Welcome, Guest. Please Login or Register
Decware Audio Forums
03/28/24 at 14:17:28 




Pages: 1
Send Topic Print
Optimal Watts for a Speaker's Dynamic Peak Needs (Read 13618 times)
beowulf
Seasoned Member
****




Posts: 1447
Optimal Watts for a Speaker's Dynamic Peak Needs
03/06/15 at 23:29:48
 
Hi all, I was just wondering ... is there a special formula that can help to determine a speaker's needs to produce dynamic peaks correctly?  

For instance on the Decware blog Steve D. is auditioning his single driver horns and stated that they needed only 10 watts to produce peaks correctly. On another forum I read that Harbeth 40.1's needed at least 40 watts to produce dynamic peaks correctly.  

How did Steve and these other people arrive at that number? Can someone explain to me ... is this a trial and error thing or is there a formula that one can use as a baseline within the speakers given specifications (such as sensitivity, impedance, etc.) of how many watts would be needed to reproduce peaks optimally for a particular loudspeaker?

I know that tube and ss watts perform differently, but just going by specs alone how do you make sure to CYA in pairing an amp to speakers?

Thanks all! Cool

Edit: Also (if this is not too redundant), what is a "dynamic peak", how is it measured and how do you know how many watts are needed to cover that for a particular speaker?
Back to top
 
 
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #1 - 03/07/15 at 16:57:56
 
-Speaker sensitivity dictates where you start from with a specified input
-Sound follows the inverse square law as it propagates through the air (a bit less so in an enclosed room)
-Power demands increase exponentially relative to spl
-Your ears dictate how loud you want it at your particular location  

http://myhometheater.homestead.com/splcalculator.html

That calculator will give ballpark ideas as to power demands, I've referenced it against my own rigs and it's fairly accurate.
Back to top
 
 
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #2 - 03/07/15 at 17:20:38
 
As for CYA when pairing speakers to amps, consider ss amps have very low output impedance, so it's more a simple matter of getting one that can deliver whatever peak power into the specific load without clipping.  Too much is just enough with ss.

Tube amps require a different approach.  Tubes clip quite differently, so absolute power, while still important, is less of a concern.  With the little zen amps, it's more about quality than quantity (although they'll still rock out when paired with speakers of high enough sensitivity).  A SE amp mildly clipping arguably enhances the sound IMO.  Another factor is that tube amps generally have higher output impedance and will thus be more reactive into a speaker load.  Because of this, you must consider tube amps and speakers as parts of a single unit.  You can make educated guesses (speakers with flat impedance curves appear better on paper), but you really have to trust your ears ultimately.
Back to top
 
 
  IP Logged
beowulf
Seasoned Member
****




Posts: 1447
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #3 - 03/10/15 at 00:06:07
 
Thanks ski bum, the calculator is interesting as well.  Are Dynamic Peaks related to wanted volume at a certain distance then?

Maybe this went over my head and I'm missing something.  But I guess I'm still not relating to how someone would know how many watts are needed for a speaker to achieve Dynamic Peaks "unless" it comes down to volume desired ... so does it just come down to how loud you ultimately want the volume to be in your room/setting?

For example does a 2 watt Zen Amp have enough watts/current to cover a dynamic musical transient peak (such as a cymbal crash) with a speaker that that is of low efficiency such as 86dB @ 6 Ohms as long as the volume is very low or would it still not cover the transient peak with enough authority?

Thanks again!
Back to top
 
 
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #4 - 03/10/15 at 03:53:09
 
If you're considering a two watt amp, it makes sense to determine it's maximum capabilities and assess if it is enough to meet one's expectations.  This should be illustrative, let's do a little thought experiment using a two watt zen amp.

Ok, so I come home from work or a hard day on the slopes, kick of my sneakers and cue up some music.  Say it's something with rather wide dynamic range, e.g. Ravel's Bolero.  Now, if you're the type with too much time on your hands, you can run the file (assuming it's digital) through an editing program to determine relative levels of the music, including the average level and any peaks.  If you're lazy, you can simply look this up in the dynamic range database.  I'm a big fan of the DR database.

So my particular recording of Bolero has 20 db of dynamic range (an extreme case).  Now, given my speakers sensitivity of 95.5 db/w (measured, genuine sensitivity, not some inflated manufacturer spec), in my small room, and with corner placement, plugging that info into the calculator indicates I can get ~98 db at the listening position.  So two watts equals 98 db, the peak or maximum output the amp can deliver to me at my chair.  The question at that point is, "Is this enough?"  

Applying that to my Bolero listening session and working backwards from the peaks on the program material, it means I can listen at average levels of 78 db and have adequate headroom to hit the peaks cleanly.  

If I play RHCP or something horribly compressed, the dynamic range is probably a paltry 5 db or so, I can listen at an average level of 93 db and still have enough amp headroom to cover the peaks in the recording.  The recording itself may sound like arse, but that's another story, this is just to demonstrate required power for dynamic peaks in the music.

That's about it when considering power as the sole consideration.  But of course, the subject of this little thought experiment being a tube amp, it gets even more complicated.  

Back to Bolero, if I listen at average levels above 78 db, those times when the recording calls for an extra 20 db exceed the amp's limits and the amp clips.  These amps are characterized by a low order, monotonic pattern of harmonics when over-driven.  Within the early stages of being over-driven/clipped, this sort of distortion pattern tends to slip right by the ear.  As it increases it tends to act as an ideal compressor (the amps electrical output plateaus out, but the harmonics within that envelope increase, adding bite/edge/presence and which is perceived as subjective loudness...this is largely responsible for the 'tube watts are more powerful' meme).  In this case, and provided I don't drive the amp into gross/excessive clipping, which sounds bad any way you slice it, I can push the average levels considerably higher; the peaks are clipped, but not objectionably so.  Is it any wonder these amps seem to bring out the low level details in recordings so well?  When run slightly in the red, they literally do bring all that low level info up in the mix, while not causing objectionable distortion on the clipped peaks.

Now, lets move the rig into a larger room, with more space to fill and greater distances.  In this more demanding situation two watts becomes a severe limit.  Peak capability will only get me to ~93 db at my chair.  Still fairly loud, but not soul stirring loud.  Not enough for "correctness" or "authority" -in the terms you asked through my interpretation.  Much easier to drive the amp to gross clipping in this case.  The answer? More sensitive speakers, or more power.  That would go for the small room situation if 98 db peak output was unsatisfactory.

In my own case, I generally listen at average levels of 85 db or so, and like rock, blues, jazz.  Typically, on a good recording, such music has a dynamic range of 10-15 db, so two watts in the small room mentioned above is right on the margins but acceptable, and inadequate in the larger room.  

Does this help?



Back to top
 
 
  IP Logged
beowulf
Seasoned Member
****




Posts: 1447
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #5 - 03/10/15 at 07:35:11
 
Actually ski bum, this is an excellent example and exactly what I was looking for in terms of an explanation.  I also appreciate the extra comparison between tube and SS watts.  Overall I really think I have a good understanding of how much power is needed for a certain level of listening and dynamic transient peaks.  So yea, it helps a lot! Cool

Just out of curiosity, what amp and speaker combo are you using?  (I may have read this somewhere at one time but I don't recall right off hand)
Back to top
 
 
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #6 - 03/10/15 at 08:05:40
 
My zen friendly speakers include some TB W8 in mltl cabs, my old Klipsch forte II w/ crites networks, and on deck (haven't pulled the trigger yet) are a pair of SEOS Tempests.  I'm looking forward to the Tempests, should be even better than the fortes, more than 3 db more sensitive and smooth, smooth, smooth response, although they don't dig as deep.

I tend to like horns the most, as they have a fighting chance of lifelike dynamics.
Back to top
 
 
  IP Logged
jsm71
Seasoned Member
****




Posts: 134
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #7 - 03/10/15 at 18:25:15
 
I'd like to offer additional food for thought.  Sensitivity is a gauge of SPL at a given wattage level.  There is way more to this than that alone.  The speaker offers up some degree of difficulty for the amp more or less based on the speaker's design parameters.  Example:

Let's say two people can lift 100 pounds over their head with similar ease, akin to two speakers rated at the same sensitivity.  Now the task is to walk 10 feet forward.  One person gets a nicely paved road, while the other has to walk on a fine sand beach.  Who works harder?

Speakers with wild impedance swings, complex or multiple crossovers, wimpy magnet structures, or other challenges will not perform as well as a speaker lacking these hurdles, even if they have the same sensitivity rating.

My current and prior speakers only varied in sensitivity by 1 dB but my current ones are way easier to drive and achieve excellent performance with far less power.  I know this both anecdotally using the same amp on both and observing, but also by knowing the basics of each's design and materials used.

So I have no model to propose, but if you know the speakers have aspects that will challenge the amp aim for more power, you'll likely need it.
Back to top
 
 

Decware ZMA, Cary SLP-98P preamp, JansZen zA2.1 speakers with JansZen speaker cables, Marantz TT 15S1 turntable, Lyra Delos MC with Bobs Devices SUT, Marantz SA8004 SACD/CD player, Morrow level 4 ICs, Decware and Shunyata PCs.
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #8 - 03/10/15 at 20:27:23
 
I think this is why we ultimately must let our own ears decide.

On paper, my full rangers would clearly be the best match for my SE84 amps.  I use bsc on them, which runs counter to the general Decware approach to speaker design, but to me improved octave to octave balance is preferable to unfiltered drivers directly to the amp.  The speakers are still ~94db/w, and the impedance is almost ruler flat.  And, go figure, with an se84 amp, I don't like them as much as the Klipsch, which on paper appear to be a much worse match (save sensivity).  Even more confusing, the full rangers sound fabulous on sand amps. Huh

Back to top
 
 
  IP Logged
beowulf
Seasoned Member
****




Posts: 1447
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #9 - 03/11/15 at 07:59:33
 
@ ski bum - the SEOS Tempests look like they will be pretty cool!  I have always thought about getting a pair of CornScalas myself.  I wouldn't know which one to get out of the 5 different models he has but my gut tells me that style D with the Faital drivers would be the most coherent.

@ jsm71 - thanks for the additional feedback I hope to someday hear the JansZen zA2.1s at some time.  I can imagine them sounding quite good in combination with the ZMA.  Just out of curiosity have you heard the stand mount zA1.1 (although they seem like they may be a little tougher to drive)?
Back to top
 
 
  IP Logged
jsm71
Seasoned Member
****




Posts: 134
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #10 - 03/12/15 at 13:11:23
 
beowulf - I only heard the zA1.1s briefly at Axpona but it was not an optimal setup.  Because the cabinets are smaller but also sealed my guess is there is more compression going on inside and thus the slightly lower sensitivity.  This is typical of small stand mount speakers that use a sealed approach.
Back to top
 
 

Decware ZMA, Cary SLP-98P preamp, JansZen zA2.1 speakers with JansZen speaker cables, Marantz TT 15S1 turntable, Lyra Delos MC with Bobs Devices SUT, Marantz SA8004 SACD/CD player, Morrow level 4 ICs, Decware and Shunyata PCs.
  IP Logged
Palomino
Seasoned Member
****




Posts: 2474
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #11 - 03/12/15 at 22:50:58
 
Yes tough to get a feel for these speakers at Axpona.  Something was off because I didn't like anything I heard in that room, yet I sensed from others they could be special.
Back to top
 
 

i7 Mac Mini with LPSU/SSD running Audirvana 3.5, Uptone Audio Regen on LPSU, Mytek Brooklyn DAC+, Ven Haus DIY Silver ICs, 25th Ann. Zen, PS Audio P5 Power Supply, PS Audio Power Cords, GR Research Speaker Wire, DIY Big Betsy and Crystal 10 Open Baffle Speakers
  IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #12 - 03/17/15 at 00:25:36
 
Hi everyone,

I thought this might be a good thread to drop a little info bomb into about our tube amp's power ratings...

I'll use our 2 watt SE84CKC for the example.

When the amp is hooked up to a 1KHZ reference sine wave on it's inputs and to an 8 ohm resistor on it's speaker jacks (the standard way to measure power) it puts out around 2 watts RMS and 4 watts peak.  When you replace the sine wave of a single frequency with pink noise which is all frequencies at the same time, the RMS power drops to 1.2 watts and the PEAK power goes up to 6.8 watts.  Pink noise is closer to actual music than is a single sine wave.  However it's only closer, it's still not music.

Keep in mind that the above power figures are meaningless because the amp is hooked to a resistor not a complex loudspeaker.

If we hook it to a loudspeaker, the 1KHZ sine wave give us an RMS of 2.8 watts and on the other end of the example, we get over 10 watts into the loudspeaker in PEAK power with pink noise.  
With distortion the PEAK power jumps to 15 watts.  Again, with pink noise.  Music is not pink noise.  With the huge dynamics in music these figures go up substantially.

When we also consider that unlike a solid state amplifier, the output transformers block back EMF from the loudspeaker, we have an amplifier that is many times more capable of driving a loudspeaker..  Solid state amplifiers hate back EMF because it reduces power output and increases distortion.  

I wrote a paper called "Usable Power" in the articles section of the site  that applies here.  The little zen has nearly 100 % useable power, while a 400 watt solid state amp may only have 10% useable power.

And then finally, there's the impedance issue.  A zen amp can drive nearly a dead short.  Years ago a customer hooked a pair of Klipsch horns, a pair of JBL L100's and a pair of JBL Control 10's to his Zen amp all at the same time only to be recently beat by a customer this winter who actually had 4 pair of speakers in parallel on each channel.  Try that with a solid state amp, but open your windows first.

So you see, there really is no locked in formula or put another way, reality can often upset predictions.

Steve   Smiley



Back to top
 
 
WWW   IP Logged
Dave1210
Seasoned Member
****




Posts: 959
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #13 - 03/17/15 at 01:06:21
 
Steve.   I think I probably speak for everyone when I say that we greatly appreciate your activity on this forum.

Sounds like transformers offer multiple benefits…blocking back EMF, delivering more instantaneous power vs. SS.  Are there other benefits?  

Also, why is your OTL the most transparent tube amp on the planet? The best transformer is no transformer at all?  Or is it not that simple?
Back to top
 
 
  IP Logged
ski bum
Seasoned Member
****




Posts: 195
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #14 - 03/17/15 at 20:37:09
 
Thanks for the insight, Steve.  Goes to show that power ratings can be misleading, and that my sort of guestimation/calculation described in previous posts are just that, guestimates.  It's nice to know there is more dynamic power available, as the calculations end up being conservative.  That helps explain why I'm able to get triple digit spl peak readings before things fall apart, and also why my little zens seem to cope so well with heavy handed eq boosts.

Maybe a parafeed style zen amp, so I don't have to apply bass equalization would be the ticket...just thinking out loud.  I suspect you, Steve, have tried just about everything under the sun, any feedback?
Back to top
 
 
  IP Logged
hdrider
Seasoned Member
****


On the path with
Decware.

Posts: 693
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #15 - 03/17/15 at 21:26:16
 
Just like Moses coming down off the mount with the laws. we have Steve shining light on our discussions. I just love this forum and the people who participate on it. Every time I listen to my Rachael with CD's or my TT into my ZMC1/ZP3/Rachael, I just get this stupid butt grin on my face and laugh like a madman. I am SO happy. Thank you all @ Decware and the forum. Happy listening, Chris.
Back to top
 
 

CSP3, Rachael, ZP3, ZMC 1, Sony ES Bluray, Schitt Bitfrost DAC,Sota Star TT w/ FR-64, Hana SL Low Output Moving coil, Omega 7XRS Mk 2 walnut, DeepOmega 8 sub, Morrow Audio PH3 and MA3 IC's, Gamma Electrostatic phones.
  IP Logged
Ace-Tone
Seasoned Member
****


8+ years with
Decware gear and
Loving it!!!

Posts: 775
Re: Optimal Watts for a Speaker's Dynamic Peak Needs
Reply #16 - 03/17/15 at 21:31:19
 
I second that emotion hdrider..
Back to top
 
 

SE84UFO25, ZP3, Zstage, ZR2, ZSB, DM947, HERESY iii
VPI-Traveler2 +SS Zephyr MKIIIes.
Cambridge CXC. Schiit Bimby DAC
IC's Audio Envy, Decware, RAC-PC's, Mapleshade Dbl Helix+Spkr wires & jumpers
Blue Circle PLC FX2X0e, Herbies, GIK, PI Audio DIY treatment
  IP Logged
Pages: 1
Send Topic Print