Q&A Thread for Dickason's "Loudspeaker Desgin Cookbook"

[MarcE]

I just finished reading Vance Dickason's "Loudspeaker Design Cookbook" (6th edition) for the third or so time, this time with pretty detailed notes and lots of questions! I was wondering if people wouldn't mind taking a stab at them

I'll post one chapter per reply and perhaps spread them out so people don't get too bored!

I really enjoyed the first chapter. Some of the things that the author compares and contrasts that I'm interested in are:

• pole geometries (straight, undercut, angled)
• voice coil (underhung or overhung)
• back EMF mitigation (high permeability metal in iron next to voice coil, shorted turn / Faraday loop, conductive coating over pole tips (or other similarities such as copper cap, copper cylinder, or shorting ring for flux stabilization)
• former composition
• two-layer versus four-layer voice coils (very interesting actually)
• cone flexing and cone resonance modes
• dust cap (solid versus porous and other related decision such as vented pole piece or vent holes in voice-coil former)

Problem is I have already decided on the woofer and midrange for my new 3-way project. So, I was wondering if anyone might know what features these drivers have in the areas outlined above, so I can reread this chapter in light of what features these drivers have? The woofer is a Scan Speak 25W/8565-01 and the mid range is a Audiotechnology C-Quenze 15H 52 06 13 SDK.

What I've managed to find so far is:

Features of Scan-Speak 25W/8565-01:

• low loss rubber surround
• SD-1 motor
• under hung voice-coil
• hard paper cone

Features of the Audiotechnology 15H

• Kapton former
• SD system
• EMF mitigation - copper capsule wrapped around the centre pole piece
• high ventilation factor through the centre pole piece, perforation to the voice coil
• under hung motor structure
• solid die cast basket
• Polypropylene cone

So, I'm just curious if people know a bit more about these two drivers than the published information in terms of the first list of items?! I'm guessing that some of the things might be lumped under the SD or SD-1 designations? In some cases, I'm curious how these two drivers address the problems in the list (i.e. cone flexing or cone resonance modes)

Okay, now a few more general questions:

[Q] The LDC doesn't really go much into cone composition comparisons (i.e. the paper-pulp cone of the 25W/8565-01 and whatever the 15H uses ... but they do talk about dust cap materials at length), can anyone outline the differences in cone material has on the driver performance and ultimately the sound (or point me to a good reference on the subject)?

[Q] I noticed in the Madisound catalog that Dupont is no longer selling Kapton to the loudspeaker industry. Anyone know why this is? It seems that Scan-Speak is using fiberglass instead. How does this compare to Kapton? What about the more exotic materials such as glass and titanium?

[Q] The author discusses inverted hard plastic dust caps. Do just plain inverted dust caps have the same advantages and disadvantages? (Also, in the case of inverted, this means inverted with respect to the direction of the code, so "sticking out"?)

[Q] The author discusses "progressive suspension systems" as something found in pro applications, but not (yet) in the DIY / audiophile market. Has this changed and if so, is there a good example of a driver that uses this?

Thanks in advance! Stay-tuned for more questions :T

[Dennis H]

I think you can drive yourself nuts trying to predict the performance of a driver by looking at all the factors you mention. You can find excellent (and terrible) drivers with about any combination of materials, geometries, etc. you can think of. There's really no substitute for measurements -- frequency response, impedance, distortion, etc. If the driver works for the intended purpose, it works and nothing more needs to be said about it. See all the various driver testing pages on the web.

[MarcE]

Dennis:

Thanks for your reply! I have indeed purchased the drivers and plan on using them 8) I just figured that it would be worthwhile to reread this chapter knowing what "my" drivers use. Also, in my opinion, a great project thread starts with a very good description of the drivers, and I would like to include such an intro in my project write-up!

[Jed]

Marc,

Do a search for the following members to get info about driver construction and implications for various designs:

Capslock
Feyz

They are a few of the experts when it comes to that stuff. Maybe they'll chime in here to answer your questions, but that will be a good starting point to get some of the info you are looking for.

Jed

[MarcE]

Jed:

yes, these threads are excellent. Perhaps capslock or feyz will pop in sometime on this thread, or someone who's familiar with these drivers.

[MarcE]

LDC Chapters 1-5

[Q] There is a definite stress in LDC that the success of vented enclosure alignment depends on Q_L (box loss) as this is used as a starting point for using the design tables. An iterative tuning method that depends on measuring Q_L is proposed and using the actual measured Q_L to then adjust the other classical alignment parameters. LDC recommends that if you are not confident in hitting Q_L for the particular alignment you are trying to implement, that you should oversize the box by 25%. Since we have tools such as Unibox at our disposal, is this method relevant? For example, I'm interested in operating my Scan-Speak 25W/8565-01 in a extended bass shelf alignment. The primary approaches for designing a vented enclosure in LDS use design tables that take Q_L, Q_ts, H, alpha, Fs/F3, and peak dB. So, unless there are similar design tables for EBS alignments (are there?), it seems the approach in LDC is not possible with this non-classical alignment. Is this correct? Should I proceed by constructing a test box with the parameters that Unibox gives me and then just tune the port using either SPL or impedance method (not sure which is better), and I'm done with no need to mess with measuring Q_L?

[Q] When dealing with a mid bass enclosure in a three-way system, how does one come up with the box volume? I will be using a sealed enclose for this driver, so how should I choose Qtc, just use the optimized value that Unibox starts off with, or choose Qtc = 0.5? The driver in question is an Audiotechnology 15H.

[Jed]

[Q] There is a definite stress in LDC that the success of vented enclosure alignment depends on Q_L (box loss) as this is used as a starting point for using the design tables. An iterative tuning method that depends on measuring Q_L is proposed and using the actual measured Q_L to then adjust the other classical alignment parameters. LDC recommends that if you are not confident in hitting Q_L for the particular alignment you are trying to implement, that you should oversize the box by 25%. Since we have tools such as Unibox at our disposal, is this method relevant? For example, I'm interested in operating my Scan-Speak 25W/8565-01 in a extended bass shelf alignment. The primary approaches for designing a vented enclosure in LDS use design tables that take Q_L, Q_ts, H, alpha, Fs/F3, and peak dB. So, unless there are similar design tables for EBS alignments (are there?), it seems the approach in LDC is not possible with this non-classical alignment. Is this correct? Should I proceed by constructing a test box with the parameters that Unibox gives me and then just tune the port using either SPL or impedance method (not sure which is better), and I'm done with no need to mess with measuring Q_L?

[Q] When dealing with a mid bass enclosure in a three-way system, how does one come up with the box volume? I will be using a sealed enclose for this driver, so how should I choose Qtc, just use the optimized value that Unibox starts off with, or choose Qtc = 0.5? The driver in question is an Audiotechnology 15H.

Marc,

I have to give you credit, you are definitely being thorough.

As for your first question, all this theory is nice, that is, if you have the actual TS parameters and measurements of the drivers at hand.

For your second question, many people use an oversized box and do not rely so much on the Q of the box to extend the low end response. Instead, an acoustic solution is used, oversized box, in conjunction with an electrical bandpass to reach the target slope. In your case, since I know you intend on using 1st order slopes, you are going to need to extend the useful range past FC, so I'd tune to a Q of .6 for the midrange.

[TacoD]

[Q] The LDC doesn't really go much into cone composition comparisons (i.e. the paper-pulp cone of the 25W/8565-01 and whatever the 15H uses ... but they do talk about dust cap materials at length), can anyone outline the differences in cone material has on the driver performance and ultimately the sound (or point me to a good reference on the subject)?

Different cone materials sound different, soft cones polypropylene (PP) tend to sound less tight in the bass department. Paper and PP have a softer presentation of the transients than metal cones (for me this sounds more natural/ neutral). Metal cones are very precise sounding with clear transients, but sometimes this can spoil the timbre of playback of acoustical instruments.

Just my 2 cents..

An other point of view: http://www.vmpsaudio.com/d-cones.htm

[MarcE]

In your case, since I know you intend on using 1st order slopes, you are going to need to extend the useful range past FC, so I'd tune to a Q of .6 for the midrange.

Right, so that's what I was planning on doing, however I ended up with a different result. As you lower Qtc, Fb decreases, F3 increases and Vb increases. So, I was looking at a Qtc in the 0.4 range, actually. This results in a Fb in the 65-70 Hz range. I'm not sure how extreme this is, but it looks good in my simulations with the 25W/8565-01 and first-order slopes My enclosure will be able to handle the large volume as well, and in fact it results in a simpler design. I was also thinking about adding a variovent to the midrange enclosure, and then noticed that David Gatti is using one with his new 3-way design that uses the Audiotechnology 15H, so now I'm even more intrigued.

Anyways, thanks for the help. I'd rather learn as much as possible before presenting my project, hence the pedantic LDC thread. It won't be long!