Questions about amplifier power specs

[Jfs]

I have some questions about amplifier specifications:

I'm using this as an example:

NAD C325BEE
2 x 50W Minimum Continuous Power into 4 / 8 ohms
110W, 160W, 210W, IHF Dynamic power into 8, 4 and 2 ohms, respectively

From what I understand dynamic power is important for music, as well as
power into lower resistance to drive speakers properly.

What would be good amounts for dynamic power for these various resistances?
should they be a percentage of continuous power?
I can understand dynamic power for 8 ohms (eg. 50 to 100 watts), but don't
understand what is good for other resistances.

Is IHF a good measurement?

How much continuous power do I need into lower resistances?
I know there should be allowances for variances in resistance over
frequencies, but I'm not sure what would be good for most 8 ohm speakers.
Some amplifiers quote the power as 50 watts into 8 ohms and 75 watts into 4
ohms. I'm not sure what the above specs mean. Shouldn't there be an increase?

Could someone give me an idea of how much power I would need. Is 50 watts
enough for most speakers. I probably will be looking for something used.


I'm trying to get an idea of what I should be looking in an amplifier. Would
appreciate help.

Thanks.

 

[greggp2]

NAD's dynamic power ratings are a source of great criticism by many people. Is 50 watts enough power? It really depends on what types of speakers you are trying to power and how many Ohm's they operate on. Typically though, 50 watts is very little power, especially for a full range speaker that you are using for 2 channel stereo..

But look for power that is continuous as opposed to just stated as 100 watts/channel. Those ratings are usually peak, not continuous...

 

[just-some-guy]

not only that. but look for ratings with 20hz -20k hz. the simple 100wpc @ 1k hz means lower power. though, in my humble experience, it wasn't a lot.

 

[PENG]

How much power you need depends on the characteristics of your speakers, room acoustic environment, and listening habits/preference (e.g. how loud, types of music, movies? etc.).

In general, pay attention first to your room acoustics, pick the speakers you like and then get as much power as you can afford. The NAD's dynamic power specs let you know how much more (only relative to its continuous rating) it can put out for a very short duration, perhaps up to 20 ms, or more. That may be important for classical music such as symphonies. It does not mean other manufacturers who don't provide such IHF figures cannot do the same or better.

Please note that music signals are a.c. by nature, so impedance (as opposed to ‘resistance’) is a more appropriate term. The impedance of a loudspeaker varies throughout its frequency response range, and is a complex number mathematically speaking. The complex nature of the speaker's impedance characteristics also mean there is a phase angle between the current drawn by a speaker and the voltage applied to it on a moment by moment basis. All these make it impractical to try and compare the power rating of amps purely on their specs on paper. You cannot simply use the basic V=IR and P=VI or P=I^/R formula. The frequency varying and complex nature of the load impedance must be factored into the calculations, again making it more difficult to compare power specs published by different manufacturers.

 

[3db]

How much power you need depends on the characteristics of your speakers, room acoustic environment, and listening habits/preference (e.g. how loud, types of music, movies? etc.).

In general, pay attention first to your room acoustics, pick the speakers you like and then get as much power as you can afford. The NAD's dynamic power specs let you know how much more (only relative to its continuous rating) it can put out for a very short duration, perhaps up to 20 ms, or more. That may be important for classical music such as symphonies. It does not mean other manufacturers who don't provide such IHF figures cannot do the same or better.

Please note that music signals are a.c. by nature, so impedance (as opposed to ‘resistance’) is a more appropriate term. The impedance of a loudspeaker varies throughout its frequency response range, and is a complex number mathematically speaking. The complex nature of the speaker's impedance characteristics also mean there is a phase angle between the current drawn by a speaker and the voltage applied to it on a moment by moment basis. All these make it impractical to try and compare the power rating of amps purely on their specs on paper. You cannot simply use the basic V=IR and P=VI or P=I^/R formula. The frequency varying and complex nature of the load impedance must be factored into the calculations, again making it more difficult to compare power specs published by different manufacturers.

I like to see the following specs in amplifier but I guess I'm dreaming

2 channels driven 20-20K through an impedance of 8, 4, and 2 ohms at phase angles of 0, 21.5, 45, 66.5 degress leading and lagging before reaching .5% distortion ( which I think is just a point for humans to detect this level of distortion )

Repeat this for all channels driven. I know r/t multichannel doesn't use all channels at full power but that doesn't make this test irrevelant. I wanna know what the power supply is capable off and the only way is all channels driven.

 

[PENG]

I like to see the following specs in amplifier but I guess I'm dreaming

2 channels driven 20-20K through an impedance of 8, 4, and 2 ohms at phase angles of 0, 21.5, 45, 66.5 degress leading and lagging before reaching .5% distortion ( which I think is just a point for humans to detect this level of distortion )

Repeat this for all channels driven. I know r/t multichannel doesn't use all channels at full power but that doesn't make this test irrevelant. I wanna know what the power supply is capable off and the only way is all channels driven.

My calculation shows that if an amp (1 channel) delivers 1A into an 8 ohm load with 0 degree phase angle (i.e. a pure resistive load) then it will have to deliver:

Into 2 ohms and phase angles (leading or lagging) of:

a) 0 degrees - 4A
b) 21.5 degrees - 4.3A
c) 45 degrees - 5.66A
d) 66.5 degrees - 10.03A
e) 75 degrees - 15.45A

If you go with ACD, for 7 channels, at 66.5 degrees, a single power supply transformer will have to deliver 7X10.03=70.22A (or 108.2A for 75 degrees phase angle on the secondary winding.

Let's assume the voltage on the secondary side is at 50V when delivering the calculated current above, and the primary side is 120V (from the wall outlet), then the current will be 70.22*50/120=29.26A, still 2 times the limit of a typical 15A circuit. So you have more than the amp's power supply to worry about if you insist on ACD.

By the way a 15A circuit can deliver a lot more for a short duration (before the breaker trips) but no matter how short the duration is, there is always voltage drop that is proportional to the load current. So theorectically there will be some sort of self current limiting effect due to voltage drop. Sometimes when you turn on a big amp your lights may dim due to excessive voltage drop if you happen to turn it on when the incoming a.c. waveform is just passing through 0V.

 

[mtrycrafts]

I like to see the following specs in amplifier but I guess I'm dreaming

2 channels driven 20-20K through an impedance of 8, 4, and 2 ohms at phase angles of 0, 21.5, 45, 66.5 degress leading and lagging before reaching .5% distortion ( which I think is just a point for humans to detect this level of distortion )

Repeat this for all channels driven. I know r/t multichannel doesn't use all channels at full power but that doesn't make this test irrevelant. I wanna know what the power supply is capable off and the only way is all channels driven.

Well, The Audio Critic used to report those specs for amps, called Power cubed by the inventor of the test gear for it Dr. David Rich who is at The $ensible Sound now, technical director/advisor. Looking at your post again, he used phase of 0, +/- 30 and 60, and down to 1 ohms.

 

[mtrycrafts]

My calculation shows that if an amp (1 channel) delivers 1A into an 8 ohm load with 0 degree phase angle (i.e. a pure resistive load) then it will have to deliver:

Into 2 ohms and phase angles (leading or lagging) of:

a) 0 degrees - 4A
b) 21.5 degrees - 4.3A
c) 45 degrees - 5.66A
d) 66.5 degrees - 10.03A
e) 75 degrees - 15.45A

If you go with ACD, for 7 channels, at 66.5 degrees, a single power supply transformer will have to deliver 7X10.03=70.22A (or 108.2A for 75 degrees phase angle on the secondary winding.

Let's assume the voltage on the secondary side is at 50V when delivering the calculated current above, and the primary side is 120V (from the wall outlet), then the current will be 70.22*50/120=29.26A, still 2 times the limit of a typical 15A circuit. So you have more than the amp's power supply to worry about if you insist on ACD.

By the way a 15A circuit can deliver a lot more for a short duration (before the breaker trips) but no matter how short the duration is, there is always voltage drop that is proportional to the load current. So theorectically there will be some sort of self current limiting effect due to voltage drop. Sometimes when you turn on a big amp your lights may dim due to excessive voltage drop if you happen to turn it on when the incoming a.c. waveform is just passing through 0V.

Not that long ago I posted about a wall meter called 'Kill A Watt' that has PF, VA, watts, current, cycle frequency, volts, watt hour. Interesting to see it operate with a system plugged in

 

[PENG]

Sorry I just noticed I made a few mistakes in my calculations. Please ignore my previous post. Here are the correct numbers.

Into a 2 ohms load:

0 degree (resistive)................4A
21.5......................................4.15A
45.........................................4.76A
66.5......................................6.33A
75.........................................7.86A

The increased in current is required to maintain the same watts output into a 2 watts resistive load (P=4X4X2=32W).

So for 7 channels, a 2 ohms load at p.f. of cosine (66.5), the total current on the transformer secondary winding would be 7X6.33=44.34A. Keep in mind this is what it takes to output only 32W per channel into a 2 ohm load with a phase angle of 66.5 degrees.

I won't try and estimate the power supply transformer primary (h.v. side) and secondary (l.v. side) currents as it will get too complicated and have to make too many assumptions. Suffice to say that on the high voltage side the current will always be lower based on the transformer turns ratio (or voltage ratio).

 

[MidnightSensi]

One thing I have fought, ignored and eventually given up on is this:

Nothing beats headroom.

If you run on the rails, get yourself a big amp and you'll never look back.

 

[Jfs]

Thanks for the replies.

I don't have much of an understanding of how amplifiers work, but I'm
trying to get a basic understanding so I know what to look for in
amplifier specs. If I can't look at specs to tell me how an amp will
perform then what good are they, and how would I know how one amp will
compare to another (If it has enough power for dynamics, enough for
dips in speaker impedance, etc.)?

I don't have a large budget so will probably looking for a receiver or
an integrated amplifier (entry level hi-fi or maybe something better
if I can get something used or discounted). I know I can't expect
perfection, but would like to be able to compare amplifiers in my
price range and know how they will perform. I did a search on the
Internet and didn't find a lot of help. I found one article that said
'dynamic power will be twice what the RMS is" and another article said
that "a bigger difference between RMS and peak power meant a weak
power supply". It all seems complicated.

Can someone recommend some good places for reviews of audio components?

Thanks for the help.

 

[3db]

Sorry I just noticed I made a few mistakes in my calculations. Please ignore my previous post. Here are the correct numbers.

Into a 2 ohms load:

0 degree (resistive)................4A
21.5......................................4.15A
45.........................................4.76A
66.5......................................6.33A
75.........................................7.86A

The increased in current is required to maintain the same watts output into a 2 watts resistive load (P=4X4X2=32W).

So for 7 channels, a 2 ohms load at p.f. of cosine (66.5), the total current on the transformer secondary winding would be 7X6.33=44.34A. Keep in mind this is what it takes to output only 32W per channel into a 2 ohm load with a phase angle of 66.5 degrees.

I won't try and estimate the power supply transformer primary (h.v. side) and secondary (l.v. side) currents as it will get too complicated and have to make too many assumptions. Suffice to say that on the high voltage side the current will always be lower based on the transformer turns ratio (or voltage ratio).

I'll just run a dedicated 30 amp circuit into the room used for my HT. Let the measurements begin.

Seriously though, I don't care if a 15 amp circuit cannot deliver. Connect the test up to a circuit that can deliver the current. I still want to see the performance of the amp. Its analgous to buying a corvette and having the car rags not test it flat out becuase the speed limit onmost highways is only 60 mph. Thats what I'm driving at, excuse the pun.

 

[3db]

One thing I have fought, ignored and eventually given up on is this:

Nothing beats headroom.

If you run on the rails, get yourself a big amp and you'll never look back.

Nothing wrong with running on rails as long as the current is there to deliver. Its semantics really and just different ways of speccing the output.

 

[Pyrrho]

I have some questions about amplifier specifications:

I'm using this as an example:


From what I understand dynamic power is important for music, as well as
power into lower resistance to drive speakers properly.

What would be good amounts for dynamic power for these various resistances?
should they be a percentage of continuous power?
I can understand dynamic power for 8 ohms (eg. 50 to 100 watts), but don't
understand what is good for other resistances.

Is IHF a good measurement?

How much continuous power do I need into lower resistances?
I know there should be allowances for variances in resistance over
frequencies, but I'm not sure what would be good for most 8 ohm speakers.
Some amplifiers quote the power as 50 watts into 8 ohms and 75 watts into 4
ohms. I'm not sure what the above specs mean. Shouldn't there be an increase?

Could someone give me an idea of how much power I would need. Is 50 watts
enough for most speakers. I probably will be looking for something used.


I'm trying to get an idea of what I should be looking in an amplifier. Would
appreciate help.

Thanks.

As others have noted, the amount of power you need depends on many things, including how loud you want to play your music, and the exact impedance characteristics of your speakers, as well as speaker efficiency, which can be dramatically different (you can see a difference between different speakers of 20 dB in efficiency or possibly even more; a 20 dB difference in efficiency would mean that one speaker would play at 100 watts as loud as another with only 1 watt).

In general, you want to look at continuous RMS output, at a specific impedance, over 20-20,000 Hz, with a specific amount of distortion. Of course, the peak capabilities are not going to be derived from that, but unfortunately there is no standardization of peak ratings for amplifiers, so it will be a waste of time to look at numbers for that.

Buy from a dealer with a good return policy, in case you decide that whatever you take home is not adequate. Of course, if you are buying used, you may not be able to do that. If it is practical to do so, take your speakers with you to try out the amplifier with your speakers. If not, then you need to be telling us what speakers you plan on using, and how loud you like to play it, and how big your room is. It will make a world of difference whether you will be using, for example, this:

http://www.magnepan.com/model_MG_201

or this:

http://www.klipsch.com/products/details/klipschorn.aspx

50 watts would be far more than needed with the second speaker, if one is listening in a normal living room at sane volume levels (i.e., at volumes that are literally less than deafening).

 

[emorphien]

I've had a NAD C320BEE for a few years and have finally settled on my ideal speakers after trying a couple of things (ranging from higher efficiency 8 ohms to lower efficiency 4 ohms to moderately efficient 8 ohms) and the 50W (whatever IHF rating NAD used) of the 320 has never been an issue. I rarely have to turn it up past 11 o'clock before it becomes uncomfortably loud in my smaller room.

Whether it would be enough power in a much larger room would depend on how loud I want it and how demanding the speakers are. I do expect down the road I'll be upgrading to a beefier amplifier but for the time being in the listening space I have I don't need more power.

I know NADs ratings have caused various conversations in the past elsewhere and maybe here. Some people seem to describe their power ratings as "conservative" meaning that they understate or that you get the impression they can't play with as much power as they do. That's somewhat misleading, but it does display the problem that manufacturers don't all choose to define their power ratings the same way. To one person a manufacturer might rate "conservatively" and to another they might overstate power output depending on what measures of that power the person prefers or understands.

I think NAD tends to use the IHF to try to combat the fact that various companies will often give a power rating that is not a sustained output. It's essentially a marketing tool meant to make them sound more competitive against the assumed performance* of higher rated products from other manufacturers.

Assumed performance because they might be higher rated but it's possible they couldn't sustain the same level of output as the NAD without distortion.

I noticed the same thing when I bought my receiver 6 years ago. H/Ks ratings seemed lower compared to other brands like many of the Sony or Kenwood receivers available at big box stores for example but the H/Ks would sound more powerful.

 

[PENG]

I'll just run a dedicated 30 amp circuit into the room used for my HT. Let the measurements begin.

Seriously though, I don't care if a 15 amp circuit cannot deliver. Connect the test up to a circuit that can deliver the current. I still want to see the performance of the amp. Its analgous to buying a corvette and having the car rags not test it flat out becuase the speed limit onmost highways is only 60 mph. Thats what I'm driving at, excuse the pun.

For the tests you specified you may want to try a dedicated 200A service. Let's say the subject amp we are going to test is rated for 200 WPC into 8 ohms, and therefore 800 WPC into 2 ohms. Now combined with the 66.5 phase angle (power factor=0.4) and ACD conditions you also specified, you are talking about testing for 800X7=5600W at a power factor of 0.4. Considering efficiency and voltage drop issues, you can see why a 200A service is in order. Are you still serious?

By the way, there are good reasons why those 60A industrial welding receptacles are typically for 480 or 600V systems. The currents required at lower voltage level would have been too high to be practical or economical.

 

[MidnightSensi]

Nothing wrong with running on rails as long as the current is there to deliver. Its semantics really and just different ways of speccing the output.

Nod. As long as the uppermost peaks are on the rails and not the majority of the content... that's my problem.

 

[Jfs]

Buy from a dealer with a good return policy, in case you decide that whatever you take home is not adequate. Of course, if you are buying used, you may not be able to do that. If it is practical to do so, take your speakers with you to try out the amplifier with your speakers. If not, then you need to be telling us what speakers you plan on using, and how loud you like to play it, and how big your room is.

Room approximately 12' x 15'. Lots of furniture. The speakers will be about 9' away from the listening position. Listen at moderate volumes most of the time. Not sure what I will be using for speakers yet. Plan on buying speakers as well. From what I've read speakers are around 85 - 90 db sensitivity for 1 watt at 1 meter. Is that a realistic range?

Thanks.

 

[3db]

For the tests you specified you may want to try a dedicated 200A service. Let's say the subject amp we are going to test is rated for 200 WPC into 8 ohms, and therefore 800 WPC into 2 ohms. Now combined with the 66.5 phase angle (power factor=0.4) and ACD conditions you also specified, you are talking about testing for 800X7=5600W at a power factor of 0.4. Considering efficiency and voltage drop issues, you can see why a 200A service is in order. Are you still serious?

By the way, there are good reasons why those 60A industrial welding receptacles are typically for 480 or 600V systems. The currents required at lower voltage level would have been too high to be practical or economical.

Your forgetting the point of audable distortion. From all the power curves I've read, the distortion figures raise exponentially at some point. I want all channels tested to the point just before audable distortion arises. That will severly limit current draw as well. Your assuming based on your calculations that an amp can deliver its rated power all channels driven, full bandwidth at distortion levels below audability. I'm saying, thats not the case...espcially if the loads are becoming increasingly reactive. Its one thing for an amp to handle a pure resisitive laod, quite another to handle a complex load such as a loud spaker.

 

[Pyrrho]

Room approximately 12' x 15'. Lots of furniture. The speakers will be about 9' away from the listening position. Listen at moderate volumes most of the time. Not sure what I will be using for speakers yet. Plan on buying speakers as well. From what I've read speakers are around 85 - 90 db sensitivity for 1 watt at 1 meter. Is that a realistic range?

Thanks.

Okay, you need to select your speakers before you decide how much power you need. The range you selected is certainly possible, and is very common. But you do not need an amplifier to drive some possible speaker that you are not actually going to use, so you have to select a speaker before anyone can tell you how much power you really need. Of course, if you had money to burn, you could just buy the most powerful amplifier you could find, and be sure that you had more than enough power for most speakers, but that would be quite expensive to do.

Also, you need to have enough power for the loudest you ever want to play it, not just what you will do most of the time. Of course, the speakers you select must be able to play as loud as you ever want to play them, too.

My advice is to look for speakers that you like, and try to put the vast majority of your money in your speakers. If the speakers you select are terribly inefficient or a low impedance, then you will need to rethink the matter, because that would mean spending extra on your amplification. But if you select something that is an easy impedance and of normal efficiency, for normal volumes you can use a fairly inexpensive amplifier. My advice is to not pay attention to those factors while looking for speakers, as you want the best sound possible, and impedance and efficiency have nothing whatever to do with sound quality. Of course, you do need to pay attention to those things before you part with your money, but it will only be an issue if you ended up liking difficult speakers. In my case, going for the sound I liked best (at my price point), without regard to such matters, got me speakers that happen to be an easy load for amplifiers.

In my case, I have Aurum Cantus (U.S. version) speakers and SVS subwoofers, and they retail for well over $6000. I was using a receiver that retailed for around $600 with them, and it sounded great (as it should, with such expensive speakers). I replaced the receiver with something that retails for about $1600 because I wanted more features. It sounds the same, unless I engage a feature that affects the sound. The more expensive receiver is capable of putting out about twice the power of the older one, but since my speakers are an easy impedance and of normal efficiency, I have no need of the extra power. It could play louder than I ever wanted it before.

If I were doing stereo only, I could use a very inexpensive receiver with my speakers and have good sound. Buying used, I would not need to spend even $100 on a receiver. One pair of the speakers retailed for $1500. Unless the speakers are inefficient or a low impedance, an extreme ratio like that will give you the best sound you can get for your money. No matter how much you spend on speakers, you will not be able to get a frequency response as flat as an inexpensive receiver, nor will you be able to get distortion levels of your speaker as low.

But the long and the short of it is that you need to select the speakers before you concern yourself with what the amplifier needs to be capable of doing.