kfr01 wrote:
Steven, I'm pretty dead-set on building my own. I understand that kits are an easy way to start, probably the better way in terms of an end-product, but I frankly don't think that the kits will help me learn anything. |
|
|
That's no problem. I just wanted to be sure that everyone understood the complexity of what they are getting into. Of course all the great speaker engineers had to start somewhere.
kfr01 wrote:
What do you (everyone out there) recommend as the best way to proceed? What do I need to buy? |
|
|
Well first I recomend doing some serious research. There are many different drivers out there in many different price ranges. And price doesn't necessarily dictate the quality of the driver.
kfr01 wrote:
The Soundeasy + Boxcad package seems to be fairly full-featured for the price. Does anyone have any experience with this software? Is it adequate, or do I need to shell out the full price for Leap? |
|
|
I'd recomend LSPCad. It seems to be very accurate and has some great features. It has a good crossover modeling section in it, and is one of the more accurate enclosure design softwares on the market.
kfr01 wrote:
Anyway, as long as the drivers are of similar sensitivity and have sufficient flat frequency response overlap are there any other major considerations regarding driver selection? |
|
|
Actually similar sensitivity isn't critical as you can always adjust the differences in the crossover network. Just remember the lower in frequency you run the speakers, the less overall output you can get from them. Also look closely at the frequency response of the mid. A lot of times you will see cone resonances that cause dips in the response early which means you will have to roll off the speaker early. This can also mean that you will have to run the tweeter lower and thereby hurt your peak output. You should also check out the response of the speakers at different axis points. Otherwise you might have good results on axis, but have a very narrow sweet spot.
You should also understand how we hear things. Sometimes something will measure badly, but will sound just fine. A perfect example of this is doppler distortion. It measures very bad in some speakers, but when you actually listen it is either inaudible or isn't objectionable.
kfr01 wrote:
After the enclosure the next step seems to be crossover development. As I understand it, ideally this involves modeling the actual driver response characteristics in software, selecting a crossover point, watching what happens to the combined response, and then correcting / tweaking for abnormalities caused by the crossover and box. Is this a correct summary of the steps necessary? |
|
|
What my boss usually does is first select the speakers. He then has me build test enclosures for the speakers. He then tests each speaker in it's place, with the microphone hooked up to his computer. Next he uses LSPCad to design a crossover for the speakers using that collected data. Then the computer is connected to an amplifier and the amplifier feeds the speakers. LSPCad allows him to play music that is then fed through his similated crossover network and into the speakers. He can actually hear the crossover before he builds it. From his listening tests he can determine whether to modify the design or not. After that he hands me the crossover scematic, I build it up and we test it.
kfr01 wrote:
Finally, I'm thinking of creating active filters instead of passive. It seems like a much more workable system. I've heard nothing but positives about 'going active' from the internet and the short blurb in the cookbook. Are there any major drawbacks to 'active' that these have skimmed over? |
|
|
Well I think having to amplify each speaker seperately is a major drawback. Honestly unless you are going with digital crossovers, I don't see any benifit.
kfr01 wrote:
Anyway. Basically I want to learn what else there is beyond selecting the drivers and "crap-shooting the crossover design." Is my brief summary accurate? |
|
|
Sorry, I hope that didn't sound rude. It's just that I've seen people do this before and they never get good results. I just wanted to make sure you knew not to do it.
I do want to add one more thing. The crossover is the most important part of a speaker system. It is what gives the speakers their voice. Second in importance is the speaker selection. And finally a distant thrird is the enclosures.
I hope this helps!
Steven Kephart
Adire Audio