matching amp with speakers

[Brian jojade]

Re: matching amp with speakers

Matching an amp with speakers has more to do with the operator experience than anything else. If you are dealing with a user that has no clue as to what they are doing, setting them up with an amplifier that has HALF of the RMS rating of the speaker is a pretty safe bet. This would prevent the user from giving the speaker too much power even if run into full tilt clipping and distortion. Granted, you still could cause over excursion damage, but the point is that the amp can't provide enough power to overheat the speaker.

As you get more experienced, you'll realize that you can use a larger amp and if you can control yourself, you can get more out of your speakers that you have. Matching the amp to RMS rating is safe for the average non idiot user. Non idiots are those that are smart enough to know that the clip light isn't supposed to be on continuously. There's a reason the light is RED on every amplifier. It's not the end of the world if it blinks occasionally.

Of course, seeing the clip light does mean that you've exhausted the power that the amp can deliver and it is potentially affecting the sound. So, getting a bigger amp is a solution. However, once you start doing this, you have to realize that the clip light is no longer a warning light. In fact, if the amp is sized high enough over the rating of the speaker, the clip light becomes an oh shit light, and means if you see this light, your speaker isn't working anymore. Better luck next show.

Depending on the type of speakers you use, they may not exhibit any audible stress signals before death. Others are designed in a way that they can handle more power than they could use. Eg, I used to run old Peavey cabinets. If I ever drove them beyond their full rated power, the speaker would sound horridly distorted. The nice thing is that I never had to worry about them blowing up because of too much power, since I'd never run them sounding that bad. Now, other speakers still sound good up to and past their rated power and thus overpowering them can be more tempting.

The type of music being played is an important consideration of how big you can go. For EDM, that has a lot of sustained notes, matching RMS rating of the amp and speaker is going to be a much safer bet. For live shows that are much more dynamic, then a larger amplifier can safely be used.

Undoubtedly, though, Bennet's use of an RMS limiter is the most fool proof setup method.

 

[Don Boomer]

Re: matching amp with speakers

I used to run old Peavey cabinets. If I ever drove them beyond their full rated power, the speaker would sound horridly distorted.

Here's the best advise so far. If it starts sounding like crap you should seriously think about what is going on.

Trying to figure where to set limiters based on manufacturer's specs is somewhat a guessing game because the specs are considering a conditioned signal which may or may not resemble your program material. So you have to make an educated guess as to how far apart you really are.

Depending on your skill level, I think for best possible performance you should use amps that can deliver the full rated peak power of your speaker. That way the built-in limiters will be clamping down should you push the amp into actually delivering those peaks. But at the time you fire off those limiters I would expect that you would be delivering somewhere around 1/8th of the continuous rated power of that amp into top boxes and maybe 30-50% with heavily compressed bass material into subs. So I don't see you ever hitting Bennet's -3dB RMS limiting point. It should start sounding poorly before that.

All that said, in my experience over excursion is probably 10-50 times more likely to cause you problems than over-powering in the thermal sense.

 

[Avi Goldberg]

Re: matching amp with speakers

Dear Bennett,
I see from the graph that although it takes 30 minutes to reach max temp, it takes only 3-4 minutes to reach 90-95% of max temp. Does it take less time to reach max temp, or the 90-95% point if you put more energy in or does it just push up the points of max temp and the rest?

Do you use the 100 deg limit because the type of Neo magnets speakers use nowadays will lose magnetism permanently beyond that point? At what room temp do you do those tests and to what temps do you think the VC/Magnet will reach on hottest days? Do you use different guidelines for Ferrite based speakers?

The RMS/Continuous/AES power spec represents the VC durability to heat. But what does the Peak power spec represent? Does it stand for the speakers maximum mechanical limits (Xmax)?

 

[Bennett Prescott]

Re: matching amp with speakers

Dear Bennett,
I see from the graph that although it takes 30 minutes to reach max temp, it takes only 3-4 minutes to reach 90-95% of max temp. Does it take less time to reach max temp, or the 90-95% point if you put more energy in or does it just push up the points of max temp and the rest?

Do you use the 100 deg limit because the type of Neo magnets speakers use nowadays will lose magnetism permanently beyond that point? At what room temp do you do those tests and to what temps do you think the VC/Magnet will reach on hottest days? Do you use different guidelines for Ferrite based speakers?

The RMS/Continuous/AES power spec represents the VC durability to heat. But what does the Peak power spec represent? Does it stand for the speakers maximum mechanical limits (Xmax)?

Hey Avi,

If you put in more real power the coil and magnet structure will reach max temp more quickly - and then exceed it. The slope will be more or less the same.

We use the 100° limit because it leaves what we consider a safe working margin, as we would like the woofer to survive its rated power indefinitely. The Curie temperature of our neodymium magnets is above 250°C, but they will begin to lose strength around 120°C. The ferrite magnets have a Curie temperature of about 500°C, so not anything we worry about. The bigger problem is the glue used in magnet structure assembly has a working temp around 110°C, so it will begin to soften and change color - convenient as an indicator of abuse. Honestly, none of these are problems we ever see, the increased temperature in some very hot environments has a much greater effect on the voice coil. It is much more likely that the coil will reach its maximum temperature before the magnet structure does, even though that maximum temperature is quite high (>300°C). Our testing is done in a temperature controlled room at about 23°C in order to simulate an outdoor environment.

Here is a good guide to magnet temperature, we use temperature grade H magnets: http://www.supermagnete.de/eng/faq/What-temperatures-can-magnets-sustain

As for "peak power", it doesn't exist. It is sometimes used as a guide to amplifier selection, but for many products there isn't an amplifier available that could provide it anyway. It cannot be used as a guide to excursion as excursion is not easily relatable to input power. At some frequencies it may be possible to overexcurse a woofer at a small fraction of rated power, at others it will accept many times rated power before over excursion (but of course the coil will burn). It is mostly dependent on box design and processing. We only provide a continuous power rating on our spec sheets to aid with amplifier selection, and because everyone else does! In reality all these numbers are just different ways of saying the same power handling with different signal crest factors. A limiter that limits to real RMS power will do so correctly regardless of input signal.

 

[Phil Graham]

Re: matching amp with speakers

Robert,

Klippel uses a pilot tone of about 2Hz and measures VC impedance rise to determine temperature. We have some cool and less cool equipment for measuring magnet temp, but there is not a strong correlation between the two. We fail power testing if the magnet rises above 100°C for most models, and if the VC rises above 300°C for most models. Some magnets and coils can survive more indefinitely, but that's a good rule of thumb. As you can see you've got to watch them both! The below image is a 21SW152 at the end of a 4 hour power test.

View attachment 9917

To expand slightly, the Klippel power testing can utilize very advanced multi-tone stimuli:

• The first tone is a very low frequency (1Hz'ish) tone to ascertain the dc resistance of the voice coil as a proxy for temperature.
• The second tone is a high frequency (e.g. 2kHz) tone that dumps the heating power into the voicecoil. The idea is to heat the voicecoil at a frequency that doesn't drive cone movement, and therefore cooling.
• The third stimuli can be used to drive cone motion at a specific frequency, or range of frequencies, to look at the effects of cone and voice coil movement on cooling behavior.

 

[Don Boomer]

Re: matching amp with speakers

A little more about peak power.

So when you run an AES power test you are using a "conditioned" pink noise signal. I was curious as to how many peaks you might be subjecting a speaker to at a given average/RMS power level. We ran some tests and found it all over the place ... from about 60 to 200 peaks per second.

The AES noise signal is very hard limited to 6 dB peaks. But real world music could be considerably different.

 

[Jimmy Hardin]

Re: matching amp with speakers

ok guys . Here is my Amp specs. I am running a 4 ohm load in stereo. and i am looking at possibly getting new speakers. Will this setup be a Bad, good or perfect match for my amp? I will be running 2 speakers per side of my amp. from what i have been told, when you have 2 speakers on a 4ohm load then the output power on the amp is split 475 watts is going to one speaker and 475 going to the other. so i am wanting to know if these speakers will be a good match for my amp or will i blow them? I hope i am making sense on this. The guy in the video got me to thinking about things when he said that when you power a 500 watt speaker with a 500 watt amp it is equal to a strainious fast run but when you power a 500 watt speaker with a 800watt amp then it will be like a slow paced walk. the speakers that i have now are rated at 400 (rms)-800 peak watts. and i am only putting 475 watts a speaker through them. from this amp. Will this blow anything?


I am running a Behringer EP4000

• Output Power
• 20Hz–20kHz @ 0.1% THD, both channels driven:
• 8Ω per channel-550W
• 4Ω per channel-950W
• 2Ω per channel-1250W
• RMS @ 1% THD (sine wave), bridged mode:
• 8Ω-1750W
• 4Ω-2400W
• Peak Power, both channels driven:
• 8Ω per channel-750W
• 4Ω per channel-1400W
• 2Ω per channel-2000W
• Peak Power, bridged mode:
• 8Ω-2800W
• 4Ω-4000W

Speakers (I am looking at getting these Mackie s515 speakers)

• System
• Frequency Range (–10 dB): 47 Hz – 18 kHz
• Horizontal Coverage Angle: 90 ̊
• Vertical Coverage Angle: 50 ̊
• Sensitivity (1W @ 1m): 101.1 dB SPL
• Nominal Impedance: 8 ohms
• Power Handling: 300 watts continuous
• 600 watts program 1200 watts peak
• Crossover Frequency: 3400 Hz

 

[Tim McCulloch]

Re: matching amp with speakers

You probably won't blow up anything if you keep the amps out of hard clipping. If the clip light flashes only on peaks you'll probably be okay.

The bigger question should be "will this combination of speakers and amp get loud enough for my uses?"

Watts are a measure of energy being transferred, i.e. work being done (or heat being generated). You don't "send" Watts to a speaker, you send voltage, which results in movement of the voice coil in the magnetic gap. The Wattage rating for an amplifier is how much energy is available to sustain the motion caused by the application of voltage.

 

[Don Boomer]

Re: matching amp with speakers

so i am wanting to know if these speakers will be a good match for my amp or will i blow them?

If blowing them up is your prime consideration then what you really need to watch is how much material below 50Hz you send to the speakers. You are probably 10 times more likely to damage the speakers with material that is below the tuning point of the box than with too much amplifier power in the mid bands.

 

[Art Welter]

Re: matching amp with speakers

If blowing them up is your prime consideration then what you really need to watch is how much material below 50Hz you send to the speakers. You are probably 10 times more likely to damage the speakers with material that is below the tuning point of the box than with too much amplifier power in the mid bands.

Don,

This is the second time you have mentioned that statistic in this thread. Back in the "olden days" of no HP filters or limiters, my experience was far more burnt coils than mechanical damage, even though the drivers were frequently driven hard below Fb. The rising impedance below Fb coupled with less magnetic force available when driven beyond Xmax with progressively stiffer suspensions at excursions above Xmax led to few mechanical failures.

What drivers are you basing a 10 (or 50) to 1 mechanical vs. thermal failure rate on?
Got photos?

Art

 

[Ivan Beaver]

Re: matching amp with speakers

Don,

This is the second time you have mentioned that statistic in this thread. Back in the "olden days" of no HP filters or limiters, my experience was far more burnt coils than mechanical damage, even though the drivers were frequently driven hard below Fb. The rising impedance below Fb coupled with less magnetic force available when driven beyond Xmax with progressively stiffer suspensions at excursions above Xmax led to few mechanical failures.

What drivers are you basing a 10 (or 50) to 1 mechanical vs. thermal failure rate on?
Got photos?

Art

I agree with Art. I also grew up in the age of no highpass filter for the lows.

I tore up hundreds of drivers. That is why I got into reconing-to save me money and make some extra on the side.

Typically the only damage from over excursion was due to worn out- or damp cone or sometimes a large pulse that went through the system.

At least "back in the day" burnt voice coils were the normal failure.

Of course we did not have the large amps capable of the large voltage swings that we do now. So it was more of an "average power" that killed the speakers.

And we did not have the adhesives that could handle the higher temps like today.

Oh how things have changed so much in my lifetime.

And what is truly sad (warning topic serve) is that we have all sorts of tools for predicting sound-measuring sound-adjusting all sorts of parameters (DSP) and yet I am not convinced that the "overall" sound quality that people hear is any better. In many cases it is worse-because now cheap gear is available to anybody who does not have clue how to use it-and all they want is "loud".

At least "back in the day" audio gear was more expensive-so that helped keep the "riff raff" out.

It was also MUCH harder to use. Just go back to when you had to match impedances for all the gear to get the most power transfer between units. If you did not all sorts of weird things would happen-including VERY altered freq response.

Just ask anybody who has tried to use a passive White EQ into a modern amp with a high input impedance-------------------- It could drive you nuts-until you terminated it with a 600 ohm resistor.

Sorry for the rant..

 

[Michael John]

Re: matching amp with speakers

Hi Bennet,

Just curious why you recommend long term RMS limiting to half AES or IEC power? I thought the AES (or IEC) power test was fairly conservative anyway?

Cheers,
Michael

Hey Jimmy,

I think if it takes 8 minutes for the video to explain it, they are being too complicated My rule of thumb is like this:

1. What's the biggest amplifier I can get my hands on? Use that.
2. Use a long term RMS power limiter (preferably built in) to limit to 1/2 rated AES power (whatever the lowest power on the woofer spec sheet is).
3. Use a proper HPF set at approximately the loudspeaker's tuning frequency.

For most PA cabinets this works, in other designs that do not control excursion well or have unusually low tunings you may run out of excursion sooner than you'd like. In general I find that for modern woofers I can't get an amplifier large enough to satisfy my peak power desires, I'd like no less than 2x AES power (often labeled "Program") and 4x (often labeled "peak) would be better. You can pretty much assume 12dB peak to average in audio, round down to 10dB and that means I need a 5,000 "watt" amp for my 500 "watt" woofer. YMMV, but I don't see many toasted woofers - except in circumstances where the power capacity was actually lower than expected for some reason.

Here is an article I wrote a while back on this subject:

http://www.bennettprescott.com/downloads/LoudspeakerFundamentals.pdf

 

[Tim McCulloch]

Re: matching amp with speakers

Hi Bennet,

Just curious why you recommend long term RMS limiting to half AES or IEC power? I thought the AES (or IEC) power test was fairly conservative anyway?

Cheers,
Michael

Power compression.

 

[Jimmy Hardin]

Re: matching amp with speakers

You probably won't blow up anything if you keep the amps out of hard clipping. If the clip light flashes only on peaks you'll probably be okay.

The bigger question should be "will this combination of speakers and amp get loud enough for my uses?"

Watts are a measure of energy being transferred, i.e. work being done (or heat being generated). You don't "send" Watts to a speaker, you send voltage, which results in movement of the voice coil in the magnetic gap. The Wattage rating for an amplifier is how much energy is available to sustain the motion caused by the application of voltage.

ok ,I dont think i would have any trouble as far as the speakers getting out there to the audience . I dont do events for like for thousands of people , just small events so i may be ok. But what confuses me about it. i have always thought that if you got a 500 watt speaker then thats the size amp you needed to run the speaker. I had an encounter before where i put a mackie 1400I amp in bridged mode through a set of speakers and the woofer was only rated at 200 watts RMS to 400 wats peak. and it blew them , I might have put that story at the first of this post. and i have been very cautious since then. and then this guy on the video that is on this thread was saying that to run a 500 watt speaker you needed a 800 watt amp. Now i did a job saturday and my top amps got pretty warm. and it was under a tent so it wasnt sun light . and i began wondering if that theory was true or not.

 

[Tim McCulloch]

Re: matching amp with speakers

ok ,I dont think i would have any trouble as far as the speakers getting out there to the audience . I dont do events for like for thousands of people , just small events so i may be ok. But what confuses me about it. i have always thought that if you got a 500 watt speaker then thats the size amp you needed to run the speaker. I had an encounter before where i put a mackie 1400I amp in bridged mode through a set of speakers and the woofer was only rated at 200 watts RMS to 400 wats peak. and it blew them , I might have put that story at the first of this post. and i have been very cautious since then. and then this guy on the video that is on this thread was saying that to run a 500 watt speaker you needed a 800 watt amp. Now i did a job saturday and my top amps got pretty warm. and it was under a tent so it wasnt sun light . and i began wondering if that theory was true or not.

Most likely you simply drove the rig too hard. Shit doesn't blow up unless you turn it up beyond the design capabilities of the speakers. Also, when you put an amp in bridge mode you increase the gain by 6dB, meaning an input of only 0.37volts will drive most amps (those with 1.4v sensitivity is stereo mode) to Full Tilt Boogie. If you were setting a console output level based on seeing the mixer's meters in a particular position, you'd be waaaaay over-driving relative to an amp in stereo mode.

Repeat after me: speakers die because we try to make them go louder and/or lower than they are designed for. They die of thermal abuse or mechanical damage or sometimes a combination of both. But it's not the Fates that determine what happens...

 

[Ivan Beaver]

Re: matching amp with speakers

ok ,I dont think i would have any trouble as far as the speakers getting out there to the audience . I dont do events for like for thousands of people , just small events so i may be ok. But what confuses me about it. i have always thought that if you got a 500 watt speaker then thats the size amp you needed to run the speaker. I had an encounter before where i put a mackie 1400I amp in bridged mode through a set of speakers and the woofer was only rated at 200 watts RMS to 400 wats peak. and it blew them ,.

What brand of speakers? Many of the cheap speakers tend to "overrate" the drivers and not use standard methods.

It is also HIGHLY dependent on the crest factor of the program material. Material with a low crest factor puts a lot more heating on the voice coil.

There is no simple answer for amp rating and speaker power ratings. There are A LOT of "it depends".

 

[John Roberts]

Re: matching amp with speakers

There has been a great deal written about this topic over the years. Even including some popular rules of thumb, but as Ivan said no simple definitive answers.

Another way to think about this is what size motor do you put in your car/truck? The truck is heavier (like a low efficiency speaker) so needs a bigger motor just to accelerate and drive similar to the car. Just because you have a possibly oversized motor in your car/truck doesn't mean you will use it at full-throttle all the time. It would probably overheat or break something if you tried.

An important part of this analogy, often forgotten in the speaker/amp size discussion is the operator. In a car/truck the bigger the motor the better, unless it is too difficult to drive safely or without breaking something. Same thing goes for amp/speaker combinations, as long as you can drive the system "safely" some extra power is generally a good thing (it's called headroom). You can overheat and/or break a too small motor/car-truck combination by pushing it too hard just to keep up with traffic.

The right balance of power for the equipment keeps everybody happy, but the driver (operator) can still break it if he pushes a properly sized system too hard.

JR

PS: Back about 20 years ago while I was working at Peavey I generated a matrix of all the amps that Peavey sold on one axis, and all the loudspeakers that Peavey sold on the other axis. Then I asked the transducer (speaker) design engineers to check the boxes for the amps that they wanted to see their speakers designs driven by. This was a useful reference for the Peavey dealers and customers while amplifiers have grown larger since then and loudspeaker power handling has evolved too. I make no secret of the fact that I favor powered speakers where once again the design engineer is making these decisions for us. Just like the car makers decide what motor to put in your car.

 

[Art Welter]

Re: matching amp with speakers

Hi Bennet,

Just curious why you recommend long term RMS limiting to half AES or IEC power? I thought the AES (or IEC) power test was fairly conservative anyway?

Cheers,
Michael

I'm not Bennett, but I use his (and other brands) of speakers rated using the AES rating :^)
The AES pink noise signal has six dB crest factor, double the crest factor of a sine wave, which has 3 dB crest factor. If the speaker absorbs the RMS average of half the AES rating, it is seeing the equivalent power of a sine wave rating.

"Normal" pink noise has a 12 dB crest factor, a 1000 watt amp just seeing clip using pink noise is only putting out about 125 watts average, a signal with 3 dB crest factor would deliver 1000 watts average.

Since digital compression (hey, it's on every channel) is capable of squashing the dynamic range of music down to ever decreasing crest factors, it is fortunate that driver manufacturers have found adhesives capable of the high heat generated in the voice coils, and cooling mechanisms capable of removing much of the heat from the voice coil.

It is interesting that many amplifiers are now unable to deliver full power (rated using sine waves) for more than a fraction of a second without going in to current limiting or popping a breaker, while LF program in some music now actually has less than 3 dB crest factor, and HF program as squashed as the AES signal.

 

[John Roberts]

Re: matching amp with speakers

It is interesting that many amplifiers are now unable to deliver full power (rated using sine waves) for more than a fraction of a second without going in to current limiting or popping a breaker,

While perhaps TMI for the OP, some of the larger modern amps put out more short term power than the line cord can supply longer term, so by definition can only do so for a fraction of the time. An amp that delivers full rated output 24x7 is arguably over-designed for reproducing music (I am old enough to remember when they did).

while LF program in some music now actually has less than 3 dB crest factor, and HF program as squashed as the AES signal.

To be less than 3 dB crest factor you are talking clipped sine wave approaching a square wave. While heavily clipped bass program can momentarily drop below 3dB, for it to spend too much time, playing that heavily clipped, it would sound more like a guitar fuzz box, than bass line.

I do not dispute that some modern genres of music (?) push very heavy duty cycle in the low frequency bandpass.

JR

 

[Ivan Beaver]

Re: matching amp with speakers

An amp that delivers full rated output 24x7 is arguably over-designed for reproducing music (I am old enough to remember when they did).

Me to. Back then the way to test the power of an amp was to drive it into a dummy load and measure the voltage at clipping.

That took several seconds-WAAAAYYYYY longer than most modern amps can deliver their "rated" power.

But you also knew what you were getting. Today the big "unmentionable" in how long they can deliver rated power. and then what the power drops down to.

For many amps it is simply embarrassing. So they hope their customers don't understand the basics and "simply believe" the specs. Without ever testing.

I wonder how amps "back in the day' would have been built/designed if going by current "standards"-which arguably is closer to what is actually required.

So much has changed in the way of amp design over the last 30yrs or so. What we currently have was completely unthinkable back then.