It's in the crossover!

[Steve Gonzales]

Hello members,

I just had an experience that I thought to share. Most all of the seasoned veterans will probably not gleen anything from this, but for the up and coming speaker guys, it will reinforce something that you might not realize.

First off, I am just beginning to explore the ins and outs of crossover design. It is complicated in a lot of ways because there is SO much to consider. Here is something that I just found out that has firmly pointed me in the right direction-for good.

I have a dedicated two channel system for my music and a pretty good Home Theater for movies. I bartered and traded, bought and sold, to get what are the top of the line suite. I don't like plugging another company too much, but I'll say that they are from England. I've noticed that the $550.00 center channels (one front and one rear) have a less than appealing mid/HF output. You all know that the Center channel handles most of the dialogue (speech) and it's frequency response and timbre, is critical to a realistic experience.

Not being satisfied, I decided to investigate the crossover and speakers. I took the connector cup/crossover out ,and to my disbelief, in my $1100.00 worth of center channels ,were but a single capacitor (first order or 6db) and the LF was running full range!!!. Now, don't get me wrong, if the design team did that and achieved a good result, who am I to mess with success?. Here comes the big BUT!.---But they didn't and it was clear to me why. Being one to tinker, I looked up the crossover points and then went to my spare parts box and found a pair of JBL L19 crossovers. The specs were within 200hz, the impedence is the same, blah, blah,blah. I took my time and figured out how to make them fit by ungluing some of the JBL components from the board and arranging them a bit closer. After a short while, the moment of truth. I fired them up and WOW, what a complete difference. SWEET voicing!. I even have a HF level control now.

That brings me to my point. If I didn't already know what a huge difference a crossover can/could make, I certainly do now. I just wanted anyone new to the game, to know how important and critical, the crossover is. I've been at this for a long time and never gave this as much attention as I now firmly believe it deserves. Maybe this will help someone out there. Steve G

[jblnut]

I agree 110% with this Steve. I could never have predicted the difference that new xovers would make in my L250's. Getting it "right" is a mix of science and art (some might say magic) but I'm here to tell you - once it's right you'll never want to go back.

In fact it's gotten me wondering about active crossovers...

jblnut

[rek50]

I've been trying to injest a "Couple" of network ideas. You know the TWO: Cap selection, Coil selection, Resistor selection, Conductor selection, Roll-Off, Congjugate, Frequency cut-off, Low Pass, Band Pass, Pass Pass,Wire the cap with the foil side facing the output, Hold your tongue right while soldering, Theil/Small, :biting: :biting: :biting: . The distilate, so far, looks like, a Biased/Charge-Coupled, Dayton Capped, Theta bypassed Hi-fallutin network is THE answer, The Holy Grail, the end. Thanks Giskard. Now all I got to do is START.

[Valentin]

i wonder if LSR 6332 will ever have charge coupled crossovers and how big a differce would this be any body tried it .

[Steve Gonzales]

I've been trying to injest a "Couple" of network ideas. You know the TWO: Cap selection, Coil selection, Resistor selection, Conductor selection, Roll-Off, Congjugate, Frequency cut-off, Low Pass, Band Pass, Pass Pass,Wire the cap with the foil side facing the output, Hold your tongue right while soldering, Theil/Small, :biting: :biting: :biting: . The distilate, so far, looks like, a Biased/Charge-Coupled, Dayton Capped, Theta bypassed Hi-fallutin network is THE answer, The Holy Grail, the end. Thanks Giskard. Now all I got to do is START. Exactly!, there is SO much to the crossover to consider. We are fortunate to have a place to go to and get this info. And I thought enclosure tuning was difficult!.... It gives one hope though, that the very same driver compliment can sound completely different with a better crossover.

[DavidF]

You may already be familiar with Lynn Olson's 2-way design called the Ariel. If not, below is an address to one excerpt relating to the crossover design. May be somewhat lengthy but in reading all the way through you may get an idea of what goes into the crossover design for a small 2-way with some similar considerations for minimalist style of crossover.

DavidF

http://www.nutshellhifi.com/Arieltxt2.html#top

[Steve Gonzales]

What I have found is that although I can grasp, vaguely, the concept of what is going on, most of these good, in depth, articles are filled with terms that are unknown to me. It's like trying to understand a car's fuel injection system without really understanding the parts. You know "what" it's doing but can't connect the dots because you are not familiar with the parts. I find myself looking at the dictionary quite often. Like the word "conjugate", that means where two things meet, etc, etc. Maybe someone has a link to a good glossary to lift the vail of mystery?. I think that once this is accomplished, there will be an explosion of hungry, budding, crossover "mechanics". Anybody out there have a link to a glossary of terms?

[Ken Pachkowsky]

I could not agree more. As you know my Westlake HR1 having been built in 1989 is long in the tooth. I have tried many different crossover combo's in the quest to find something that "would do" while getting the Westlake rebuilt.

Nothing has even come close to the sound generated by the HRX. I have found quieter circuits but they lack the soundstage and imaging of the HRX. To be fair to Klark Teknik and Ashly, I am sure if either company designed a specific network for the HR1 it would at the least match or surpass the sound quality of the HRX with this particular system.

My lack of budget (13,000.00 for a new HRX) has led me to building the custom design done by Ian McKenzie in cooperation with Nelson Pass of Pass Labs. Ian's description of his 2-way version convinced me to go with a switchable 4/5 way design. This will let me experiment with addding the 2403/076's to the HR1 to see if extending the UHF to above 16k will add even more life to the system. I in no way wish to permanently change the physical design of the HR1 so will add the 076's in walnut boxes that match the large horn. These will hang over the edge of the top cabinet and be adjustable left or right and be set back passed the baffle as in the SM1 design.

I look forward to your visit Steve. We'll have a nice barbeque and enjoy some good Jazz.

If anyone wants access to the thread please message Ian McKenzie for access.

Ken

[Robh3606]

Hey Steve

Yes! Well designed passive crossovers are very difficult to do and seperates them men from the boys. I am still cutting my teeth with a binky on this issue. I have done my own crossovers and they ended up measuring and sounding very good but it was a difficult and very time consuming process to get it right. When you do get it right you know it. It just all comes together the driver balance is right, you don't get fatigued, the imaging in good and the clarity and tonal balance are right. It's just right and not on one or two disks but it sounds right on everything unless the recording truely does suck. As you go through your voicing you go one step back as many times as you go forward. It's worth it though.

What I find a big help is access to the JBL Tech Sheets and the information available on the Pro Site. We have most of the schematics for the crossovers available and they can be very instructive. Asside from the crossovers themselves you can see the nominal crossover points as well as the slopes used. This gives you a good idea where it is "safe" to use the drivers. You can also clone a system having the original schematic where you would be guessing without it.

"In fact it's gotten me wondering about active crossovers..."

Hello jblnut

They have there own little pit falls. Some of the drivers well all love have some surprises that the JBL networks manage to tame or work around. A couple of examples are the rising 2122H response and the 4430/35 compensation. Running a 2122H active you need a notch filter from about 800Hz till 1.1K to tame a peak in that range. The 4344 crossover does a great job of flattening that peak out. The 4430/35 compesation is not a canned curve and the best way is still to use the passive compensation built into the 4430/35 network with the biamp switch set in biamp mode.

Then you have to balance and voice it which on the surface sounds like a snap but really is not. You should have an RTA as a minumum to set-up an active system. With it you can verify driver phase and basic levels. Once you have all the phase issues tamed you have to tweek the driver levels to voice it. I always set them up flat and then tune from there. Doesn't take much of a change in level only a dB or so the change the balance. Also have to match the levels Right and Left or you won't image correctly and will loose depth. I like the actives and they are a good altenative but don't think for a minute there is not a good deal of work left to get it all sounding right.

Rob:)

[Steve Gonzales]

Hey, I want to put out the call for a glossary link. I'm VERY intrigued about this stuff and as I've stated, it is a subject that needs to be given a lot more respect. I've seen some crossover changes recently that on the surface seemed not such a big deal but now I've given it a lot more thought. Like here recently, a change from a N150 to a N150a crossover changed the whole character of the two seemingly identical speakers.

[Zilch]

You should have an RTA as a minumum to set-up an active system.Same with tweakin' passives, I've found. I was just peein' in the dark without an RTA.

It's my third ear, the objective, realistic one.

[Now, I can see the rope.... :p ]

[DavidF]

I have not come across any comprehensive websites that bring the total into one glossary. Audioexpress sells a couple of reference books that provide a substantial amount of info:

http://www.audioxpress.com/bksprods/bkindexldesign.htm (http://www.audioxpress.com/bksprods/bkindexldesign.htm)

Also, to save time and some frustration, you are going to want/need at least a multi meter, tone generator, & sound meter. More can be spent on software design and test programs. Even at a base level, you will need a variety of parts (caps, coils, and resistors) available for empirical testing.

Crossover design is one component in the entire system design but it must also be the most time-consuming. Assuming a well planned, tested and voiced crossover design is already in place, changes to that design that result in an "improvement" are a less likely result than changes that result in a "difference". Those "improvements" are likely to be very costly on an incremental basis (as I contemplate a charge couple modification). Still, any difference you create may be better to you because you have different voicing preferences and this does create an enjoyable aspect to tweaking the original. Due to the nature of compromise though, you can also introduce one or two negatives for each positive that you create. Also, you can end up chasing-the-tail in a never ending cycle of changes, tweaks, upgrades. It can be fun, it will be frustrating, but remember to always keep the original circuit handy.

David F

[Steve Gonzales]

Here is some very interesting information: http://www.silcom.com/~aludwig/Sysdes/Crossove_Design.htm

The more I research crossovers, the more I am in AWE of those that can "speak" the language. Funny how most of us take this for granted!!!. There are so many things to consider. When I see a T/S parameter data sheet now, I have a completely different view. I truely understand now that "it's in the crossover!". The best drivers on the planet (JBL) are lost without a good one. I'm gonna dive in and try to ingest a small fraction of what is possible. Right now, it's like trying to plug all the leaks while inside a ship's hull and someone is on the outside shooting holes in it with a MACHINE GUN!. ;)

[Harryup]

Not tried a JBL DCS280 or a similiar product?
That will make a huge difference since you get the room correction as well.
When you have tasted that cup of tea there are just no way to return back.
And you will be able to mix amplifieres to your choice of bass, mids and highs.

[4313B]

http://www.jblpro.com/pub/electronics/dsc-280.pdf

[johnaec]

If anyone's interested, I have my DSC260A, (little brother to the DSC280), up on eBay right now: http://my.ebay.com/ws/eBayISAPI.dll?MyeBay&LogUID=johnaec&CurrentPage=MyeBaySelling&ssPageName=STRK:ME:LNLK

John

[dynamo]

My buddy jblnut has just spent a ton of money on new crossovers for his very high end JBL speakers, and I am here to tell you that the difference in the sound quality of his entire system is nothing short of incredible and amazing in every way. Especially considering that his speakers are allready formidable 4 way top of the lines with upgraded drivers. I don't know if it's the batterys keeping the caps charged up, newer polypropelene caps, the hot rod design generally speaking, or all of the above? I can recommend to anyone based on audition, that the crossover is a shamefully underestimated component in the audio chain and worth looking into for the serious among us.

[jblnut]

My buddy jblnut has just spent a ton of money on new crossovers for his very high end JBL speakers, and I am here to tell you that the difference in the sound quality of his entire system is nothing short of incredible and amazing in every way.....

Hey - welcome aboard ! Looks like you got my email with the links to this site.

Just so everyone is has the background, my old friend dynamo has been a vinyl junkie from the very beginning. When everyone was tossing their LP's, he was building an incredible collection of music. He brought a favorite of both of ours over this weekend - the MFSL LP release of Queen's "A Night at the Opera". We proceeded to give the 250's and the new amps a real workout. By the time we got to Sweet Lady the needles were peaking over 200w :rockon2: . It was awesome and I'm glad to have an objective person hear the system at different stages. We all make changes to our systems and it's sometimes hard to really make an accurate judgement on whether it's "better" or not.




jblnut