Building a crossover and not sure where to start

[bre's dad]

[greywarden]

Re: Building a crossover and not sure where to start

Well the easiest way to do this really, is going to be this:

Read the Speakerbuilding Bible in the Sticky Section. Then Go to Paul Carmody's website and read his simulation walkthrough.

Go to Zaph's website - Zaphaudio.com - and find those 3 drivers (I'm pretty sure he's measured them all - if not, the tweeter is the only one) and download SPL Copy.

Now trace all graphs.

Now run them through Response Modeler

Now Import them into PCD

When you've done all that, create a thread where you post the living crap out of your simulation pictures and a couple weeks later you will have streamlined the crossover enough, without measuring to build it.

Then you'll build the crossover. And voice it, and change it a couple times. Then give it one final build.

And Voila! You're done.

[JimHRB]

Re: Building a crossover and not sure where to start

As greywarden is trying to tell you, designing a 3-way crossover for a newbie is a monumental task. Unless someone is willing to step in and do the simulation for you, you are facing a long (and probably disappointing) road. What are you unhappy about the speaker currently sounds? You may be able to improve things by adding zobels to the woofer and midrange to even out the impedance and let the crossover work better, and then adding L-pads to even out the levels from the drivers. Will that be equal to a properly simulated and designed crossover? No! But who knows, you may be happy with the improvements.

[BeeTL]

Re: Building a crossover and not sure where to start

Well the easiest way to do this really, is going to be this:

Read the Speakerbuilding Bible in the Sticky Section. Then Go to Paul Carmody's website and read his simulation walkthrough.

Go to Zaph's website - Zaphaudio.com - and find those 3 drivers (I'm pretty sure he's measured them all - if not, the tweeter is the only one) and download SPL Copy.

Now trace all graphs.

Now run them through Response Modeler

Now Import them into PCD

When you've done all that, create a thread where you post the living crap out of your simulation pictures and a couple weeks later you will have streamlined the crossover enough, without measuring to build it.

Then you'll build the crossover. And voice it, and change it a couple times. Then give it one final build.

And Voila! You're done.

Oh...is that all?

;)

Seriously though, that's a nice little summary.

[Chris Roemer]

Re: Building a crossover and not sure where to start

Right now you must have some screamin' machine there. I'm working on a 17 element XO right now. Had to cut both the tweeter and the mid by over 12dB.

I'll be back.

Chris

[greywarden]

Re: Building a crossover and not sure where to start

Oh...is that all?

;)

Seriously though, that's a nice little summary.

I really, really can't tell you how many times I've typed that. I've probably dumped $700 into the rediculous project I've working on for the past 15-16 months. I should have started with either A) someone else's designs or B) purchasing measurement equipment and using measurements.

[donradick]

Re: Building a crossover and not sure where to start

I really, really can't tell you how many times I've typed that. I've probably dumped $700 into the rediculous project I've working on for the past 15-16 months. I should have started with either A) someone else's designs or B) purchasing measurement equipment and using measurements.

There's an old saying: "Too late ve get schmart".
Seriously, you know what you're talking about, and you now have a couple of boxes of crossover parts - those will come in handy. Don't be like me, though, and have a shelf full of $5 buyout drivers that you will probably never use. :eek:

Strongly advise a measuring setup. Sims only get you so far. Sometimes they work out great, but other times there are "troubling" peaks and valleys in the measured response that weren't there in the sim.

[bre's dad]

Re: Building a crossover and not sure where to start

As greywarden is trying to tell you, designing a 3-way crossover for a newbie is a monumental task. Unless someone is willing to step in and do the simulation for you, you are facing a long (and probably disappointing) road. What are you unhappy about the speaker currently sounds? You may be able to improve things by adding zobels to the woofer and midrange to even out the impedance and let the crossover work better, and then adding L-pads to even out the levels from the drivers. Will that be equal to a properly simulated and designed crossover? No! But who knows, you may be happy with the improvements.

There are a few things. Mainly I am wanting to do it as a learning experience. Second the tweeter is crossed over too high. 3000 is recommend. Currently its XO'd at 5000. PE does have a crossover that would work but I really don't want to spend $100 (about $50 each) if I don't have to. Plus that one crosses the woofer too low @ 375 Hz. I haven't found a premade 3 way XO that has the cross over poiints I'm looking for. I also already have L-pads on the tweeter & midrange.

[Chris Roemer]

Re: Building a crossover and not sure where to start

Right now you must have some screamin' machine there. I'm working on a 17 element XO right now. Had to cut both the tweeter and the mid by over 12dB.

I'll be back.

Chris

I don't suppose you'd care to push your mid up to touch the tweeter? There's no good reason to put it so far away (and a lot of bad ones).

[Paul Carmody]

Re: Building a crossover and not sure where to start

PE does have a crossover that would work but I really don't want to spend $100 (about $50 each) if I don't have to. Plus that one crosses the woofer too low @ 375 Hz. I haven't found a premade 3 way XO that has the cross over poiints I'm looking for. I also already have L-pads on the tweeter & midrange.

I understand what you're thinking, but passive crossovers don't work like that.

[morris]

Re: Building a crossover and not sure where to start

I don't suppose you'd care to push your mid up to touch the tweeter? There's no good reason to put it so far away (and a lot of bad ones).

This should be easy enough for the OP to do. Just slap a fresh baffle on top of the old one and cut out new holes. It would be an opportunity to dress is up more if needed.

Chris, If you do not mind, how are you simming the triangular baffle of the OP's original box? I really have to hunker down and learn "The Edge." Or buy Omnimic and learn to measure.

[Chris Roemer]

Re: Building a crossover and not sure where to start

This should be easy enough for the OP to do. Just slap a fresh baffle on top of the old one and cut out new holes. It would be an opportunity to dress is up more if needed.

Chris, If you do not mind, how are you simming the triangular baffle of the OP's original box? I really have to hunker down and learn "The Edge." Or buy Omnimic and learn to measure.

I won't be. I'm only taking him 80% of the way from "textbook" to "as nearly perfect a sim as you can get". I AM shaping the rolloffs properly using real F/Z files, and the phase probably won't be perfect, but like I said, with a stock XO, the tweeter and mid were probably at least +12dB, which isn't really very good. If he had an "F" (or D-) XO, he can have maybe a "B" (or B+). Looks like about 4-5 dB of generic BSC.

Chris

[morris]

Re: Building a crossover and not sure where to start

I won't be. I'm only taking him 80% of the way from "textbook" to "as nearly perfect a sim as you can get". I AM shaping the rolloffs properly using real F/Z files, and the phase probably won't be perfect, but like I said, with a stock XO, the tweeter and mid were probably at least +12dB, which isn't really very good. If he had an "F" (or D-) XO, he can have maybe a "B" (or B+). Looks like about 4-5 dB of generic BSC.

Chris

A "B" or B+ XO is nice. especially since those drivers are not really cheap.
You are very generous Mr. Roemer Sir!