More crossover questions

[ocdSCHACK]

Hello to all... this is my first post so I'll try to not ask too many questions.

I want to build a 3-way speaker that sounds great, but can also play really loud. I am going to use the the Eminence ASD 1001 for the high, Eminence Alpha 6a for the mid and I think I've decided on the Dayton 12" RSS315HF8 for the woofer. I can't find any pro woofers that that can go that low without being 15"+ and needing the worlds largest box; plus the Dayton RSS315's have received great reviews. I would like to build a simple 6db 3-way crossover for all drivers, but use an 18db slope in the tweeter.
Questions:

1) I know I will need L pads on the Mid and Tweeter... will the L pad Resistors change the values needed for the crossover points I choose? I've read they don't.

2) Should I use a Zobel Network for the woofer (and mid-range?) I hear they help but am worried after I add Zobels and L-pads, nothing will work as I plan it too.

3) Because I'm using L-pads on the Mid and High, do I have to worry about whether I wire and the speakers in parallel. Almost every 6db and 18db crossover diagram shows all speakers in parallel.

4)Has anyone had any real experience using a 6db crossover with any of these drivers? I would plan on using crossover points somewhere around 450 and 4K Hz.

Thanks for you help

[Pete Schumacher]

Re: More crossover questions

Hello to all... this is my first post so I'll try to not ask too many questions.

I want to build a 3-way speaker that sounds great, but can also play really loud. I am going to use the the Eminence ASD 1001 for the high, Eminence Alpha 6a for the mid and I think I've decided on the Dayton 12" RSS315HF8 for the woofer. I can't find any pro woofers that that can go that low without being 15"+ and needing the worlds largest box; plus the Dayton RSS315's have received great reviews. I would like to build a simple 6db 3-way crossover for all drivers, but use an 18db slope in the tweeter.
Questions:

1) I know I will need L pads on the Mid and Tweeter... will the L pad Resistors change the values needed for the crossover points I choose? I've read they don't.

2) Should I use a Zobel Network for the woofer (and mid-range?) I hear they help but am worried after I add Zobels and L-pads, nothing will work as I plan it too.

3) Because I'm using L-pads on the Mid and High, do I have to worry about whether I wire and the speakers in parallel. Almost every 6db and 18db crossover diagram shows all speakers in parallel.

4)Has anyone had any real experience using a 6db crossover with any of these drivers? I would plan on using crossover points somewhere around 450 and 4K Hz.

Thanks for you help

The RSS315HF won't be as sensitive as your other choices, so you'll be padding down the mid and tweeter a lot in order to keep a balanced sound. Also, with that mid, you don't want a crossover point above 2KHz due to off axis issues.

[ocdSCHACK]

Re: More crossover questions

I know the woofer is way less sensitive, but couldn't find a pro woofer that size to play low enough... Is that difference in sensitivity a bad idea? Would the 4 ohm woofer be any better of a match?
Thanks again.

[fbov]

Re: More crossover questions

You sound like a reasonably intelligent guy who has done a lot of research and thinks he's ready to dive in. You're not ready.

Everything you've said shouts "I've read the book." No one's told you the book is only a starting point. You will get something that works with your approach, but let me suggest an alternative, and one I'm still very new to - design your own system.

You'll need some measurement equipment, a mic and mixer to do frequency response measurements on your PC, and something PE sells called a WT3 (woofer tester 3) that measured impedence and other Theil-Small parameters. And you'll need an enclosure, preferably the one you plan to use.

Measure the FR, phase and impedence for all 3 drivers in the enclosure. Input the data into XO design SW like PCD (Passive Crossover Designer) and then select the kinds of XO circuit elements you've read about. You'll see predictions for total FR and impedence, which will allow you to model padding resistors and the like as well as how to arrange values to get a bump here, a dip there, a shelf from here-to-here, etc.

The point is that actual loudspeakers are nothing like what books must assume in order to teach you something. I have done as I suggested above and found I stilll needed to pad the tweeter some, but my measured FR was darn close to the model's prediction. If you've chosen wisely, you'll also like the way they sound.

But I will close with the admonishon that you should start with a 2-way as I'm told 3-way's are far harder to get right. If you love those drivers, by all means continue, but remember that the best place to start is with a proven design....

Have fun,
Frank

[ocdSCHACK]

Re: More crossover questions

Thanks for your reply Frank.
I have done a lot of research, but hit some road blocks and thought posting some questions might improve my learning curve. I have built a set of speakers before, but nothing complicated. I knew very little, and when I came to a problem, I read a book, the internet, etc... and found the formula I needed to move on. My learning technique was similar to cramming for an exam. I built a pair of 3-ways with each speaker using a nice value Hivi tweeter, 2 Hivi 2" mids that were on sale for 5 bucks, and a 10" Dayton classic woofer. I used adjustable L-pads to simplify the build and help protect against my ignorance. I put one speaker together, and chose to lower the frequency of the highs... Ordered bigger capacitors and built the other speaker while my second order was on the way. Using a simple 6db crossover actually worked well and they sounded great to my surprise. A year or 2 later, I now wish I used a bigger mid-range and padded down the tweeter a little more despite already having an L-pad on it. The speakers play loud and clean, but I want to build something that will blow them away. I know my speakers won't win an Audiophile contest, but they'll easily beat anything you can get at Bestbuy. Sorry to bore you with my life story, but I thought it might give you a better idea of my experience. I know I'm not really ready, but thought a forum of experts might be able to simplify the process. The more I learn about this stuff, the more I realize I don't know. I was hoping there was an easier way than buying a bunch of test equipment, or a cheaper way than buying every type of capacitor, resister, and inductor to play the mix and match game until the drivers sound good.

[johnnyrichards]

Re: More crossover questions

1st order @450 Hz = bad idea for the Dayton subwoofer. Probably no less than 4th order at ~250 Hz. The Dayton has a relatively low sensitivity compared to the mid and tweet, compounded by baffle step loss.

I suggest a high sensitivity MTM with a pair of 15" subwoofers with a ton of power to get the high, full range SPL you are looking for.

[Wolf]

Re: More crossover questions

I know I'm not really ready, but thought a forum of experts might be able to simplify the process. The more I learn about this stuff, the more I realize I don't know. I was hoping there was an easier way than buying a bunch of test equipment, or a cheaper way than buying every type of capacitor, resister, and inductor to play the mix and match game until the drivers sound good.

You don't HAVE to buy a lot of test equipment if you aren't in this for the long-haul.

You can trace the curves, alter them to your specs, and apply them to a xover optimizer. I will be writing a blog-spot on this soon, as it seems there are soooo many ways to skin this cat. Once you can setup your spreadsheet or program correctly, your procedure should be repeatable between projects (if the graphs are reputable), and give you a reasonable outcome to tweak with.

To this point, I have not 'measured' a lot of FR plots. I have traced quite a few. I have however, 'measured' a lot of T/S parameters and Impedance plots with my WT2. I highly recommend a WT3 to double check the driver consistency, and design for your pair. A good LC meter and DMM is also a good thing to have handy.

I can comment more later, should you be interested,
Wolf

[maynardg]

Re: More crossover questions

You don't absolutely have to have measurement equipment.

You do need measurements from reliable sources made under known conditions (often manufacturers are not reliable). You can trace images of such measurements with SPL trace (free), and process the output with other free software such as Response Modeler 2.0 and PCD 6.2. Follow some of the guides on these that can be found via search here. Ask questions from the context of that approach. WT3 is well worth the money.

Listen, listen, listen - trust your ears.

Very respectable results are quite attainable. I've won the voting at some of the speaker contests and have never used my own measurements.

[ocdSCHACK]

Re: More crossover questions

I want to thank y'all for the comments so far. I respect all of your opinions, so I won't rush this project until I start getting replies without the "bad idea" undertone. I just bought parts to build the "overnight sensations" to fix my DIY itch for the time being. One problem I have is that all the programs everyone recommends are Windows based and I have a Mac. I'm good with Excel... can I use this instead. It sounds like SPL trace could be done with 3 columns: SPL, Frequency and Impedance. Can I just just read the FR graph and input the numbers into 3 columns on a spreadsheet to make my own charts? I've become totally overwhelmed with where to start... I'm guessing I need a thorough list/chart that has the entire FR like "Speaker X at 1000Hz is 95dB with 5.5 ohms?" I can easily make a spreadsheet for this, but I'm sure it not that simple...is it?

[ocdSCHACK]

Re: More crossover questions

I just finished the "overnight sensations" this weekend. I'm very impressed with the sound. They seem to play every sound without any particular frequency range draining out the others. Although I don't really know, they seem like they could make decent studio monitors with little change (if they make passive monitors.) I liked having someone else's design to work from; it involves much less effort. Most importantly, I know from listening to these that I'm asking questions and getting advice in the right place. I also want to thank Paul Carmody, Bill Schneider and WWWJD for the great writeups and ideas. Bill's version of the removable back is so much more attractive than methods I have used in the past; I might steal his idea for every new build I do.
But back to my speakers...
I purchased the ASD 1001 tweeter and the Alpha 6A midrange (both Eminence.) I placed them on an open baffle to experiment and I think I hear potential. Something about the Alpha 6 in an open baffle sounds like the artist is really there. It seems to "flood" the room with sound. Any ideas on where to start if I want to do an open baffle with this. I'm guessing an active sub will be my next project. I just downloaded the demo for Fuzzmeasure because I have a Mac. I'm tempted to buy a bunch on Inductors, Capacitors, and Resistors and play with them using the Frequency sweep. Any better ideas?

[billfitzmaurice]

Re: More crossover questions

I know the woofer is way less sensitive, but couldn't find a pro woofer that size to play low enough... Is that difference in sensitivity a bad idea? Would the 4 ohm woofer be any better of a match?
Thanks again.

With the options being going low and having high sensitivity you can't do both. Pro woofers don't go as low as their hi-fi counterparts because they have high sensitivity, and vis-versa. If you want to go low with high sensitivity the options are either horn loading or multiple woofers; both work, neither is small.

[wg_ski]

Re: More crossover questions

I know the woofer is way less sensitive, but couldn't find a pro woofer that size to play low enough... Is that difference in sensitivity a bad idea? Would the 4 ohm woofer be any better of a match?
Thanks again.

A better match would be the Series 2 12", or the Quatro. Higher sensitivity, reasonable size vented box, low to mid 30's. And about in the same 'class' as the mid/tweet.

[Ludo]

Re: More crossover questions

I've done a couple of designs using the Alpha-6A which might give you some additional ideas.

The first is a 2-way indetended to be used with a subwoofer and the second is a 3-way with an integrated subwoofer. In the 2-way the Alpha-6A is tuned to run down to 80Hz while in the 3 way I am crossing it at ~250Hz. I used the Morel CAT-378 tweeter crossed at 2500Hz. I wound up using fairly complex x-overs to extract a sound which was to my liking out of the Alpha-6A across a wide variety of music material.

http://techtalk.parts-express.com/sh...highlight=ludo

http://techtalk.parts-express.com/album.php?albumid=198

Louis

[ocdSCHACK]

Re: More crossover questions

I'm going to get some testing equipment before I beat my head into a wall one more time. I just got my Parallels program working, so now I (hopefully) can run all the programs everyone recommends on my Mac.
so...
What are the necessities of testing equipment? Obviously a microphone, but do you have any suggestions? I went to radio shack and the widest range one could only go down to 30Hz; I assumed this wouldn't work and don't really think that highly Radioshack stuff anyway. What should I be looking to spend? I'm guessing $200-$300, but know this is a subject where you could spend as much money as a new car would cost. I would prefer cheap. I also want to do it right the first time, so cheap isn't a requirement. At the very least, I would like to run a frequency sweep and/or pink noise and reliably see the problem areas of a speaker or crossover setup.

[Wolf]

Re: More crossover questions

Any RTA program can get you going. SynRTA (free), Room EQWizard (free), or the True Audio stuff from John Murphy for sale here at PE can get you there.

The microphone should be a measurement mic, usually of the electret variety, and will typically cost you about $60 without calibration.

A soundcard if necessary can be had fairly cheap, and depending on which program suite you utilize, should have certain features. I don't think you require it for the above 3.

There is a mic to USB interface that can connect ot the PC for the RTA stuff. I don't know if a mic-preamp is required, but I would wager that it is. Some microphones have a requirement for Phantom power, and some do not.

You also need a pink-noise source and an amp and speakers for the RTA style measurements.

Using the MLS sweeps and a full-measurement suite will require the soundcard, and a full-duplex unit at that, the mic-pre, amp, your Fuzzmeasure software, etc.

Here is a good source for what gear you'll require:

http://soundhobby.com.istemp.com/measurementsystem.htm

The tracing is a lot more economincal for the end user, as you can probably see now.
Later,
Wolf