Speaker matching.

[Josh 2]

Hi all,

 

As many of your forum posts begin... 'I know this has been covered before', I hate to repeat it, but I have searched and can't quiet find an answer. I also appreciate my question is basic, but every time I think about hooking up new/different speakers I get confused. Could someone help me understand please (excuse any incorrect terminology!):-

 

I believe that I understand correctly that a speaker should optimally (safely) be rated at ½ of the amplifier output, so eg: a 1000w amp would ideally drive a 500w speaker with a matching impedance. Is this correct?

 

If the above is correct... 1, can an amp rated at 4Ohm safely drive an 8Ohm speaker and if so what does this do as far as wattage is concerned. ie: with a 1000w/4Ohm amp would you still, optimally, use use the 'half rule', therefore a 500w/8Ohm speaker or would it be either a 250w/8Ohm or 1000w/8Ohm or is there a totally different matching rule?

 

I wish I could find a way to remember this and get it stuck in my brain :)

 

Thanks for any help.

[ramdram]

http://www.gcaudio.com/resources/howtos/spkramp.html

 

http://www.ehow.com/how_5915892_match-speaker-power-amp.html

 

http://www.horrorseek.com/home/halloween/w...ceMatching.html

 

HTH

[paulears]

Don't forget that amp ratings just give you an idea about the closest to a short circuit they can tolerate. So, the amp workest hardest with the lowest impedance load. Hence why when the 4Ohms turns into 8, the apparent output goes down. Using loudspeakers that are higher impedance means less volume - but don't get thrown too much, because then if you added an extra pair of speakers, the power handling of the combined system goes up, the impedance goes down and as long as the things match, all is well.

 

Use the figures to work out where mismatches occur. Some brands of amp quote the absolute maxium the amp can provide simply because 1000W looks better in an advert than 500W which would be what they'd probably put if they quoted 8 Ohm specifications.

 

The obvious upshot of all this specification one-upmanship is that some people do like to have surplus amp capacity - and treating it a bit like a car, you don't have to drive the V8 at 100mph, but if you have to .... Others disagree and feel that all that is required is undistorted power from the amp and this won't hurt loudspeakers actually rated as being unable to handle the power applied. Another group are absolutely certain that speakers with a higher capability, power wise, than the amp are therefore the 'secure' bit, and can operate at their highest potential quality as they're not stressed. To be honest, there's physics in these arguments - but I think my own view is that it's important to use the specs sensibly, taking the makes and style of kit into consideration. Trying to find a magic formula that guarantees success is difficult. I can think of many speaker systems that might well have very large power handling capability - but when pushed are known for producing horrible sound. This isn't something specs alone can predict.

 

So matching a 1000W (any flavour of Watts) with speakers with a 100W label isn't going to be a good match - but if levels are kept low, could be quite listenable.

 

A Pair of wopping great 1K capable speakers, driven by a 100W amp isn't really going to be damaged, but sensitivity (which often goes down as ratings go up) might mean you can talk easily over the amps maximum volume, and flat out it isn't going to sound nice.

 

My current feeling is that you just need to match reasonably closely, but match based on paper specs, price, sound quality and any other factor that might be important - not stick rigidly to some kind of maths.

[Josh 2]

Thanks for the links ramdram and many thanks for writing the detailed info paulears.

 

I know it's my own stupid fault, but even after reading and more reading I'm still not sure I fully understand... my apologies guys.

 

I guess I have added to my initial understanding that I can safely (within reason) increase the impedance of a speaker in as much as a 4Ohm rated amp can drive an 8Ohm speaker load, but I shouldn’t decrease the speaker load, for example an 4Ohm speaker load on an 8Ohm rated amp. However, it would be preferable to have an exact match (given the parameters of manufacturers specs). Also now understood that increasing the speaker load will reduce the speakers effectiveness (volume)... am I getting there?

 

Paul, I'm now getting confused again about what I thought I initially understood about speakers being ½ the amp wattage (considering matched load for a moment). Are you saying that there are different opinions and some advocate speakers rated higher than the amps output?. I guess it sounds logical that the speakers would be less stressed... just when I thought I had grasped one concept! :)

 

I still can't seem to work out what wattage would be optimal (given various opinions exist) if I drive a 8Ohm speaker load from a 4Ohm rated amp... is the a rule of thumb?

 

Am I missing something or being stupid.

[Dmills]

Most amps rated at 4 ohms will deliver a little over half their 4 ohm rated power into an 8 ohm load.

So am amp rated at 1000W per channel into 4 ohms will probably deliver just north of 500W into an 8 ohm load.

 

Now speaker power ratings are a can of worms, mostly because of the 3 or 4 ways speaker power ratings are given....

There is usually an "RMS" (properly continuous) power rating, then there is a 'program rating' (usually about double the continuous rating), then there are the completely made up numbers the car audio crowd use (PMPO and the like).

 

Now my personal preference tends to size the amps to roughly match the program power rating at the appropriate impedance, so if the box was good for say 500W RMS (they mean continuous), 1000W program and was an 8 ohm unit then I would be looking for an amp that did about 1000W per side into 8 ohms, which if the amp is specified into 4 ohms means I will be looking for something that does just shy of 2000W per side into 4 ohms....

Now with that combination it is entirely possible to blow that speaker, but that really does not change much as amps get smaller until you are down to really tiny amps, clip any of them hard enough and you will take the horns out.

 

Really you are coming at this from the wrong direction, you need to start with knowing how loud you want to be, and what coverage, then pick boxes and amps to suit that requirement, don't start with the speaker, start with the room and the required volume in the audience, it may be that you end up with a pile of boxes that can handle way more then you could ever need, in which case you can save some money on the amps.

 

Regards, Dan.

[paulears]

You're getting on the right track. As for people's opinions - yep, there are conflicting personal preferences, many discussed over the year on here. The real key to all this is simply down to Ohms law. The comments about the speaker impedance are quite common talking points. In many cases, a manufacturer designs a single box, with a maximum rating on the cabinet labelled perhaps as say, 300W - 8 Ohms. They are quite aware that some people will wish to get power output up for bigger rooms, and the simplest way would be to buy two of them, and then connect them in parallel. If the amplifier can handle driving loads of 4 Ohms, then if it produces 600W at 4 Ohms, each speaker gets it's maximum input level, and it's louder. So when you look at the amplifier output powers, you can see why the output is different at specific impedances. As one measurement goes up, another goes down - just as Ohms law states.

 

It's quite handy, really - because one amp, connected to a single cabinet will be quite happy, and as long as it's design can cope with the impedance halving when another cabinet is attaches in parallel - it will provide sufficient extra to cope with both!

[Josh 2]

Thanks again Dan and Paul. It's extremely kind of you both to take the time. I do feel like a proper moron, because I keep reading and when I think I understand another spanner goes in the works!

 

To some extent, I think I have approached this with a basic misunderstanding. 'If' manufacturers quoted (or where able to quote) accurate and consistent figures for equipment, I guess that physics and Ohms law could be applied and an exact match could be found. However, I guess the science can't be exact due to power and impedance being arbitrary and variable, ie: amps output and speaker impedance varying across the operating range due to electro/mechanical considerations. Therefore the best that can be achieved is a perceived ('perceived' due to differing opinions) optimal match... am I still grasping this?

 

Dan, sorry I got a bit lost yet again. You mention that a 4Ohm amp driving an 8Ohm load would give a few more watts, but you add that your preference for a 1000W/8Ohm amp would be to use a 1000W speaker at 8Ohm or -2000W speaker at 4Ohm. Can I ask (given that there are different preferences/opinions) isn't this going the wrong way round and also a rather big difference or am I misunderstanding.

 

Paul, again my apologies, you use a 300W – 8Ohm example and if I read correctly, suggest this would provide 600W at 4Ohm... there is something I'm obviously still not getting here where Dan mentions a small increase in output.

 

I'm truly sorry guy's, I must be coming across as really dim and I appreciate your patience.

[paulears]

No - It's my fault for over simplifying - At random I grabbed the specs from a couple of manuals handy.

 

1000 W RMS per channel 2 Ohms

700 W RMS per channel 4 Ohms

400 W RMS per channel 8 Ohms

 

1200 W RMS per channel 2 Ohms

750 W RMS per channel 4 Ohms

500 W RMS per channel 8 Ohms

 

The figures are pretty close really.

 

Dividing by two seemed to highlight the maths element, rather than 1.4ish. Output wise, as Dan correctly states, it's quite small. 1000W is not much more than 700W really.

 

Remember that there's no such thing as a 4 Ohm amp. They are designed to work over a range, and it's just electrical matching that enables the amp to deliver what it's output devices and power supply can manage.

 

Let's try swapping from audio to school physics. Did you ever do experiments with batteries and small bulbs - parallel and series connections?

 

If you wire the lamps in parallel, then they'll be full brightness, wire them in series and the resistance is doubled, and there is half the voltage across each one. Unless you try to draw too much current from the power supply, any combination of load arrangement is fine - the battery doesn't care, but maximum light output only happens when the individual lamps are driven to their rating, and the power supply keeps within it's capability. All these things can be calculated.

 

When we're doing loudspeakers, rather than worrying about this in electrical terms, think about quality of sound. Speaker measurements are usually a bit of a joke anyway, in brands other than the higher quality ones. If they slap a label on saying 300W max, what does it really mean. What are Watts? Most people quote RMS Watts - 'average' power, but others have a more unusual way of describing them. The word 'average' nowadays is better replaced with 'mean' which does have a mathematical base, and of course is linked to the Root Mean Squared RMS description. Peak Power is another method of labeling speakers - but means very little. So consider what the label is saying, if, it says anything at all.

 

If I bought a loudspeaker from a reputable manufacturer then if I had a biggish amp and would be driving the loudspeaker at a sensible level I could control, I'd be happy. I wouldn't worry if the speaker was over or under the amp rating. I tend to have bigger amps for usefulness - and hanging a pair of SX300 plastic boxes on the end wouldn't bother me - nor would a pile of big boxes. However, if somebody gave me an unbranded, chipboard box that said 1000W in big print on the back, I'd probably take very great care not to destroy it.

 

 

I'm not sure we can really explain this properly because to do so involves not doing what I've done - simplifying for clarity, because this prevents accuracy - but if we did it accurately, we'd confuse. It could be done face to face, because then every time there's a blank look, you can reverse a bit and try a new route through the jam - something that doesn't work in text.

 

 

Remember Rolls Royce, the first time Clarkson asked them the BHP of their new supercar. They said "sufficient".

[Josh 2]

Thanks for clarifying Paul.

 

It's funny that you mention school. On a purely personal note and by way of explanation, I could well have done those experiments, but I wouldn't know!. Some 20 years ago I had a m/cycle accident resulting in loss of memory and some cognitive skill (why I'm stupid, I guess). It took me years to re-learn language, how many pennies in a pound, days in week etc etc. Hence I am always wary about asking questions in case I'm making a fool of myself (not unusual :) ).

 

However and strangely, I had been thinking about light bulbs and reasoned it in a different (reverse) way. I was working on the theory that if I screwed a 100W bulb into a 13A socket it would have x efficiency and y longevity due to the filament (ignoring other factors), so I concluded that the filament was the load/resistance (impedance). If I screwed a 60W bulb in the same socket, I would be decreasing the efficiency (light output), but increasing the longevity because the filament would be stronger (thicker). Then I started thinking again and decided this can't be right, because if the filament is thicker in a 60W bulb then the resistance (impedance) must be higher and draw more power, but a 60W light bulb is cheaper to run... at which point I reached for the aspirin :)

 

I completely understand what your saying about a text forum v face-to-face and there would sure be plenty of those blank looks from me and I do remember the Clarkson quote which seems very apt for this topic.

 

Paul, guys, I really don't want to impose and protract the topic because of my own failings, so can I conclude by thinking that the parameters for speaker amp/matching are (given decent equipment, driven at sensible levels), really quiet wide in the real world and that I have been worrying too much about a perfect match?

[back_ache]

the parameters for speaker amp/matching are (given decent equipment, driven at sensible levels), really quiet wide in the real world

 

You've hit the nail on the nail on the head there, good equipment, operated by good people is inherently flexible.

 

(However, if the equipment is to be put in an environment where the people operating it are going to be under the influence of alcohol, adrenaline or whatever, over-spec it!)

[Josh 2]

Thanks back_ache and thanks everyone for the time and patience put into the replies... very much appreciated.