Thanks Julian,

What your looking at are the original response curves and finalized response curves of the speaker. The crossovers are adjusted and re-tested with modifications. I had a friend of mine run the speaker through their paces in test cabinets (he's an engineer for a pretty well known loudspeaker company). When the speakers are finished, these tests will be done again, and then more adjustments will be made. Actual listening tests will start after that and then a final listening round will be done with the owner of the speakers. I will make any further adjusemtns after that.

I'm sure comments will follow from the peanut gallery.

 

Nice work Fozzz, you've obviously done your home work. I can't wait for build pics. This is going to be a very interesting thread. Not that it hasn't been amusing thus far. lol

 

when do I get to hear the damn things already???????...I'm sure defros right and they'll suck.

I've seen some of your work on home speakers Fozzy one and you know I'm kidding about them sucking.

nice job and thanks Ju for getting him back into this, mines next year...lol

 

I've got about 10 sets of half completed home speakers in the shop. You welcome to a set. I'm running out of room!

 

So, what do you use, (apart from ears & experience) to determine the value of your x-over components?
It's one of the most arcane and dark areas of the audio arts, so I'm always intrigued by what folks are using.
One of the most disappointing speakers I've heard, is the MBL mini-monitor. I found out that they actually use a passive high-pass filter before the woofer. So a heavy-gauge inductor is the first series component between the amp and drivers. Could explain that sub-83db sensitivity figure. Generally, the MBL stuff I've heard has been very good, but way too rich for my bank account. These ones seemed to have been designed on an off day........

 

Usually the first thing I have done is a frequency response plot of each speaker in the system. Some speakers have a natural roll off that can help or hurt you depending on what you do with it. Other speakers have serious spikes that will either need to be controled directly or crossed over early enough that they arent a problem. You also need to take response plots with the speakers in their cabinet. The width of the cabinet can cause problems, as can tunning frequencies and even resonance issues are thingsthat need to be delt with before you put your final design togther.

It takes a long time.

I'm only working on these speakers during weekends for short periods. The final design is actually very nice (IMHO) and will keep the future owner quite happy. I will post pics in February of how the cabinet was built. They are big and heavy, and my back is less than happy with me right now!!

 

I know this was directed to the o/p but I wanted to offer an option I like to use. I found a software for PCs that allows you to use your multichannel sound card as an active crossover. This lets you quickly make crossover changes on the fly. Once you have it sounding good you can make a fixed frequency active or passive setup. I use KX software but there are several other free ones. Console is not free but it's easier to use and LSPCad has it built in.

high pass would have a capacitor in series no?

 

Have you ever tried LSPcad? I have used it with success though it could be more user friendly. Also have you ever considered elliptical filters? They are a nice tool to tame resonance and achieve a steep cutoff at the same time. Good luck with your design.

 

A steep cutoff is not always better. This particular design is using 6db per octave slopes.

I have tried active filters as a bench mark to help pinpoint frequencies. But in the end, the passive systems react differently and sound differently than active. So most of the work was waisted.

 

Right. Having a software compatibility issue between by brain & hands.
High-value, heavy gauge inductor in series to take off the bottom end of the system.
Lots of wire in series = high insertion loss. Generally thought to be an inadvisable approach in crossover design. Certainly, there will be exceptions.

Fozz, main brunt of the question referred to crossover topology. Once you've made your driver measurements & determined enclosure format, what do you use to determine component values relative to the passive crossover components?

I've used LEAP/LMS, CASD/CACD, LFDES/FLTDES and the ever-unfriendly CALSOD.
Each one has benefits & drawbacks, too numerous to list in a reasonable amount of space.

Just curious what others are using in their design kitchen. Lotsa ingredients and recipes available these days. It's a great time to be involved in this area of audio!