Does a high-pass filter lower the needed power for a speaker?

[jfburk]

I'm looking at getting an outboard amplifier for my front three speakers so I can add some other zones off my receiver. They're Ascend Sierra 2-EX's and the Horizon center.

All three speakers are crossed over at 80Hz, so does that mean I can opt for a lower power amp than I might otherwise? In other words, does crossing them over allow me to get the same output from, say, a 100 watt amp as I would running them full range with a 200 watt amp (setting aside what is happening below 80Hz)?

 

[Blumlein 88]

Probably a yes to that.

 

[Tim Link]

The Siera 2-EX never drops below 6 ohms, so should do fine with lower power amp, especially when crossed over at 80hz.

 

[RayDunzl]

I'd like to see some measurements to support the idea.

If I look at a peak spectrum of a "typical" modern recording, it would seem to demand the same voltage extremes both above and below 80Hz, at least up to 1kHz or so.


Steely Dan - Everything Must Go - Track 1

Power = the voltage squared divided by the impedance

If the impedance is the same, and the apparent voltage swings are the same, where's the power difference?

(I get to be wrong, so have at it)

 

[LTig]

I'd like to see some measurements to support the idea.

If I look at a peak spectrum of a "typical" modern recording, it would seem to demand the same voltage extremes both above and below 80Hz, at least up to 1kHz or so.
[..]
If the impedance is the same, and the apparent voltage swings are the same, where's the power difference?

(I get to be wrong, so have at it)

I fear you are - wrong, I mean. The spectrum does not tell you the amplitude in the time domain (TD), which is the sum of all amplitudes in the frequency range. If you cut off a part of the frequency range the amplitude in TD is reduced (ergo voltage swing and power).

 

[Blumlein 88]

Steely Dan Two against nature. Upper is 80 hz and lower. Lower is everything above 80 hz. Surprisingly the higher frequency side is at a 5.47 db greater RMS level. Nevertheless you are reducing the total power on the speaker even if the majority is still there. If these levels were equal you'd reduce power by 50%, but in this case the reduction is less than that. Momentary peak power is about equal for both halves.

 

[RayDunzl]

Explain it harder, please.

I see lower voltage on the bass, higher voltage on the treble, above.

(brain wants to class this a Urban Legend or similar)

 

[RayDunzl]

What's the RMS for the original track?

 

[RayDunzl]

Seems like the above 80Hz material requires more voltage after the split, ergo, a more "powerful" amplifier could be needed.

(I'm so wrong...)

 

[Trouble Maker]

Seems like the above 80Hz material requires more voltage after the split, ergo, a more "powerful" amplifier is needed.

(I'm so wrong...)

I think you are thinking BlueMelon did Upper Graph=Full Range and Lower is above 80Hz.

Upper is 80 hz and lower. Lower is everything above 80 hz.

But I think they did Upper is below 80hz and Lower is above 80hz
The idea would be if they were equal, half of the power is needed for either of them e.g. only half power amp for 80hz and up.

I'm going to go down some quick assumption rabbit holes that I haven't full explored, but I think intuitively I'm on the right track.
Of course most (logical, science based) people don't want to spend more than they need to. But if one is ok with class D (should be), those watts are cheap and incrementally very chap. I don't see much reason to get let's say something based on a Hypex NC122 over a NC252. For most people the NC122 would probably be enough for these speakers but maybe not enough for their next speakers that may be much more power hungry. The NC252 would probably be enough for speakers that 90%+ of even people on here end up with. For example at March Audio the NC122=$725, NC252=$895 and NC502=$1075. Only $170 more to probably future proof your amp from a power perspective. If you end up with a huge space with some monsters you probably have enough money to eat cost of that NC252 at that point or re-purpose/sell it and get something bigger like a pair of NC1200s or NC2sk. There is also a possibility that we are seeing commoditization of clean powerful amps right now with these products. So it's also possible that 5 or 10 years later that equivalent to the NC252 or larger cost much less than now in relative dollars. Now if one feels they want or need a more expensive amp topology that might be a different story.

If they only target is amps for zones at the lowest cost possible, I think there are cheaper ways to do it. I've been going down this rabbit hole right now planning my new surround sound system and 2 other zones.

 

[RayDunzl]

I think you are thinking BlueMelon did Upper Graph=Full Range and Lower is above 80Hz.

Using 80Hz 24db filter:

Original, High Pass and Low Pass filter applied.

The High Pass display is clipped - needs to lose 5.241dB to not clip.

 

[RayDunzl]

Attenuate the original by -6dB to eliminate the clipping in the High Pass remnant:

Original, High, and Low Pass:

 

[Trouble Maker]

I was actually wondering about that, and I think you alluded to it earlier about RMS vs Peak. Your idea is Peak VS RMS for Blumelin's analysis right?

 

[RayDunzl]

The idea would be if they were equal, half of the power is needed for either of them e.g. only half power amp for 80hz and up.

The prime determinant for an amplifier's power rating is the voltage it will swing across a specified load.. It will need to provide current to maintain that voltage across the load presented to it.

If the high-passed signal into the same load calls for more voltage, that indicates the need for a "higher powered" amplifier for the high-passed material.

 

[RayDunzl]

A confounding factor:

My speakers can be bi-wired.

They are rated as 4 ohm speakers, and measure as such (statically) when jumpered.

When separated (jumper removed), each half measures 8 ohms.

---

So that would split the power requirement for bi-amping by half (ohms doubled on each leg, current halved), but still, the high-pass could need a relatively "more powerful" amplifier to accommodate the higher voltage swings.

Whether or not the high-passed drivers consume more "power" than the low-passed woofer is still debateable, but it would appear the higher voltage amp should go on top.,

 

[pjug]

A confounding factor:

My speakers can be bi-wired.

They are rated as 4 ohm speakers, and measure as such (statically) when jumpered.

When separated (jumper removed), each half measures 8 ohms.

---

So that would split the power requirement for bi-amping by half (ohms doubled on each leg, current halved), but still, the high-pass could need a relatively "more powerful" amplifier to accommodate the higher voltage swings.

Whether or not the high-passed drivers consume more "power" than the low-passed woofer is still debateable, but it would appear the higher voltage amp should go on top.,

I don't think HP filtering will always cause higher peaks; peaks can be greater or smaller after filtering. So I think you are right that the answer to the OP question is no, but then don't go the other way and say you need more power.

 

[RayDunzl]

So I think you are right that the answer to the OP question is no, but then don't go the other way and say you need more power.

Too late!

High Pass - peak negative -0.8dBfs

Low pass - peak negative -9.8dBfs

Difference 9dB

The high-pass amplifier needs to swing 2.82 x the output voltage of the low-pass amplifier.

If the low-pass amp is rated 200W, the high-pass amp should be 1590W.

Low Pass:

High Pass:

 

[pjug]

Too late!

Ha ha. Now try it with this one. I think the high pass reduces the peaks.
http://www.lindberg.no/hires/test/2L-145_01_stereo_01.cd.flac

 

[RayDunzl]

I don't think HP filtering will always cause higher peaks; peaks can be greater or smaller after filtering.

Ok. Let's see.

Another handy sample:

Pipe Organ with at 16Hz pedal.

 

[RayDunzl]

Ha ha. Now try it with this one. I think the high pass reduces the peaks.
http://www.lindberg.no/hires/test/2L-145_01_stereo_01.cd.flac

Think again.