How important is doubling down power to 2 ohm or 1ohm on Hi Fi Speakers on high dynamic range music?

[witwald]

From the impedance plot, the magnitude of the impedance at 50 Hz is magnitude about 20 ohms. That means that not a lot of current is required. The fact that there is a phase shift of about -50 degrees at 50 Hz just means that there will be a phase difference between the voltage output and the current output. This means that when the voltage is zero, the amplifier will still be supplying quite a bit of current, as the peaks in the current output are not in phase with the peaks in the voltage output. The magnitude of the current that amplifier produces is computed using |I| = |V|/|Z|, where Z is the complex impedance. As |Z| at 50 Hz is relatively high, the current that is produced will be relatively low, compared to say an impedance |Z| = 8 ohms.

 

[ahofer]

From the impedance plot, the magnitude of the impedance at 50 Hz is magnitude about 20 ohms. That means that not a lot of current is required. The fact that there is a phase shift of about -50 degrees at 50 Hz just means that there will be a phase difference between the voltage output and the current output. This means that when the voltage is zero, the amplifier will still be supplying quite a bit of current, as the peaks in the current output are not in phase with the peaks in the voltage output. The magnitude of the current that amplifier produces is computed using |I| = |V|/|Z|, where Z is the complex impedance. As |Z| at 50 Hz is relatively high, the current that is produced will be relatively low, compared to say an impedance |Z| = 8 ohms.

As I asked in the other forum in a different way - the amp meters probably only measure apparent power. The overstatement of apparent power to real power will be highest right at the steepest phase angle, which is at the dominant frequency of this track, right? If I have this right (and no guarantee that I do) the meters would significantly overstate (real) power delivery on this particular track.

My limited references on this: https://www.pema.ie/TECHTALK LIBRARY/APT/True vs. Apparent Power.pdf
http://www.cromptonusa.com/Watt_Var_VA.pdf
https://www.audioholics.com/loudspeaker-design/understanding-impedance-electrical-phase/page-2

If any good samaritan AC power experts want to straighten me out, I’d appreciate it, but this seems to cast some doubt on amp power meters to determine peak power, esp. near crossover frequencies. (@amirm ?)

 

[pjug]

As I asked in the other forum in a different way - the amp meters probably only measure apparent power. The overstatement of apparent power to real power will be highest right at the steepest phase angle, which is at the dominant frequency of this track, right? If I have this right (and no guarantee that I do) the meters would significantly overstate (real) power delivery on this particular track.

My limited references on this: https://www.pema.ie/TECHTALK LIBRARY/APT/True vs. Apparent Power.pdf
http://www.cromptonusa.com/Watt_Var_VA.pdf

If any good samaritan AC power experts want to straighten me out, I’d appreciate it, but this seems to cast some on amp power meters to determine peak power.

I wonder how they measured power too. When I tried the same track on my 8 ohm, 87dB speakers playing about as loud as I could stand it, peaks were less than 30V (measured with an oscilloscope).

On the other hand I have tried more dynamic tracks where peaks were much higher, so I would have expected such a demonstration to use extremely dynamic music to prove the point. Somehow I think there was some gaming in that video and the power measurement, with that weird track.

 

[Spkrdctr]

I agree with some of the comments on here. That music was picked especially for the demo.

1. Heavy bass content with no subwoofer meant that it all had to come from those amps.
2. The volume was turned up very high as the room was quite large, probably 30% or so larger than most listening rooms at home.
3. The second track of violins really showed what I am trying to point out. It had very little bass and they were hitting a peak of 28 watts at a very high volume on the preamp.
4. In reality, most people who want accurate bass in large amounts get a quality subwoofer. This will reduce a good portion of the bass and deep bass from the amplifiers. That alone would have dropped those 750 watt peaks by a good bit. It may have shown peaks of only 500 watts which is a big drop. But we will never know as that was not the point of the test.
5. I want to comment again as it can't be overstated. A large room and high volume was used. As I believe he stated it was much higher than anyone would listen to normally. Those violins were screeching like crazy and now your in hearing damage territory. But I can't really tell for sure as an important technical point was left out. What was the listening SPL range? It doesn't have to be specific, just a range. If your in the 60 to 70 range you know that, if in the 100 to 105 range that is something you can also tell with a cheap meter. But to have no idea at all was a let down of the "test". I am making an assumption as 500 watts RMS, 750 watt peaks into most speakers will be the max (or over max) they can play at over the long term, they are beating the drivers pretty hard. I have tested drivers to failure many, many times. That was almost a torture test and as torture tests go a very short one that didn't cause any problems. After a few hours they may (we will never know) have let the smoke out of the speaker box!

To get me to respond like this shows that this is a fantastic thread! Very enjoyable.

 

[ahofer]

As I asked in the other forum in a different way - the amp meters probably only measure apparent power. The overstatement of apparent power to real power will be highest right at the steepest phase angle, which is at the dominant frequency of this track, right? If I have this right (and no guarantee that I do) the meters would significantly overstate (real) power delivery on this particular track.

My limited references on this: https://www.pema.ie/TECHTALK LIBRARY/APT/True vs. Apparent Power.pdf
http://www.cromptonusa.com/Watt_Var_VA.pdf
https://www.audioholics.com/loudspeaker-design/understanding-impedance-electrical-phase/page-2

If any good samaritan AC power experts want to straighten me out, I’d appreciate it, but this seems to cast some doubt on amp power meters to determine peak power, esp. near crossover frequencies. (@amirm ?)

I talked this through with one of my engineer sons. He agrees that the meters are probably just showing apparent power and the real power delivered is likely much lower. He noted that he has the same issue figuring out power usage of his heat pump.

 

[Jukka]

How does one know whether the Amp has good peak or how high of peak is needed?
Does the doubling down infer higher ability of peak power than non-doubling down amp?

Good point. Although I don't have an answer, I raise you a counter question: wouldn't it be smarter to favor higher impedance speakers?

Nowadays we get hundreds of watts from small and cool amps, why not use higher impedance speakers, as it should improve efficiency of amplification.

As a guy who likes DIY, favors big pro drivers and has a small d**k, I'm going to try this with my next speakers, which are going to be fully active with 16 ohm drivers (if I can find suitable). The wattage is going to drop (which means energy savings), but having dedicated amps for each driver mitigates the issue and higher impedance versions tend to be more sensitive. The big boys also say that higher impedance versions of the same drivers sound better all else being equal. -EDIT- also larger drivers are more efficient to a point (better acoustic coupling) and have better natural bass extension --> less eq is needed --> less phase angle changes.

 

[pjug]

I talked this through with one of my engineer sons. He agrees that the meters are probably just showing apparent power and the real power delivered is likely much lower. He noted that he has the same issue figuring out power usage of his heat pump.

The CH Precision manual says they measure power into the load:
1.2.1 Power monitoring
Each power amplifier board is equipped with a DSP that monitors the instantaneous output voltage and current delivered to the
loudspeaker. Both values are sampled at around 100 kHz, ensuring peak values are properly detected.

However, JA measured a CH amplifier and it seems the peak power meter is not very accurate:

From https://www.stereophile.com/content/ch-precision-m11-power-amplifier-measurements

When I started testing the amplifier's output power, I fed the balanced input at 1kHz tone at 188mV, which resulted in exactly 1W into 8 ohms; the M1.1's meter read 2.3W

The display is peak power, which should have been 2W, not 2.3W. Just 15% error in this case but I wonder what kind of error there might be at much higher volumes.

edit: Stereophile footnote says:
Footnote 2: CH Precision says the meter on the M1.1 shows peak power, not RMS power; the rest of the discrepancy they attribute to a measuring technique that avoids adding components in series with the output signal.

So the current measurement uses current mirroring or something instead of a sense resistor. I suspect there is some problem with this as there are other videos where these CH amps show insanely high power levels. I really would like to see someone put a scope on the output of one of these amps to see if the displayed peak power numbers are credible.

 

[MakeMineVinyl]

So the current measurement uses current mirroring or something instead of a sense resistor.

We use a special Hall effect sensor to measure current. I believe several other manufacturers use this device also.

 

[pjug]

We use a special Hall effect sensor to measure current. I believe several other manufacturers use this device also.

Thanks. Do you have any thoughts on the CH power numbers where videos show many hundreds of watts peak even with what seem not to be overpowering SPL?

 

[MakeMineVinyl]

Thanks. Do you have any thoughts on the CH power numbers where videos show many hundreds of watts peak even with what seem not to be overpowering SPL?

I'm not familiar with what CH is doing - too focused on what I'm doing.

 

[ahofer]

e CH Precision manual says they measure power into the load:
1.2.1 Power monitoring
Each power amplifier board is equipped with a DSP that monitors the instantaneous output voltage and current delivered to the
loudspeaker. Both values are sampled at around 100 kHz, ensuring peak values are properly detected.

Sounds like apparent power-instant voltage and current multiplied.

 

[MasterApex]

So if my speaker is a complex load (not efficient , low impedance, high phase angle, etc) and I want to be able to get the full dynamic of any music out of my speaker, what is the spec I should look for in choosing Amplifier?
Double down = more headroom ?

 

[audio2design]

I experimented with Levinson 333 300W@8ohm , doubling down to 1200W @2ohm , Krell KSA 200s 200W @8ohm that double down to 1600W @ 1ohm vs a few highly rated class D 200W - 400W @8 ohm amplifiers.

On my B&W N801 speaker, the sound quality is comparable below 75dB....do show different "brightness" and "bass punch" which can be subjective in term of preference.

But when playing high dynamic classical music like 1812 overture at peak (not sustained) volume above 94dB, the Class D amplifier sounds compressed relative to Levinson or Krell.

Why is that?

OMG! --- have you been running into GeorgeHiFi over at Audiogon?

It is not important. All that matters is your amp is stable and not clipping, whether voltage or current.

Oh, and phase-angle/ EPDR (was that brought up?) only matters w.r.t. thermal loading of primarily BJTs. Tube and MOSFETs don't experience the same SOA issues of BJT's, and Class-D don't run in the linear region, hence they don't have similar significant thermal issues.

Your KRELL can't do 1600W continuously into 1ohm, and realistically, how many speakers hit 1 ohm?

Class-D amplifiers will generally hard clip on current. 1812 Overture (Telarc?) is not normal music. The dynamic range is extreme. I am not sure when you are measuring that 75db, but from the more quieter peaks to the cannons, the jump is significantly more than 19db. What you are likely hearing comparing the two is performance into clipping, but then again, maybe your speakers are flapping in the wind the Class-D is compressed but more accurate? I know with that piece, played properly, it is crisp, but in a less "capable" system, it comes across louder, but not well defined.

 

[audio2design]

50Hz is right at the phase angle extreme. I'm not expert in these things, but I think past +/-45 degrees will draw a ton of current?
View attachment 128351

Current draw is define by the impedance. The lower the impedance, the higher current. Where phase-angle comes into play in an amplifier is thermal loading of the output devices, which is mainly only an issue for BJT output devices. It does impact MOSFETs and TUBES, but they don't experience the catastrophic devices failures of BJTs. Class-D do not operate in the linear region, so phase angle has far less impact on device thermal loading.

 

[MasterApex]

OMG! --- have you been running into GeorgeHiFi over at Audiogon?

It is not important. All that matters is your amp is stable and not clipping, whether voltage or current.

Oh, and phase-angle/ EPDR (was that brought up?) only matters w.r.t. thermal loading of primarily BJTs. Tube and MOSFETs don't experience the same SOA issues of BJT's, and Class-D don't run in the linear region, hence they don't have similar significant thermal issues.

Your KRELL can't do 1600W continuously into 1ohm, and realistically, how many speakers hit 1 ohm?

Class-D amplifiers will generally hard clip on current. 1812 Overture (Telarc?) is not normal music. The dynamic range is extreme. I am not sure when you are measuring that 75db, but from the more quieter peaks to the cannons, the jump is significantly more than 19db. What you are likely hearing comparing the two is performance into clipping, but then again, maybe your speakers are flapping in the wind the Class-D is compressed but more accurate? I know with that piece, played properly, it is crisp, but in a less "capable" system, it comes across louder, but not well defined.

I don't know GeorgeHiFi. was just trying to find objective measurement that can explain my/our observation of the reduced dynamic range (sound compression).

There is no doubt there is very good Class D Amp. Just stating in our experiment, the class D Amp (brand P....) is inferior in term of reproducing the dynamic, sounds less life like compare to Levinson 333 or Krell KSA200s.

Yes, 1812 overture (telerac) is not a typical background music. It will sound aweful on Sonos or entry consumer amp/speaker system because its huge dynamic (the soft was will not be very audible and the loud part could cause distortion).

When I do active listening , I like playing the 1812 overture. Its recording demands good amp and speakers to reproduce the life like presentation. The SPL meter set of dBC (C-weighting) does go from low 50dB to 98dB (peak).
When it does, it made me feel like I was listening to life presentation and justified the extra $ spent on Amp and speakers.

A lot of speakers sounds good (including my Sonos that I listen to a lot) but its sound is clearly coming from a box.
My personal preference for my HiFi system is to reproduce life like presentation (soundstage, etc).

 

[audio2design]

You may not like this, but the odds are, your B&W 801's cannot do the 1812 overture by Telarc justice. Most speakers cannot, not matter how much power you give them. This is one time that efficiency (and serious subs) really does come into play. It is called an 8 ohm speaker, but is much closer to 5 or 6 ohms, and the efficiency is 91db/2.83V, not 1W. Really it is closer to 88db/w. In the lower base it is closer to 6-8 ohms, so the Krell is peaking at 200-300W, and down to 3 ohms in high bass, so the Krell is peaking at 500W. Its sounds like a lot, but your speakers while not inefficient, are not efficient. The first time I heard a system do the 1812 overture justice was a custom set of massive horn loaded speakers with equally massive subs. Total power was several thousand watts and efficiency likely in the high 90's . It was the first time I heard the cannon come across as anything truly life life. After that, I realized how hard most people were pushing their systems playing this and how terribly distorting they were. That "compressed" version I would not be so certain are wrong. Woofer flopping is not how this piece actually sounds. I have heard this sound better with relatively cheap electronics and really sensitive speakers compared to uber expensive systems that just did not have the "oomph" everyone seems to turn it too loud for their system.

 

[MasterApex]

I am pretty sure there are many better speakers than my prior N801
I am hoping to learn what spec / objective measurement that may explain the dynamic sound differences among the amplifiers I tested.

Compressed sound does have its use....I enabled Dolby Volume or DRC in my other living room when playing movies (different amp and speakers, cheaper ) so that the dialogue is more audible and the explosion is not annoying people who were there for passive listening.

 

[audio2design]

If you have test equipment, you can measure both output of the source, amplifier, even the speaker output and compare them to understand what is happening. Otherwise we are both guessing unfortunately. The N801 is great, but not for playing insane volume peaks. Other speaker can play those peaks better, not will fail at the 99.9% of the other stuff the N801 does well.

 

[DSJR]

After spending a wonderful afternoon many years back with Harbeth 40.1's (good company, great music and no restriction on playback volume except the amp), I can tell you that speaker model eats power like you wouldn't believe, or appears to because I found the sound clean and unstressed, rather than 'HiFi Edgy' as some of its dearer competition from Europe can be. The amp used was a vintage 200WPC (300WPC or so into 4 ohms) and the 'IOC' lamps which compare input and output signals and illuminate when the distortion starts to climb were coming on regularly. The sound wasn't perceived as loud until I nearly had to shout to my pal in conversation. It certainly wasn't apparently compressed either. I only mention it as in a large room with plenty of people in it, I suspect 400W peaks were quite achievable in reality, especially in the bass.

I just don't get B&W these days. The M801 was quite good as I remember but the N versions seem to put style way over ability despite the hi-tech which goes into them. the one occasion I heard N801's I thought them needing an upper mid driver as it was all bass and sparkle but nothing in between - flutes sounded extremely odd to me and this was in quite a large space too so not strictly room limited (I can't remember the rest of the system).

 

[Balle Clorin]

Regarding Power meters.
Many meters -Accuphase analog needle is one example- Simply reflect the output voltage, but the scale is converted to watts at 8 ohm. So the meters are a voltmeter , not a power meter.
Accuphase LED bar meters show the true power if I read the manual correctly.