Go Back   Home > Forums > Blogs > abraxalito

Rate this Entry

What is dynamics?

Posted 5th March 2013 at 03:02 AM by abraxalito
Updated 11th March 2013 at 06:35 AM by abraxalito

Fascinating discussion going on about the nature of 'dynamics' over on this thread : http://www.diyaudio.com/forums/multi...nger-what.html

Pano writes (post 1174):

There is a consistency across several decades and cultures that high efficiency speakers sound more dynamic than low efficiency. That's a common subjective assessment, so we might ask "why?"

I'm not going to chip in on that thread because I have bigger fish to fry, but ISTM they're all barking up the wrong tree. That's because the question itself isn't quite posed correctly even though its a great start. Nobody listens to speakers alone, they're always powered by amplifiers. The answer to dynamics lies in amplifiers, not speakers. Put another way 'Its the electronics, stupid!'.

Higher efficiency speakers tax amplifiers (here meaning their power supplies in the main) much less - so there's no way to do an apples-apples comparison between a high efficiency and a lower efficiency speaker, because amplifiers get noisier the more power they put out. Its the absence of the amplifier noise (noise modulation) that gets heard as 'dynamics'.

As an example of a speaker that lacks dynamics, they do seem pretty much agreed the Orion is it. Well no surprise there - the Orion in its native form is powered by LM3886 chipamps with ultra-wimpy power supplies. Nuff said

Interesting article here : http://www.audioholics.com/education...f-an-amplifier which covers a lot of ground but almost totally misses the elephant in the room. I say 'almost' because on the 2nd page there's a bit where its noted Julian Hirsch says that very small changes in SNR can be very important.......

<update> twest820 has recently flagged up an earlier post of his which seems to be hinting he's come to the same conclusion. He's aiming for a PSRR (mains fed supply) of 140dB : http://www.diyaudio.com/forums/power...ml#post2790519

Added recording of the Nitro's power supply noise playing Beethoven - music at the top attenuated 50dB wrt the bottom (the psu noise). This on a chipamp with flat PSRR around 60dB over the audio band. Here the total supply capacitance is close to 0.1F for the two channels, made up of over 30 10uF ceramics, 8 low esr 390uF and 42 2,200uF. This large array of caps is being fed by an LM338K linear regulator.
Attached Thumbnails
Click image for larger version

Name:	nitro-psu-noise-6hexa.png
Views:	353
Size:	8.6 KB
ID:	909  
Posted in Uncategorized
Views 1429 Comments 11
Total Comments 11


  1. Old Comment
    Agreed. Fascinating that people are able to so adroitly separate contributions to the end sound into nice parcels like that -- a little bit of lazy thinking goes a long way!

    And, most appropriate timing: my poor worn-out HT cheapie is struggling, age and solid workouts are catching up with it -- the CD transport sometimes take countless tries to initialise a disk, the electronic volume goes ta-ta's now and again; and just a day ago the main chip amp finally decided it had had enough of the heavy duty treatment. Still working, but it regularly goes into fits of squealing and buzzing. How does this relate to dynamics? Well, it also now sounds like a typically crappy amplifier -- OK at pleasant listening levels, but starts to fall apart when asked for a bit of volume: it's lost its mojo! Some major internal crack or fracturing has kneecapped it, but luckily there are two spare, identical chip amps, unused normally, that can be pressed into service! Just have to rev up a bit of energy, and do some track surgery ... :)

    Posted 5th March 2013 at 01:10 PM by fas42 fas42 is offline
  2. Old Comment
    abraxalito's Avatar
    I see you got a response from Earl to your experience that its 'the electronics, stupid'. He must have missed reading where you said using high efficiency speakers (his are) was one way to reduce the demand on the electronics. I've had interactions with him in the past - these days just give him a wide berth.

    Was looking at the latest behemoth amplifier last night - Boulder 3050 - 1500W/8ohms monoblock. By our standards its seriously short-changed in the capacitor department having (from what I can see) no more than 60 * 4,700uF 100V cans. By which I calculate the peak ripple (into 8R load) as 1.1V assuming perfect ESR.
    Posted 8th March 2013 at 12:55 AM by abraxalito abraxalito is offline
  3. Old Comment
    Why don't you run a spice sim to prove your hypothesis? Take an amplifier circuit, attach a real speaker load that is, say 85 db/2.83V, and another that is 100 db/2.83V. Then run them both at 90 db and 100 db and compare the distortion profile.
    Posted 8th March 2013 at 06:28 PM by ra7 ra7 is offline
  4. Old Comment
    Already been there, done that ... take a real amplifier, real power supplies, real mains quality, real loads - i.e. not "perfect" instances of any of the chain - do reasonable sim's of that as a total "thing", and you marvel at the fact that the sound is subjectively as decent as it is most of the time. It's quite easy to see how the "grunge" gets into the final result once you stop making assumptions ...

    Posted 9th March 2013 at 01:07 AM by fas42 fas42 is offline
  5. Old Comment
    abraxalito's Avatar
    @ra7 - my hypothesis is one of noise modulation, not distortion. I could run a sim in LTSpice using a music wav file as input though, but I'd want to be sure first that PSRR was correctly modelled, not sure it is in many active models. Thanks for the suggestion.

    Found a Stereophile mention of the Boulder - seems it has only 48 caps so that's 24 on each rail 113,000uF - way under-capitalized.
    Posted 9th March 2013 at 03:19 AM by abraxalito abraxalito is offline
  6. Old Comment
    You may recall me mentioning once an Australian designer doing it right, over 20 years ago: [url]http://www.me-au.com/ME_1500_data_2.jpg[/url]

    Posted 9th March 2013 at 05:42 AM by fas42 fas42 is offline
  7. Old Comment
    abraxalito's Avatar
    Took a look (the link doesn't work directly) and loved his purple heatsinks but don't see why the caps are so far from where the hot bits are. Anyway good eye candy
    Posted 9th March 2013 at 09:17 AM by abraxalito abraxalito is offline
  8. Old Comment
    I'm pleased to see that Basspig Mark is giving a few people a solid kick in the appropriates about how effective realistic levels of clean sound are, in the "Better than Live??" thread. He's achieved it using very efficient pro speakers and amplifiers way more powerful than necessary, so the normal audio squibbing doesn't occur. But, the key point is that he is able to experience convincing sound, a proof of concept happening...

    Next trick, do that with equipment that's not OTT, that doesn't have huge margins of capability ...

    Posted 9th April 2013 at 11:41 AM by fas42 fas42 is offline
  9. Old Comment
    abraxalito's Avatar
    I looked at that thread - Mark Basspig's speakers are 112dB efficient - a bit different from anything I've experienced.

    The PNAS video was interesting - even on my crappy laptop speakers I heard that the reproduced voice had digital 'edginess' - confirmed by the one of the guys interviewed who said he could hear 'details' more clearly. I put this, and the loss of ambience reported firmly at the door of the digital bits in the Korg recording box.
    Posted 9th April 2013 at 03:22 PM by abraxalito abraxalito is offline
  10. Old Comment
    triode_al's Avatar
    Could a reason for dynamics of high efficiency speakers be that they are underhung, meaning the voice coil is shorter than the gap depth? And that they are most often powered by flea triode amplifiers with rather low damping?
    And yes I have ESL57 (who said anything about compression and needing something like a Krell or Accuphase) and a Lowther/L3702 in a build stage.
    Posted 18th September 2013 at 07:47 AM by triode_al triode_al is offline
  11. Old Comment
    abraxalito's Avatar
    DaveC's RMAF 2016 Best In Show: Austin Acoustic

    Notice the amps for the top end are fleapower type DHT. Hence the drivers need to be uber-efficient which quite possibly accounts for the stratospheric prices.
    What's needed to achieve this level of dynamics from normal efficiency drivers is exceptional amps....

    Oops, I see Amir didn't much like it, based on music he didn't much like....
    Posted 11th October 2016 at 01:31 AM by abraxalito abraxalito is offline
    Updated 11th October 2016 at 04:03 AM by abraxalito

New To Site? Need Help?
Copyright 1999-2017 diyAudio