, 05-22-2008 at 02:41 AM (1579 Views)
I don't always take harmonic distortion arguments to heart, as they aren't always within the best range for the driver's implementation. If you use a driver well, in the range best applied, most will sound fantastic with optimal xovers, and the distortion might not even be audible. I build projects to experiment, and experience different drivers we have available to us. This is a hobby, and not life-or-death struggles. I don't always understand every nitty-gritty detail discussed, but at this point, I don't find that much is lacking from my experiece to make decent projects. Some are definitely better than others, but you'll have that.
I don't always use BSC, as project-applied BSC is more specific. For small-drivers, say 3", 6dB BSC is not a good idea. You are killing the best sensitivity of the driver for a "seemingly" more balanced bottom-end. In turn, the top-end will seem right, when the driver starts to overload, or breach Xmax. Sometimes vented enclosures can be utilized to lessen Xmax, whether using multi-chambers or not. Most of the time, we use these little boogers with a sub anyway, so the receiver will usually have a high-pass applied, and unloading below tuning becomes a non-concern. In the case of my "Dragon-Foal" project, I achieved a +4.3dB at 140Hz or so using the 0.9 Qtc M3N in a vented enclosure. This alignment lacked the baffle-step lost in a lot of listening perceptions, and was rather balanced. The mn-series have the rep for good bass in vented enclosures, despite the high Qts of the driver. It's all how you use them, not the size.
For xovers, I typically use pegboard and zip-ties, and maybe some adhesive. Point-to-point wiring is preferred, but jumpers have been employed in the past for them. You can paint the pegboard a color if desired, so you can get a different look, and zip-ties come in a plethora of colors. The more compact the layout, while adhering to proper inductor placement, the better, IMO. If I can cram it on a 2.5" square, so be it! I am not opposed to any typical topology, so series xovers and parallel xovers are always a possible outcome in planning stages.
I sometimes go a little out on a limb by using more of the premium-grade xover components, mainly capacitors, but am perfectly happy with good quality mylar caps as well. Non-polar-electrolytics (NPE's) have been used in the past as well, but in larger-values being a requirement.
IMO-CrossCaps are smoother than Solen caps for series applications, but Jantzen (CC, Z-Silver, Z-Standard, Z-Superior), Solen, Bennic, Dayton, ASC, Carli, Aerovox, Wima, ERO, Mundorf, Hovland, Panasonic, General Electric, (NPTM, MKP, and Film/Foil types) have all worked well for me, without failure. I found Sonicaps to be "hard" sounding in the standard-fare. All of them sound good, but the Z-Silver has my vote for the best I've used. EDIT: I prefer Solens in woofer shunts over a lot of other brands- it seems to have a better damping effect on the bass range. Paralleling caps might have merit, I've tried it, and it sounds fine; EDIT: BUT in some cases it can make the treble smeary, and sound like it's being emanated by 2 sources instead of the proper single. This may be application/component specific. Take this info however you want, as it's my opinion, and I stand by it.
As for coils, I typically use 20/18 AWG air-core, but steel/metal core types are fine in woofer circuits. I have used foils on tweeters when using uber-caps on the xover. I have also used a LOW-DCR C-Coil in a series xover application to reduce parasitics. This was favorable. I use P-Cores a lot. EDIT: In one recent case, I had a lower DCR 16AWG coil and added resistor sound better than a smaller gauge coil with the same net DCR as the resistor + DCR in the former configuration. This was an eye opener, as I had never considered that DCR vs. Resistor would be that different.
For resistors, I have normally employed sand-cast or 1% panel-mount-aluminum (Arcol/Dale) types with great results when optimal. Mills are sturdier than the sandcast for wire leads, but I hear no difference between the 2.
If you can get a similar result in a smaller box with the same drivers, I say go for it! Smaller is easier to build, and more aesthetically pleasing for the WAF to be high. If they aren't pretty, then this is the modus operandi you should take until they are better aesthetically. 5 Hz is only 5 Hz, when you are talking a difference of half a cube in volume. I use Unibox to emulate the mox model, and own a WT2 (V 4.04) to measure real specs of the drivers in question, including tweeters and mids. Real measurements of specs and xover component values are a task overlooked by some. I measure all of my parts to maintain similarity for left and right channels, though small tolerance is usually negligible, and notch filters are the exception. They are required to be more accurate to achieve the results desired.
I now own a microphone and preamp, and have software on hand to use. I am needing to acquire a soundcard to complete my measurement rig. Most likely, I will be using ARTA's freeware as my choice. I'm not a stickler when it comes to being a flat horizontal line from 20 Hz to 20kHz. If I can get within +/-2.5dB in a simulation, I'm pretty happy. I might not like full 6dB BSC, and 3-3.5dB might be adequate. Measurements are just measurements, and models are also their namesake. Your ears become the final judge, so tweaking is always a must. I trust my ears, and know when something sounds *off*. Jim Holtz will tell you I can hear differences in wire, as I did a few years back. I believe playing an instrument of the non-amplified variety has given me a pretuned ear, and therefore rely heavily on listening. As for what I've done in the past, I've used measured, measured/traced, and traced FR curves in simulations, and not all are accurate or optimal for the endgame that your drivers will have to accompany. In simulation, I'm not an over-user of xover components, but instead use what is necessary to get close to the goal in the simulation. If that takes 14 parts a side, it takes 14 parts a side!! Minimalism is not always the best method, as you can sacrifice a lot in 2 parts omitted. I use Jeff B's RM/PCD spreadsheets program for my modelling simulations. "Minimalism with respect to necessity", as I've said quite often.
Finishes are finishes, and I'm "out-there", but it's all fun and learning what looks different.
It's all application, optimization, realization, customization, prolification, and relaxation. In that order. We don't want our projects to become sediment in the electronic rock of the future. This is all my perspective, and the pitons are in the grooves..... I'm still climbing the mountain until I can smooth sail making this easy... It's a project and goal scenario, ease of progression is not on a logarithmic curve. I know this was long, but I have pretty much everything in print on my convictions for the task at hand. I love this hobby!!!
Your #1 Source for Audio, Video & Speaker Building Components
Deal of the Day
View Our latest
Order our FREE 336 Page Full Color Catalog
Speaker Component Categories
Home Audio Speakers
Professional Audio & Guitar Speakers
Car Audio Speakers
Measurement & Design Tools
Subwoofer Plate Amplifiers
Full-Range Plate Amplifiers
Cabinet Hardware & Speaker
Subwoofer System Kits
Speaker Repair Parts