instantaneous flux modulation compression.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Was reading the ESP sound pages website, and came across this page:

Power Handling Vs. Efficiency

He discusses and tests power compression and 'instantaneous flux modulation compression. '

Interestingly, he also states that typical home use speakers really need to be over 90db/1w to be good.



I have searched for more tests/information but have found very little apart from a factory tour of ATC (yes the guys who do THAT mid dome)

This intrigued me as my favorite sounding mid is indeed that dome. The engineer talks about the same things as ESP when explaining what makes it sound so good.

Can anyone suggest more reading on the subject?
 
I have searched for more tests/information but have found very little
3rd party non-linear distortion tests of speaker drivers:
Zaph|Audio
Measurements and compare | HiFiCompass
and my own here:
Timothy Feleppa's Pages

There is not a direct relationship between sensitivity and distortion although there may be a loose relationship. There are many reasons for this including:
-Drivers with larger surface area need less excursion to reach a specific output level. Lower excursion means less distortion, and increasing surface area increases sensitivity. Larger diaphragms come with disadvantages however, such as lower frequency breakup nodes and poorer off axis response.
-Underhung motor designs (voicecoil shorter than magnetic gap) typically achieve lower distortion and higher sensitivity but typically have lower peak excursion capability for their overall motor size compared to overhung motors (voicecoil longer than magnetic gap). Tweeters tend to be underhung. Woofers tend to be overhung.
-Larger more powerful motors achieve higher sensitivity. Larger motors tend to remain more linear through a given amount of excursion than smaller ones. This very much applies to the ATC 3" dome - 'solving the problem with brute force'

That said, there will always be drivers that buck the trend - low sensitivity drivers with excellent distortion, and high sensitivity drivers with horrible distortion.

It's also worth pointing out that distortion varies with amplitude. Driver A may be lower distortion than driver B under 100dB, but driver B may be lower distortion than driver A above 100dB. Underhung motors tend to have low distortion up to the point where xmax is exceeded and distortion rises sharply. Overhung motors tend to have slightly higher distortion initially but it increases more gradually as xmax is exceeded. "Tend to" is important here - a good overhung driver might be better in every way compared to a badly designed underhung driver.

Distortion also varies vs frequency due to a multitude of reasons.

The best way to approach the problem is to take real world measurements of drivers by doing harmonic distortion sweeps and consider how different drivers might be able to be used together to achieve full-range (20Hz to 20kHz) sound with minimal distortion by avoiding using drivers through frequency ranges which they have distortion issues.
 
Last edited:
Well this is very interesting, is this an important reason why ScanSpeak and others use modular magnets? To improve cooling?

Pics below from a ScanSpeak Ellipticor and Focal Utopia subwoofer.
 

Attachments

  • AirCirc.jpg
    AirCirc.jpg
    160.8 KB · Views: 198
  • g09133WX2-b.jpeg
    g09133WX2-b.jpeg
    19.7 KB · Views: 194
could it partly be cost? Are the magnets energized after being placed in the return circuit? One large energized neo magnet could be dangerous to handle. I have W8-1772 made that way - it seems to waste flux as there's opportunity for screwdriver to be yanked to the puck magnet's outer edge.
 
Well this is very interesting, is this an important reason why ScanSpeak and others use modular magnets? To improve cooling?

Pics below from a ScanSpeak Ellipticor and Focal Utopia subwoofer.

Looks darn good too!

My reason for asking all this is I'm looking to find reasons why the speakers I like the sound of are so good to my ears.

That ATC mid dome appears to reveal detail I simply cant get close too any other way. It could of course be psycho acoustic, but it sounds like its revealing more detail to me, and not just a bit more either.

I hear very little if any difference between amplifiers (within operating range) and none between good CD players.

Big horns have something special though.

My favourite speaker is the Living Voice Olympian. This is why my current project is a JBL2452SL JMLC 350hz horn with a AE TD10M mid driver below, and possibly a JBL 2206H (12 inch) or 2226H (15 inch) bass driver mounted on the side of each cabinet for 60-200hz use.

The AE TD drivers measure extremely well (Augerpros tests indicated they were the best)

The Ideal outcome would be the etched detailed mid of the ATC SM75 (s), but with the aesthetics of my horn design and the dynamics of the Living Voice system. God knows what I will actually get though!

Having now 3D printed the connecting tube between the 2452 driver exit (1.5 inch) and the JMLC horn throat (2 inch) , I ran a full bandwidth signal (at a low level) to it. It currently has the non Aquaplas coated version diaphragm.

It sounded awful.... Hopefully when crossed over it wont be so bad!
 
Last edited:
Well this discussion sure has me thinking about the Velodyne idea of using accelerometers on the cones, or some other means to detect actual cone movement vs. intended.

It also has me thinking about line arrays and power dissipation in a new light. I never bought the idea that a small woofer was "faster." Distortion figures at low Hz and high output prove it, but then ... I wonder how much easier it is for them to dissipate heat?



Not to mention, I'll stay away from passive crossovers in high power systems from now on. :)



Best,


E
 
Last edited:
Yes, real magnets are always magnetized after assembly; it´s absolutely impossible to do otherwise.
Those are ceramics (2 rings stacked), it´s even more impossible with Neos which to boot would repel each other .
Actually, assembly of magnetic structures using fully magnetized neos is done a lot in the accellerator, light source, and particle physics world. That said, it is very dangerous if done stupidly. Fixtures used are non magnetic, if one gets a hand or finger between two very much in love magnets, the outcome will always be "sub optimal":( .. Fingers usually go first, and victims generally have to wait for others to bring non magnetic tools to pry the parts.
Only one incident here so far, and that was just a wrench caught in the field, well, with two fingers in the way..

Jn
 
Well this is very interesting, is this an important reason why ScanSpeak and others use modular magnets? To improve cooling?

Pics below from a ScanSpeak Ellipticor and Focal Utopia subwoofer.
Not sure about cooling (might even be worse due to lower heat conduction from the coil to the magnets?) but it certainly would reduce chuffing/fluttering noises that occur from air getting trapped behind the spider and dustcap.
 
Well this discussion sure has me thinking about the Velodyne idea of using accelerometers on the cones, or some other means to detect actual cone movement vs. intended.

It also has me thinking about line arrays and power dissipation in a new light. I never bought the idea that a small woofer was "faster." Distortion figures at low Hz and high output prove it, but then ... I wonder how much easier it is for them to dissipate heat?

Not to mention, I'll stay away from passive crossovers in high power systems from now on. :)

Best,

E

My thinking has followed similar lines in the last few days..


YouTube

This is Bob Carvers discussion on his line array speakers..... very interesting!

-Can't believe the ribbons play down to 400hz though:eek:
-Using the speakers as a Mic for room correction is genious
-using the drivers to create a set image height is also v smart.
-The bass drivers being THAT small is really crazy though. They prob add up to a single 8 inch drivers swept area?

I think my next speaker may be a line array...(if I ever finish this one!)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.