I broke my hi-fi system and my hi-fi speakers.
May 6, 2012 11:22 AM Subscribe
I replaced old 6 ohm woofers with new 8 ohm ones. Now, my amplifier overheats. Can someone here explain the science of what I have done wrong?
My Sony amp goes into "Protect" mode when my music gets way too loud. But only when I use the speaker in which I replaced the rotteing-foam 6 ohm woofer with a brand new 8 ohm one. I am just wondering, scientifically, what I've done wrong. Is the amp detecting that it has to drive/load one speaker a lot more forcefully than the other one, and then giving up? Or does putting an 8 ohm woofer where a 6 ohm once was just cause overheating in general?
P.S. I don't have the smarts to redesign the crossover, nor the money to just buy a new pair of speakers. Thanks MetaFilter, any hint would help.
My Sony amp goes into "Protect" mode when my music gets way too loud. But only when I use the speaker in which I replaced the rotteing-foam 6 ohm woofer with a brand new 8 ohm one. I am just wondering, scientifically, what I've done wrong. Is the amp detecting that it has to drive/load one speaker a lot more forcefully than the other one, and then giving up? Or does putting an 8 ohm woofer where a 6 ohm once was just cause overheating in general?
P.S. I don't have the smarts to redesign the crossover, nor the money to just buy a new pair of speakers. Thanks MetaFilter, any hint would help.
Response by poster: I am almost feeling CRAZY enough to install another 8-ohm woofer into another speaker where a foamrotted 6-ohm woofer used to be (((yes, I have several))) just to weight the L and R speakers evenly to see if it's uneven Left-Right that's tripping the Sony amplifier Protection mode, or, as you say, tweeter pulling too much power. http://i.imgur.com/oATY5.jpg
posted by shipbreaker at 11:49 AM on May 6, 2012
posted by shipbreaker at 11:49 AM on May 6, 2012
token-ring has the gist of it. You've introduced an impedance mis-match and your amp is tripping over it. Basically, the amp it trying to drive the remaining 6ohm components and is over-driving the 8ohm woofers. The amp goes into "protect" in order to not damage the woofers.
You should have simply replaced the old woofers with another set of 6ohm drivers matching the old ones. There really isn't an easy way around this.
posted by Thorzdad at 11:51 AM on May 6, 2012
You should have simply replaced the old woofers with another set of 6ohm drivers matching the old ones. There really isn't an easy way around this.
posted by Thorzdad at 11:51 AM on May 6, 2012
Also, you probably haven't borked the amp. My Sony amp got tetchy about connecting speakers improperly and went into protect mode repeatedly, but once I got everything sorted the amp worked fine.
posted by LionIndex at 11:54 AM on May 6, 2012
posted by LionIndex at 11:54 AM on May 6, 2012
Response by poster: An 8ohm woofer is _harder__ to
drive than a 6ohm woofer, so the tweeters are not the components that need protection.
I may yet go through with putting another 8 ohm woofer into a speaker that used to have a 6 ohm, just to see if it's the Left - Right imbalance causing the Sony amp to go into Protection mode.
posted by shipbreaker at 12:14 PM on May 6, 2012
drive than a 6ohm woofer, so the tweeters are not the components that need protection.
I may yet go through with putting another 8 ohm woofer into a speaker that used to have a 6 ohm, just to see if it's the Left - Right imbalance causing the Sony amp to go into Protection mode.
posted by shipbreaker at 12:14 PM on May 6, 2012
Response by poster: The woofers I am installing are goldwood GW-8PC/8 , they have a frequency response of 25Hz -- 2800Hz, SPL 91dB, Impedance 8 ohm and they handle 330 Watts of maximum power. I have no fear that I will blow the woofers. Instead I worry about the amplifier.
posted by shipbreaker at 12:17 PM on May 6, 2012
posted by shipbreaker at 12:17 PM on May 6, 2012
An 8ohm woofer is _harder__ to
drive than a 6ohm woofer, so the tweeters are not the components that need protection.
Right. My bad.
So, then, the amp is over-driving the tweeters in order to drive the woofer. Either way, it's a mis-match. Sticking another 8ohn woofer in the box isn't going to correct the situation. You need a woofer that matches the rest of the 6ohm system.
posted by Thorzdad at 12:19 PM on May 6, 2012
drive than a 6ohm woofer, so the tweeters are not the components that need protection.
Right. My bad.
So, then, the amp is over-driving the tweeters in order to drive the woofer. Either way, it's a mis-match. Sticking another 8ohn woofer in the box isn't going to correct the situation. You need a woofer that matches the rest of the 6ohm system.
posted by Thorzdad at 12:19 PM on May 6, 2012
You could use a 4ohm dummy load in parallel to get you back to 6 ohms.
posted by sanka at 12:35 PM on May 6, 2012 [1 favorite]
posted by sanka at 12:35 PM on May 6, 2012 [1 favorite]
I doubt the receiver is overloading because of the tweeter. They don't take much power. But what I think the receiver is seeing is the mismatched load (6 and 8 ohm components on the same circuit) and having a freakout. Because the resistance is variable over the frequency spectrum, it might be seeing excess load (even if it isn't really overloading).
Wait, did you say that the L and R speakers aren't the same? This could very well be the cause.
posted by gjc at 2:00 PM on May 6, 2012
Wait, did you say that the L and R speakers aren't the same? This could very well be the cause.
posted by gjc at 2:00 PM on May 6, 2012
Response by poster: Yes, and as I mentioned, I am tempted to "rig" another speaker ((Infinity Reference Four)) with an 8 ohm woofer ((((because the 6 ohm woofers lost all their foam)))) to test that theory!!
My test is to turn Pink Floyd WISH YOU WERE HERE up to MAX volume, then press Play. Usually it's at the five minute mark when all the hyperloud stuff starts that my Sony amp goes into Protection mode.
posted by shipbreaker at 2:05 PM on May 6, 2012
My test is to turn Pink Floyd WISH YOU WERE HERE up to MAX volume, then press Play. Usually it's at the five minute mark when all the hyperloud stuff starts that my Sony amp goes into Protection mode.
posted by shipbreaker at 2:05 PM on May 6, 2012
I am guessing here, but I'd be surprised if the mismatch between left and right has anything to do with it. I also don't think it's the 8 ohms that matter, but rather that the sub has a much lower (or non-resistive) impedance than that in certain frequency intervals where you drive it. Are you sure the sub is not meant to be driven from a special low-pass filtered output?
posted by springload at 2:09 PM on May 6, 2012
posted by springload at 2:09 PM on May 6, 2012
Not to derail, but if you replaced the woofers because the foam edges were rotten, another option might be to replace the foam rather than the entire driver. This place sells DIY kits to do so -- I've used them to repair home audio woofers with good results.
posted by zombiedance at 2:18 PM on May 6, 2012
posted by zombiedance at 2:18 PM on May 6, 2012
Best answer: Wait wait: Did you just swap the bass element in your existing cabinet for a different kind? I bet this has screwed up the cross-over filtering. The loudspeaker as a whole routes high frequencies to the smaller element and low frequencies to the bigger one. In both extremes only one element is active, and the amplifier sees that element's impedance. Around the crossover frequency, both elements work at half power, ideally keeping the total impedance close to the nominal one. If you change the filtering characteristics, however, you may have an extended crossover regime where both elements work at the same time, drawing more power than the amp is made to deliver. Read the wikipedia page on Audio crossover.
posted by springload at 2:34 PM on May 6, 2012
posted by springload at 2:34 PM on May 6, 2012
Best answer: You could use a 4ohm dummy load in parallel to get you back to 6 ohms.
No you couldn't. Don't do this. The 4 ohm dummy load will absorb twice as much power as the 8 ohm speaker, and the overall impedance of the parallel combination will be 1/(1/4 + 1/8) ohms = 2.7 ohms, not 6.
posted by flabdablet at 5:33 PM on May 6, 2012
No you couldn't. Don't do this. The 4 ohm dummy load will absorb twice as much power as the 8 ohm speaker, and the overall impedance of the parallel combination will be 1/(1/4 + 1/8) ohms = 2.7 ohms, not 6.
posted by flabdablet at 5:33 PM on May 6, 2012
Depending on the particular drivers you have, you might be able to find replacement foam surrounds and simply replace the rotten ones. I did this with a set of op-shop speakers and got quite a good result for very little money. It's fiddly, but it worked.
posted by flabdablet at 6:26 PM on May 6, 2012
posted by flabdablet at 6:26 PM on May 6, 2012
> You could use a 4ohm dummy load in parallel to get you back to 6 ohms.
No you couldn't. Don't do this. The 4 ohm dummy load will absorb twice as much power as the 8 ohm speaker, and the overall impedance of the parallel combination will be 1/(1/4 + 1/8) ohms = 2.7 ohms, not 6.
But (thanks to flabdablet for providing the formula) you *could* use a 24 ohm resistor in parallel with your 8 ohm speaker to give an effective 6 ohm load. But that ignores all frequency-dependent loading differences of course.
posted by anadem at 6:58 PM on May 6, 2012
No you couldn't. Don't do this. The 4 ohm dummy load will absorb twice as much power as the 8 ohm speaker, and the overall impedance of the parallel combination will be 1/(1/4 + 1/8) ohms = 2.7 ohms, not 6.
But (thanks to flabdablet for providing the formula) you *could* use a 24 ohm resistor in parallel with your 8 ohm speaker to give an effective 6 ohm load. But that ignores all frequency-dependent loading differences of course.
posted by anadem at 6:58 PM on May 6, 2012
No you couldn't; don't do that, part 2:
Moving-coil speaker drivers are primarily inductive rather than resistive. As such, their impedance is heavily frequency-dependent; the "8 ohm" rating is purely nominal, and about all you can tell from it is that the speaker probably has roughly twice the impedance of one rated at 4 ohms for any given frequency, all other things being equal; but in practice all other things never are equal.
Most speaker cabinets contain multiple drivers, and if you have a speaker cabinet containing multiple drivers that only has two terminals to connect the cabinet to an amplifier, those drivers will be fed from a passive crossover. The crossover is a bunch of inductors, capacitors and resistors whose job it is to split the available amplifier signal across the various drivers in such a way that each driver is fed only those frequencies it's designed to reproduce well.
Because the speaker drivers are themselves inductive, they interact with the crossover in frequency-dependent ways. The crossover in any given speaker cabinet will be designed to work hand-in-hand with the drivers the cabinet was designed for in order to achieve a reasonably flat frequency response across the audio spectrum. Substituting a different driver - especially one with a different nominal impedance - will alter the frequency response of the crossover output connected to that driver. There's no general rule describing exactly what driver substitution will do to the frequency response of a given crossover, as the results depend too sensitively on the exact components involved. If the new driver has a different free-air resonant frequency rating from the old one, the effects of that difference could easily outweigh a nominal impedance increase of 2 ohms.
It's quite likely that what's happened in the OP's case is that the combination of the new driver and the old crossover has resulted in the cabinet as a whole having a lower impedance in some parts of its bass frequency range than it used to with the old driver, causing the amp to overload if the signal contains significant content within those frequencies.
Basically, the amp is going into overload because it sees itself driving something that looks unacceptably close to a short circuit already, and adding ohmic resistance in parallel will make things worse, not better.
Finding an equivalent replacement woofer, or repairing the surround on the broken one, are really the only reasonable choices. You could redesign the crossover given sufficient engineering expertise, but that would likely cost much more than replacing the woofer, and would in any case result in asymmetry and a muddied stereo image.
posted by flabdablet at 4:50 AM on May 7, 2012 [3 favorites]
Moving-coil speaker drivers are primarily inductive rather than resistive. As such, their impedance is heavily frequency-dependent; the "8 ohm" rating is purely nominal, and about all you can tell from it is that the speaker probably has roughly twice the impedance of one rated at 4 ohms for any given frequency, all other things being equal; but in practice all other things never are equal.
Most speaker cabinets contain multiple drivers, and if you have a speaker cabinet containing multiple drivers that only has two terminals to connect the cabinet to an amplifier, those drivers will be fed from a passive crossover. The crossover is a bunch of inductors, capacitors and resistors whose job it is to split the available amplifier signal across the various drivers in such a way that each driver is fed only those frequencies it's designed to reproduce well.
Because the speaker drivers are themselves inductive, they interact with the crossover in frequency-dependent ways. The crossover in any given speaker cabinet will be designed to work hand-in-hand with the drivers the cabinet was designed for in order to achieve a reasonably flat frequency response across the audio spectrum. Substituting a different driver - especially one with a different nominal impedance - will alter the frequency response of the crossover output connected to that driver. There's no general rule describing exactly what driver substitution will do to the frequency response of a given crossover, as the results depend too sensitively on the exact components involved. If the new driver has a different free-air resonant frequency rating from the old one, the effects of that difference could easily outweigh a nominal impedance increase of 2 ohms.
It's quite likely that what's happened in the OP's case is that the combination of the new driver and the old crossover has resulted in the cabinet as a whole having a lower impedance in some parts of its bass frequency range than it used to with the old driver, causing the amp to overload if the signal contains significant content within those frequencies.
Basically, the amp is going into overload because it sees itself driving something that looks unacceptably close to a short circuit already, and adding ohmic resistance in parallel will make things worse, not better.
Finding an equivalent replacement woofer, or repairing the surround on the broken one, are really the only reasonable choices. You could redesign the crossover given sufficient engineering expertise, but that would likely cost much more than replacing the woofer, and would in any case result in asymmetry and a muddied stereo image.
posted by flabdablet at 4:50 AM on May 7, 2012 [3 favorites]
Here's a pretty good article containing a number of graphs showing the impedance of assorted "8 ohm" speaker cabinets at various frequencies. These are impedance responses, not overall sound output responses (those would typically be much flatter).
Note all the weird humps and dips in the bass regions: these arise from interactions between the driver, the crossover and resonant effects due to cabinet design. It's not hard to imagine that altering any of those parameters would shift the peaks and troughs around, potentially making one or more of the low-impedance minima even lower.
posted by flabdablet at 4:58 AM on May 7, 2012
Note all the weird humps and dips in the bass regions: these arise from interactions between the driver, the crossover and resonant effects due to cabinet design. It's not hard to imagine that altering any of those parameters would shift the peaks and troughs around, potentially making one or more of the low-impedance minima even lower.
posted by flabdablet at 4:58 AM on May 7, 2012
This thread is closed to new comments.
If I had to guess, it's more likely to be the amp is sensing the impedance mismatch between the speakers, or the fact that the overall impedance is now not 6 ohms on that channel.
posted by token-ring at 11:42 AM on May 6, 2012