4 ohm speakers, a transformer is needed to convert the high voltage output of the tubes into a high current output for the speakers. The primary side connected to the tubes has lots and lots of turns of very fine wire. The secondary side connected to the speakers has fewer turns of thicker wire. The tubes act as current sources. If no speaker or a speaker of too high an impedance is connected, the tubes can present the transformer with very high voltages that can damage the insulation on the transformer windings. If the speaker impedance is too low, the tubes can push excess current through the windings, causing them to heat up. Neither of these is ideal. Generally, the transformer's secondary will have 2 or 3 taps for common speaker impedances to make matching as simple as selecting the right impedance on a switch.

In the case of solid state amplifiers, you can have a similar problem with an unloaded amplifier damaging itself by generating high voltages internally. The cause is the same: the output transistors act as current sources, and if the impedance is too high it will result in high voltages. Modern amplifiers are generally designed either to avoid this problem entirely, or they have internal loads that are permanently connected across the output terminals to put an upper limit on the impedance that the amplifier sees.

In terms of amplifier power output, most amplifiers actually have 3 output limits - voltage, current, and power. If the impedance is small, you hit the current limit. If the impedance is too large, you hit the voltage limit. If you pick just the right size impedance to hit both the current and voltage limits at the same time, you will probably hit the power limit. The voltage limit is determined by the amplifier's supply voltages. The current limit is determined by the output drive transistors. And the power limit is generally a thermal limit - if you exceed it for too long, the amplifier will overheat.

You can harm a speaker in several ways. One is putting too much power through it. Another is putting too much power through it on frequencies outside of its design frequency range. e.g. don't put bass through a tweeter. Another is amplifier clipping. When the amplifier voltage or current limits are reached, it cuts off the top of the waveform, generating lots of high frequency harmonics. These can damage a speaker by violently jerking the speaker cone around at frequencies it's not designed to operate at. Also, if the clipping is not symmetrical, the cone can inch its way into or out of the speaker. If it gets far enough out, the coil will leave the groove in the speaker magnet and it can be damaged if it misses the groove when it comes back.

You can harm an amplifier by overloading it or underloading it, impedance wise. There is no problem with connecting a 4W speaker to a 1/2 W amplifier so long as amplifier is fine with driving the speaker's impedance. It just won't be very loud.

\$\endgroup\$
3
  • \$\begingroup\$ I wonder to what extent the "sound" of tube amplifiers is a function of their higher output impedance and its interaction with frequency-dependent speaker impedance? \$\endgroup\$
    – supercat
    Dec 26, 2013 at 19:12
  • \$\begingroup\$ There are tubed amplifiers without output transformers. They (at least most of them) use output stages with a (fairly) large number of tubes in parallel to provide more output current. A search for something like "OTL tube amplifier" should turn up more details for those who care. \$\endgroup\$ Jun 10, 2020 at 15:55
  • \$\begingroup\$ @supercat: that's almost certainly at least part of the equation, anyway. In particular, the "tighter" bass of a solid state amp is almost certainly due in large part to the lower output impedance. \$\endgroup\$ Jun 10, 2020 at 15:57
5
\$\begingroup\$

First of all, it's pretty rare for a speaker's impedance to be anywhere close to flat. The impedance curve normally looks vaguely like this:

enter image description here

The peak is fs, the speaker's free-air resonance. The rated impedance is the first minimum in the impedance curve above resonance. The DC resistance will usually be a little lower than that, but not typically a lot lower (e.g., might be around 6 ohms for a speaker rated at 8 ohms impedance). The DC resistance is also affected by other factors though--for example, a speaker intended to handle more power will typically have thicker-gauge wire in the voice coil which will reduce the DC resistance, but have almost no effect on impedance at higher frequencies.

When you mount that driver in a box, you typically add at least one (and often a couple) more, smaller peaks at lower frequencies that reflect the resonant frequency of the cabinet and any ports it may have.

I'm not sure where you got the idea that voltage is constant (or even close to it). Like any other circuit, P = I * E. So, for example, one watt through an 8-ohm speaker is 2.83 volts (square root of 8, since P = E2/R). Perhaps you're thinking of the fact that most amps will be rated for a maximum voltage swing (but it'll usually be higher than 16 volts).

As to what happens if you connect a 10 Watt amp to 4 speakers (presumably in series-parallel to maintain the same impedance), you'll typically gain at least a little efficiency, because most speakers are at least somewhat non-linear. For example, a speaker might be rated at 92 dB SPL at one watt (under some standard testing conditions). In theory, that means it should produce 95 dB SPL with 2 watts of input, or 102 dB SPL with 10 watts of input. In reality, three or ten more dB of input won't usually produce (quite) three or ten more dB of output though. By separating the power from the amplifier into four separate speakers instead of one, you'll minimize this effect, so you'll get (slightly) more acoustic output for a given amount of electrical output from the amp.

As far as too powerful an amp damaging a speaker: it depends. If you completely overpower a speaker, yes, that can happen. For example, if you connected a 500 watt amplifier to a little 3 inch speaker and just cranked it up to anywhere close to maximum power, the speaker would almost inevitably fail fairly quickly. Depending on the design, it's a little hard to be certain what would fail first -- you might overheat the voice coil, and a wire would simply vaporize, or you might generate a stronger magnetic field than it's designed for, and push/pull the speaker cone further than intended and destroy the surround (in my experience, voice coil failure is a lot more common though).

Much more common is destroying a speaker by driving an amp past its rated power. This is particularly problematic with bipolar amplifiers, since they tend to have quite harsh clipping characteristics. Here, however, you're saved by the fact that intentionally producing various forms of distortion is pretty common, so when you're dealing specifically with a guitar amp and speaker, you're not quite so likely to destroy things (very quickly anyway). With something like a normal stereo, clipping will typically increase high-frequencies in the signal a lot in a hurry -- that, in turn will result in a lot more of the power going to the tweeter than intended, which may destroy it very quickly.

Hurting the amplifier depends. The short summary is that failure in a solid state amp will typically happen if you connect too low impedance of a speaker. That will try to draw more current than the amp can deliver, leading to overheating and (if you go too far) melting down the output transistors.

Conversely, tubed amps tend to be damaged more often by connecting too high impedance of speakers. The amp is designed for the speaker to load the output. Without enough loading from the speaker, the amp will produce higher voltage than intended. When/if a speaker wire comes loose, you effectively get infinite impedance almost instantly. Depending on the design, either your protection circuit kicks in and shuts down the amp, or else the last sound you hear before repairing the amp is a loud pop as the output tubes burn out.

\$\endgroup\$
2
  • \$\begingroup\$ The last paragraph is an excellent explanation as to why you don't EVER want to hook your 250 ohm headphones directly into your tube Amp's speaker output. \$\endgroup\$ May 8, 2014 at 16:00
  • \$\begingroup\$ The fact that an amplifier wouldn't want to be open-circuited would suggest that its output impedance is very high; most amplifiers nowadays are designed to have a very low impedance. I would expect that a high-impedance amp driving the above speaker would feed it maximum power at frequency fs, while a low-impedance amp would feed minimum power then. How much power should be fed to a typical speaker at fs to yield a sound power level comparable to that at other frequencies? \$\endgroup\$
    – supercat
    Dec 29, 2014 at 19:55
2
\$\begingroup\$

The specifications for speakers are a bit of a mine-field but for amps they are simpler. If an amplifier is rated at 10W RMS then that is the sinusoidal power it can deliver to a specified load (2 ohms to 8 ohms usually) at a certain distortion level. Usually the distortion is because the amplifier is delivering the sinewave at the onset of clipping.

So, if it has +/-10 V internal power rails it will just about be able to deliver 17.9 Vp-p with some amount of small clipping into an 8 ohm load. The same amplifier may also be able to drive a 4 ohm load with just about the same output amplitude and in this case the amplifier may specify that it is a 20W amplifier.

An amplifier will tend to have a very low output impedance and this is generally the case for transistor amps using negative feedback - the feedback tends to keep the output constant irrespective of the load. However, there will be a point (if the load impedance is reduced) that the amplifier smokes or a current limit circuit kicks-in to "save" the amplifier from destruction.

For a speaker, it will have a rating and hopefully this rating will be in the same form of units that a power amp is specified in but this needn't be the case and you do have to ensure you are comparing apples with apples. A speaker's rating will also include the frequency response that it is rated at and this is important to note because you can't push bass (at the speaker's rated power) into a tweater and expect it to survive and neither can you pump deep sub-bass into a standard bass driver and expect it to survive.

\$\endgroup\$
2
\$\begingroup\$

There are some really good answers here so there is only a little for me to add because things are pretty well covered by everybody else .Class AB solid state audio amps are reasonably flexable about speaker impedence within the limits of current and volts already covered.Class D is a different story because there is usually a lowpass filter that has a cut frequency above the highest audio frequency of interest and below the switching frequency .Example cut frequency 30KHz and switching frequency 150KHz .The filter will be designed to be nice and flat across the audio band .If you say run 16 ohm speakers on say a 4 ohm amp the filter could get peaky and it could sound terrible or even damage things if the filter is outside the feedback loop .If you are running class D do not muck around with the speaker impedences unless you really know what you are doing .

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.