# 50W Amp for 15W Speaker

I am building a portable speaker and I am looking for a decent amp for my two 15W Phillips speakers. It is tough to find an amp that is efficient and cheap so I was wondering if there is something wrong with buying an amp that is rated for 50W*2 while my speakers are only 15W*2. Thank you in advanced.

• Why do you want to destroy your speakers? Surely there are easier ways than buying a special-purpose amp? – WhatRoughBeast Jun 24 '16 at 22:22
• Are you suggesting I use a 15w amp or I do not use one at all? – Bawad Jun 24 '16 at 22:24
• What do you think will happen if you attempt to drive a 15 watt speaker with 50 watts? Just curious. – WhatRoughBeast Jun 24 '16 at 22:26
• @WhatRoughBeast that would be why you don't crank the amplifier to "10" to start with. 50W is 50W maximum. As answered below, you have plenty of room to dial back the output power of an amplifier in an audio application. – user2943160 Jun 25 '16 at 1:48
• Put a knob on that only goes to '4'. ;^) – Transistor Jun 25 '16 at 22:55

That has been a subject of debate for decades. Should the speakers be overrated or the amplifier?

1. If the speakers are overrated the amplifier might be strained at high volume levels and produce distortion (clipping) that not only sounds bad, it is bad for the speakers as they faithfully try to reproduce the distortion.
2. If the amplifier is overrated (but not by an outrageous amount like 10:1), then you get clean sound until the speakers reach their limit. The woofers may pop against their frame and the sound becomes distorted. This is a warning to back off the sound level or risk burning up a speaker.
3. Because there is seldom anything like a perfect match between speaker wattage ratings and that of the amplifier, the better match is to have an amplifier that is a little over rated, and be careful with the volume control. This way the amp is never strained or distorted, giving you clean sound up to the maximum volume the speaker can handle.
4. It is easy to burn up 15 watt speakers, but they are often very efficient and will be very loud at a safe maximum volume. A 50 watt amplifier will drive these speakers with ease and not distort from working too hard.

EDIT: For better mechanical range of the volume control, look for a gain adjust control on the amp or buy 2 RCA attenuators from most any stereo installation shop(car or home). Ones with -3db rating will cut signal level to 70.7%. Those with a -6db rating will cut sound level by 50%. Any of these options will allow you to use more of the mechanical 'rotation degrees' of the volume control before you reach the limit of the speakers. (To avoid confusion, I am referring to signal level, not power levels.)

• I wonder why amplifiers are not routinely constructed with a power scaling control (which could be implemented as a second pair of pots in series with the volume pots, or a multi-position switch, etc.)? It shouldn't be hard to design one that way, and it would encourage people to buy a bigger amp than they need now so they'd be ready if/when they upgrade their speakers. – supercat Jun 24 '16 at 23:02
• +1 for rationality. There is too much paranoia about matching amp and speaker ratings, and an under-powered amp is at least as likely (maybe more likely, by clipping) to damage a speaker as an over-powered one. – Brian Drummond Jun 24 '16 at 23:03
• Part of the issue is that no one single number can characterise the problem : 5-10W at the main LF resonance can cause excessive cone excursion, rattling and damage, while at other frequencies. short 50W bursts will merely heat the voice coil, and if they stop before adhesives burn or cones soften, there may be no harm done. – Brian Drummond Jun 25 '16 at 11:34
• @BrianDrummond: I wish speaker makers would publish multiple specs: (1) worst-case concentrated-frequency power output at which speaker will meet THD specs; (2) worst-case power output at which speaker will meet THD specs if no half-octave band exceeds #1; (3) worst-case power output that would not prevent speaker from meeting expected lifetime if applied for 15 seconds once a day; (4) worst-case power output that speaker could endure continuously while meeting expected lifetime. For amps, there should be specs for power meeting THD spec, and will-not-exceed power. – supercat Jun 25 '16 at 13:23
• @BrianDrummond: There is huge value in having the will-not-exceed power below #3 or, for units that may run unattended, below #4. Provided the above are met, there is value in having the amplifier's THD-meeting power being at least equal to that of the speakers'; depending upon one's audio taste, #1 or #2 might be more important for that. If an 8-ohm speaker sounds like garbage with anything over 10W, but can withstand any 30VRMS waveform one might throw at it (111W at 8 ohms), it should be safe to connect to a 100W amplifier, but a speaker which sounds great at 25W but can be... – supercat Jun 25 '16 at 13:30

There are many audio power amplifier products based on the Tripath TA2020 which is a 20W + 20W Class-T chip. This would seem to be an ideal solution for your product. Many of the most respected powered monitors available today use Class-T power amplifiers internally.

Well, if you absolutely had to use the 50W amp with 15W speakers, and wanted to be able to crank the amp to "10" without blowing them, then there is another option: add a series resistance (pseudo-pad) to the speaker.

If the speaker and amp are 8Ω, through which 50W can be delivered, then the volts must be: $$V = \sqrt{W \cdot R}$$ $$\sqrt{50W \cdot 8\Omega}$$ $$\sqrt{400}$$ $$V=20$$

To calculate the current (just for reference), $$I=\frac{E}{R}$$ $$\frac{20v}{8\Omega}$$ $$I=2.5A$$ Note already it should be apparent that if a 16Ω speaker were used instead, the amperage would be half, and thus the power delivered would be half also.

To spec the exact resistance needed to reduce a 50W amp to 15W output, one could setup an equality (using the inverse of resistance) and cross-multiply: $$\frac{50W}{8\Omega^{-1}}=\frac{15W}{x^{-1}}$$ $$50x^{-1} = 8^{-1}\cdot 15$$ $$50x^{-1} = 0.125\cdot 15$$ $$50x^{-1} = 1.875$$ $$\frac{50x^{-1}}{50} = \frac{1.875}{50}$$ $$x^{-1} = 0.0375$$ $$x = 26.667\Omega$$

Now subtract the 8Ω and that gives $R_{PAD}=18.667\Omega$. Add that to the speaker and the risk of "blowing it" by over-driving from this amp should be mitigated. Since in a worst-case scenario up to 15W could be dissipated by Rpad, choose high-wattage resistor(s); carbon, metal film, or other non-inductive types should be used. Preferably several in parallel, to spread out any heat.

Note that this does change the impedance as seen by the amplifier. For most small audio amps, a higher resistance is usually not a problem. (A class-D amp which depends on a low-value load being there may be a whole other story worthy of an "L-Pad", but that depends on the amp.)

• Of course it changes the impedance seen by the amplifier. Otherwise how could it affect the current? The issue is whether the impedance change matters., which in an SS amplifier it almost certainly doesn't. – user207421 Jun 25 '16 at 1:59
• What is wrong with proper use of the volume control? Gain setting or RCA attenuators can bring the mechanical range of the volume control in range of the speaker limits. Thanks for the math. – Sparky256 Jun 25 '16 at 4:23
• The problem with adding series resistance to the speaker is that it ruins the ability of the amp to control the speaker. The speakers are highly inductive and require very low impedance amplifier output to drive them properly. That's one of the reasons for using heavy speaker cable - low impedance. – Transistor Jun 25 '16 at 23:00
• @Sparky256: I think the problem with signal level attenuators is that they only limit signals of correct amplitude. If someone plugs in a device with a boosted output - headphone level rather than line level, for example, it may still overdrive. To protect properly needs a cut-out or compressor but that's overkill for most applications - although I've considered it for PA in a youth club where all sorts of devices could be plugged in. – Transistor Jun 25 '16 at 23:04
• @transistor. I agree. The best way to protect any transducer is for the driver to have preset limits via feedback loops. For the audio/PA industry they expect the end user to know when an amp or speaker has reached its limits. Most do not, as I made a good living in the 1970's and 1980's repairing car and home stereo's. The introduction of CD's and their wide dynamic range led to more blown speakers and amps. To this day, for audio equipment, we are on our own. – Sparky256 Jun 26 '16 at 2:58