|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: Minimizing woofer dynamic compression (RBP??) posted by Duke on April 16, 2002 at 03:10:02:
Hi Duke!I will cut and paste...if that is ok...
Q:. Other than efficiency, what factors make a significant difference as far as minimizing woofer dynamic compression? This will help me narrow down my woofer choices (specific recommendations enthusiastically appreciated).
* Look for a unit that has the least gap distance in the pole piece/magnet stucture and preferably a flat wire voice coil. Losses in the gap are a major concern for dynamic integrety. JBL professional woofers (2226H for a 15") would be a great choice...but follow the parameters for proper loading of that motor.
Q. 2. Do muliple woofers give you better dynamics than a single woofer of the same model? In this application size and maximum volume aren't really an issue, but best possible dynamics is.
* You will lower distortion with multiple woofers and increase dynamics. Distortion is lowered by having less cone motion of each woofer for a given output. Once most woofers move farther than an 1/8 inch on a half wave (1/4" peak to peak)...distortion from differential field bias is measurable and audible, not to mention "cone cry" and other moving part anomilies. I love 3 equal units in unequal enclosure sizes, close but not the same to spread out unwanted resonances over a broader band. Stay within the parameters of the driver. You can vary 20%, making the ideal one the one closest to ear leave at your seat, or focus position.
Q. 3. If multiple woofers are used, what's the best way to hook them up to minimize compression? For example, should two 8-ohm woofers be hooked up in series or in parallel? Should four 8-ohm woofers be hooked up in series, parallel, or series-parallel?
* Look at the system impedance you need for your amplifier. In horn loaded enclosures, I prefer to use 2 16ohm units in parallel to get 8 ohms. I did a Klipschorn bottom at 2 ohms once (dual voice coil single driver 4 ohms each coil in parallel and had great results because the amp I was using was comfortable with low DC resistance and the bass horn raises acoustic resistance appreciably. The above means that you can vary the final impedance to suit your amplifier power requirments...dampning, heat, drive current average...etc..
I go parallel before I go series in 95% of cases.Choose your minimum DC resistance and wire according to number of drivers, parameters and amp choice. There is not "ideal impedance" other than a flat impedance curve...which is damn near unattainable, unless you go to resistive loading, and that will fuck up your dynamic integrety. (tradeoff again)
All designs are tradeoffs. You have to live with the ones that are acceptable for your application and desires.
Hope this helps.
Email is audiowkstation@aol.com if you want to get deeper.
Follow Ups:
Hi! I happened upon your Q and A and found your response very informative and also intriguing. KEF used to use a technique called "conjugate loading" e.g. in the KEF 104/5 to provide the amplifier with a friendly load, but I don't know what that amounted to, and whether it is similar to the "resistive loading" you mentioned, and involves a trade-off. Are there any do's and don't's for gettig a favorable impedance curve?Thank you,
Peter
Basically, a zoble circuit will control the impedance rise at resonant frq. that occurs in the circuit (woofer, cabinet, room air, crossover, and amplifier..plus speaker leads.All these are contributors. I know of no company that has a totally flat impedance curve unless they have found a way to eleiminate wire in general..be it speaker wire, crossover wire, or voice coil wire. Even esls have wire..so absolutly, phase shift free and flat impedance curve, power curve through the entire range is impossible. even if it could be done in therory, just moving your body to another location in the room would change the air loading behavor , thus causing some interaction with the motor system that would change the character just enough to rule out "perfection". The zobel circuit, carefully used with controlled internal enclosure dampning, and predictable frontal cone loading with the crossover as close to the amplifier as feasable does yield promising results. I would say the obvious "least trade offs" would be to map the drivers inductance curve (if using moving coil units) and use said driver only within the range that the inductance is most stable with the zoble circuit applied. Formulae for zobels usually does not account for changing inductance, changing power band and changing frequency. Sometimes just the "right" enclosure loading without a zobel will yield a flatter impeadance curve than with a "formulated" zoble. What it all comes down to is integrating the speaker system into a vast number of room varables and getting as predictable musicality as possible. This is where the magic lies. VNF listening (very near field) does 3 things that are of advantage. 1. you are masking room acoustics. 2 You are listining to the speaker at or about where it's frequency response was documented. 3. you can have low power and modulation..which equals lower distortion. Obviously, this is uncomfortable at best to be right in the pattern of most speakers..but if loudspeaker manufactures would design their products with the 3 meter rule in effect...we all would enjoy higher fidelity..or even 4 meter to 5 meter. When I design and build a loudspeaker system, the crossover is usually designed in my lap or seating position and tweaked until the focal lenght is appropriate for that distance.
Their is much work left on high fidelity loudspeaker design for the future. We are just scratching the surface after 77 years of usage.
Volumes could be written on the inabilities of loudspeakers to behave as needed in all circumstances.
Once again you have taken the time to teach me, and I thank you. Your influence on my understanding of and direction in high realism loudspeakers has been and continues to be dramatic.And truer words were never processed than these: "All designs are tradeoffs."
I appreciate your offering me your e-mail address, and I'll be writing you a follow-up.
Best regards,
Duke
This post is made possible by the generous support of people like you and our sponsors: