Newbie Crossover selection question

[fire01]

After reading many of your projets here on this forum,i m amazed to see how many different projets look very apeeling and are more than hi end audio.I would like to start one also but me skills are very limited in crossovers and all the specs surronding speakers.As for the cabinets and finishes i probably could manage something pretty good.

My question is"Are pre made crossovers a good start for any type of speakers or are they low end and the sound is crap?"Since i never built a crossover before and they seem very technical i find the pre made ones an easy way to start.PE has a nice selection of already made ones.

Could i use one of those for lets say any projet?Or do i have to select certain types of tweeters,mids,woofers and inclosures?

Any help would be very helpful :T


Thanks
Danny

[ThomasW]

Generic premade crossovers are fundamentally useless..

The best bet for a n00b is to build one of the tested designs in our Missions Accomplished section.

http://www.htguide.com/forum/forumdisplay.php4?f=39

[Hdale85]

We need to put this in a sticky or something...seems to get asked once or twice a month.

[bluewizard]

"Are pre made crossovers a good start for any type of speakers or are they low end and the sound is crap?"

Well, that question is indeed at the very heart of a great debate. Others will say they would never use off-the-shelf or text-book-made crossovers under any circumstances. Though sadly those individual rarely if ever offer resources to create alternate 'good' crossovers.

If you are striving for audio perfection and a perfectly flat frequency response where every little deviation from flat has been trapped out, then you need some pretty complex Crossovers.

But, it you simply want to duplicate the equivalent of some nice reasonably priced commercial speakers, then certainly off-the-shelf crossovers will work. Though it doesn't take that much more effort to design your own 'text-book' crossovers. There are sources of on-line crossover calculators as well as free software that will calculate basic straight forward crossovers for you.

Keep in mind we have a HUGE disparity on what people consider 'entry level' speakers. To some of us, that means $100 each, to a few it means $100 a pair, to others it means $1,000 each, and to still others it means speakers substantially higher than that. Keep in mind that you can easily spend $10,000 on a pair of speakers, and $50,000 is not unheard of.

If your goal is some decent REASONABLY priced 'consumer' speakers, then certainly off-the-shelf crossovers will get the job done, assuming you have the knowledge to pick the right crossover. Which pre-assumes you have the knowledge to pick the right speakers.

One step better than that, is to design and build basic crossovers yourself. The Key here is to look at the frequency response graphs and use the actual speaker impedance at the crossover point rather than a generic 4 ohms or 8 ohms.

In addition to that, it would be wise to add a Zobel network to the woofer. This is a simple resistor and capacitor in parallel with the woofer to stabilize the woofer impedance at higher frequencies. Again, on-line calculators are available.

Next, midrange speakers and more so tweeters tend to both sound louder and to actually be louder than woofers. So, you need to add some L-Pads, which are simply two resistors one in series with the tweeter and another in parallel with the tweeter. The combination of the two will lower the signal level by a pre-selected number of DB, to match the levels of the other speakers.

Purely as an illustration, if you have a woofer with an SPL Sensitivity rating of 94 db and you have a tweeter with an SPL rating of 100 db, you would want to trim 6 db off the tweeter to keep it at the same level as the woofer.

Enter your speaker impedance and the required drop in DB, in this case -6 db, and an on-line calculator will give you the two resistor values you need.

In practice, it is a little more complicated than this, but 10watt power resistors are about $2.00 each, so you can get a selection across a range of attenuation, and use the combination that sounds best to you.

The basic 12db/2nd order 3-way crossover is four coils and four capacitors, plus a few power resistors. It's not that complicated.

Again, I'm certain people will tell me I am dead wrong, and I would actually agree with them. Every argument you could make for detailed speaker specific custom designed and built crossovers is correct. When we are talking about some one with a lot of experience, a lot of resources, a lot of time, an urge to study and learn, a huge pile of money burning a hole in their pocket, and a plan to build as close to the ultimate speaker as possible in their several thousand dollar budget.

But that's not everyone. Some of us just want a good budget $200 to $300 pair of speakers that exceed what the equivalent would buy us on the open consumer market. If you fall into that category. Then either off-the-shelf or generic 'text-book' crossover will do. Many consumer speakers have little more than that in them and even many reputable budget on-line speaker plans have little more that straight forward crossovers in them.

But let me remind you that you must first pick the correct combination of speakers; that's not as easy as it sounds.

Next, you must select the right crossovers for that combination of speakers.

That's why it is better to build from a known and proven speaker plan your first couple of times. Even I would recommend this. Give us an idea of what you had in mind, that is how you pictured these speakers, and give us a general budget, and we can probably point you to several proven speaker designs.

So, the answer is Yes and No, depending on how much money, time, talent, resources, tools, knowledge, and experience you have.

Few will agree with me, but I stand by what I said.

steve/bluewizard

[Landroval]

"Are pre made crossovers a good start for any type of speakers or are they low end and the sound is crap?"

If you are striving for audio perfection and a perfectly flat frequency response where every little deviation from flat has been trapped out, then you need some pretty complex Crossovers.

But, it you simply want to duplicate the equivalent of some nice reasonably priced commercial speakers, then certainly off-the-shelf crossovers will work. Though it doesn't take that much more effort to design your own 'text-book' crossovers. There are sources of on-line crossover calculators as well as free software that will calculate basic straight forward crossovers for you.

Partly true, but mostly misleading. Pretty much all commercial speakers use some kind of custom crossover, even if it's just one cap and coil. By using an off-the-shelf xo you might find one driver combination out of 20 that'll work ok, while all the rest are a disaster. And finding that one combination is harder than building a custom xo anyway.

If looking for an easy way out, fullrange drivers can provide very good sound reproduction with no crossovers or maybe a baffle step correction cirquit of one coil and one resistor. Depending on the price range many 3" Tangbands, CSS FR125S or Jordan JX92 are well worth taking a look. And no xo's needed...

[Dave Bullet]

Hi Danny,

Build one of the designs in the section as posted by Thomas. No really - you want to and will be happy you did.

If I'd never designed an aeroplane before - yet insisted I design one and build it and fly it - you'd declare me crazy. If I was building a proven aeroplane kit, you might think that was achievable / sensible and would probably be a good spend of my time and money.

So - build the kit. You will still learn heaps and have something worth your time and effort. With all the stresses life has - who needs more!

Cheers,
DAvid.

[cjd]

If I'd never designed an aeroplane before - yet insisted I design one and build it and fly it - you'd declare me crazy. If I was building a proven aeroplane kit, you might think that was achievable / sensible and would probably be a good spend of my time and money.

Bringing this analogy a little closer, it'd be like finding a great deal on a 747 wing, an old VW aircooled engine, and a stash of sheet metal screws and thinking they might bolt together to make something that flies. You might happen across components that DO fit together just right, but... not in most cases. People go for big bad woofers and cheap little tweeters and wonder why a 3/4" dome and a 12" woofer don't quite sound right crossed at 2500Hz 2nd order using a circuit designed for 8ohm flat when the drivers are 6ohm rising on the woofer and 4ohm on the tweet (also not flat of course... probably has Fs right at the XO point...)

You have to know everything you need to know to design a crossover to get a system to work with an off-the-shelf crossover - you just design "backwards" - fixed XO point, expected nominal impedance, so your job is to find drivers that fit (and probably add zobel).

C

[Dennis H]

Danny,

If the concern is with reading the schematic and soldering the crossover together, some of the places that sell crossover components will assemble the crossover for a nominal charge. There are good designs out there in about any price range you can name so there's no reason to stumble along trying to design it yourself when you admit you're new to the whole driver/crossover thing. Tell us your budget and we can make some recommendations. Besides this site, Zaph has a lot of good designs.

http://www.zaphaudio.com/

[bluewizard]

Well, though it would seem illogical given my position, I agree with what every one said.

Let's cut the the chase of my long-winded statement -

But let me remind you that you must first pick the correct combination of speakers; that's not as easy as it sounds.

Next, you must select the right crossovers for that combination of speakers.

That's why it is better to build from a known and proven speaker plan your first couple of times. Even I would recommend this. Give us an idea of what you had in mind, that is how you pictured these speakers, and give us a general budget, and we can probably point you to several proven speaker designs.

I know off-the-shelf crossovers can work because I've made them work. But I have a woofer with a rated frequency response of 30hz to 3000hz, a midrange with a rated response of 500hz to 12,000hz, and a tweeter with a response of 2,500hz to 25,000hz. That gave me a lot of room to work with. (Mid and Tweeter are horns.)

I settled on an 800hz/5000hz crossover from MCM Electronics. Well within the working range of all the components I chose.

Now I certainly admit, this constitutes some compromises, but the results were good. I'm fiddling to see if I can improve the port tuning, but, at the moment, have no desire or need to modify or replace the crossovers.

So, the first step is to truly select speakers that are able to work together. That REALLY is NOT as easy as it sounds.

Rated frequency response on woofers is something like 25hz to 2500hz, but if you look at the frequency response graphs, few of them are usable above 1,000hz without some special effort on the designers part.

In many cases, the problems with a woofer are outside the normal rated usable range. Take the aluminum cone Dayton Reference speakers. They are nice in their working range, but have horrible break-ups and peaks above their rated range that are going to come into play and are going to be heard. It takes some additional crossover work to make sure that break-up region doesn't diminish the quality of sound of your final speaker.

So, if you are using off-the-shelf crossovers, that makes it difficult to select 'hard cone' woofers (no matter how COOL they look) and get good results. Best to stick with good paper cone woofers.

Probably the most difficult task is to find a woofer, midrange, tweeter combination that will work together. A Mid with workable crossover frequencies that works with your woofer is not likely to work with your selected tweeter. Once you get a Mid/Tweeter combination that works, it is not likely to work with your woofer. Very difficult to select truly compatible speakers relative to workable crossover frequencies.

Now if you are very knowledgeable in crossover designs, you can make some seemingly incompatible speakers work well together. It's part of the magic of actually knowing what you are doing. Yet, no one starts out at this level. We have to build a few speakers to get there. We even have to build a few wrong speakers to truly get there.

I agree, even if you are going with a basic generic 12db crossover, it is better to make it yourself. You have a lot more flexibility, AND as circumstances and improved skill level demand, you can tweak them until you have reached a higher more accurate level of crossover.

Start where you are at and keep going until you are where you need to be.

Other than this one very easy combination of speakers, I have had a very difficult time finding Woofer, Mid, Tweeter combinations on my own that can easily work together.

So, for beginners, it just makes sense to build a known proven design that appeals to you. Gradually as you become more skilled and more knowledgeable, you can take on greater amounts of the design process yourself. Eventually you'll be designing some pretty sweet speakers ... eventually.

So, Landroval is right in saying -

By using an off-the-shelf xo you might find one driver combination out of 20 that'll work ok, while all the rest are a disaster.

Finding two or three speakers that will work together is VERY difficult, then finding a stock crossover with the appropriate frequencies for your selected speakers is even more difficult.

I can be done with good results for good budget speakers, but as you can see even the easy route still has its difficulties.

Find an existing design you like and build it. That is always the advice that is given to those starting out in speaker building.

ZillaSpeakers -

http://www.zillaspeak.com/zillaspeakspeakers.asp

Troels Gravesend -

http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm

Zaph Audio -

http://www.zaphaudio.com/

Humble Homemade Hi-Fi -

http://www.humblehomemadehifi.com/

Parts Express DIY Projects-

http://www.partsexpress.com/projectshowcase/projectindex.cfm

RJB Audio - Speaker Projects -

http://www.rjbaudio.com/projects.html

Just passing it along, in my usual long-winded but well intended fashion.

Steve/bluewizard

[fire01]

Thanks alot bluewizard for your help.All start from a existing design and work from there.

This is what i m really considering for the moment:

Mini Statement,Statement Monitor with close to wall crossover,SR71 and some different MTM designs.All of these look very nice and the reviews seem good but wich one?I belive the SR71 are the easier build cause i could order the Madisound kit and buy the baffles seperatly to start.Looks like a great design but the Mini and monitor are three way designs and seem more appeeling for music and HT.Now for MTMs i belive are lower end designs with less sound quality,correct me if i m wrong.

Now if i go with the mini i dont need stands vs the monitor i ll have to buy stands or build them myself,wich brings the cost to even more than the minis for parts.Its a kind of situation that only listening test could help me decide i guess.

Thanks
Danny

[Landroval]

Find an existing design you like and build it. That is always the advice that is given to those starting out in speaker building.

Couldn't agree more! And great links to sites with options for any need.

[chrisn]

Thanks alot bluewizard for your help.All start from a existing design and work from there.

This is what i m really considering for the moment:

Mini Statement,Statement Monitor with close to wall crossover,SR71 and some different MTM designs.All of these look very nice and the reviews seem good but wich one?I belive the SR71 are the easier build cause i could order the Madisound kit and buy the baffles seperatly to start.Looks like a great design but the Mini and monitor are three way designs and seem more appeeling for music and HT.Now for MTMs i belive are lower end designs with less sound quality,correct me if i m wrong.

Now if i go with the mini i dont need stands vs the monitor i ll have to buy stands or build them myself,wich brings the cost to even more than the minis for parts.Its a kind of situation that only listening test could help me decide i guess.


Thanks
Danny

All of the specific designs you've listed are good choices. I don't think there is anything inherently lower end about MTM designs, and there are several good MTM designs in the Mission Accomplished section. The Nat P's come to mind...

If you like a monitor or mtm design you could always build it as a floorstander. Just keep the baffle width the same, as well as the driver distance relative to the top of the cabinet. You can block off any extra volume with a shelf brace. You could also build Bass bins as your speaker stands, something I've been meaning to do, yet haven't.

Many of the build threads here have very detailed schematics and photos of the crossover assembled. It may help you become more familiar with XO's before you start. We all had to start somewhere :B

I actually did have a simple TM design that used one cap, one coil, and one resistor and was fairly close to LR2 slopes. It sounded okay at lower levels, but the tweeter couldn't keep up at higher levels.


Hope this helps 8O