• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Crossover design question, Experts needed!

Yester_Year

Yester_Year

New Member
So I have been looking at a lot of factory schematics from Cerwin Vega and I have noticed that in the early to mid 80s they started using inductors and capacitors to bridge the positive lead to the ground. Sometimes when they did it with the cap they also added a resistor. An example is in their DX9 speakers they used a 2.3mH coil the before the woofer they placed a 100uF cap and a 1.0 Ohm 5W resistor tapped into the positive lead and ran it to the negative. Here is the question: what specific purpose does this serve? I have seen many crossovers with the inductor and the cap inline on the positive to make a band pass. Also when CV started doing it this way it was the beginning of the digital age and that was their sales pitch for the DX9, PD9, as well as the 380 SE speakers. This maybe obvious to some of you folks and a lame question but I would really like to hear/read the answer. I have included the DX9 schematic.
Thanks,
Y_Y
 

Attachments

Register to hide this ad
L1 and C1 make a 2nd order low pass filter, similar to C4 and L4 on the tweeter making a 2nd order high pass.

While the C1 and R1 components are in the right position for a Zobel, I'm still thinking they're more for filter duty. The one formula I found for calculating a Zobel shows that R1 would be sized at 1.25 x the voice coil DCR, which doesn't quite jive with the 1-ohm rating for R1.

Perhaps R1 flattens the low pass filter a bit? It's too late tonight for my mind to think about series and parallel impedance in terms of frequency.
 
HTWillie said:
L1 and C1 make a 2nd order low pass filter, similar to C4 and L4 on the tweeter making a 2nd order high pass.

While the C1 and R1 components are in the right position for a Zobel, I'm still thinking they're more for filter duty. The one formula I found for calculating a Zobel shows that R1 would be sized at 1.25 x the voice coil DCR, which doesn't quite jive with the 1-ohm rating for R1.

Perhaps R1 flattens the low pass filter a bit? It's too late tonight for my mind to think about series and parallel impedance in terms of frequency.
Click to expand...
I think you're right. I hadn't paid any attention to the values for the components, but the configuration and position within the network just seemed obviously a Zobel. But the numbers don't correspond to the formulas I have at hand. Hmmmm....... That's about the extent of my street knowledge of these things. Smarter heads will need to jump in at this point. Drats.

GeeDeeEmm
 
I had wondered about this configuration allowing for a siphoning type action (I am sure that is not the right nomenclature). The maintaining a constant resistance makes a lot of sense since CV speakers are ultra efficient throughout their power range and the DX9s are good to 400w and are 4Ohms.
 
Sounds like they moved from 1st to 2nd order filters.

The diagram in the 1st post shows an almost textbook 2nd order low pass, textbook 2nd order band pass for the midrange, and textbook 2nd order high pass filters.

Im not sure if the resistor is there to smooth the roll-off of the filter, or if it there to indeed smooth the woofer impedance.
 
bowtie427ss said:
Sounds like they moved from 1st to 2nd order filters.

The diagram in the 1st post shows an almost textbook 2nd order low pass, textbook 2nd order band pass for the midrange, and textbook 2nd order high pass filters.

Im not sure if the resistor is there to smooth the roll-off of the filter, or if it there to indeed smooth the woofer impedance.
Click to expand...
My bet is that the resistor is part of a hybrid 2nd order crossover/Zobel scheme (your second guess). I think to smooth out humps, you have to have the resistor in parallel with the filter cap/coil.
I could be full of it, and welcome being called an ignorant fool.
 
The coil, cap and resistor all act together to insure only low frequencies signal apply any voltage across the woofer. The way to think about this is the coil looks like a very, very low resistance to low frequencies. Thus the woofers sees almost all of the low frequency voltage across it. At the same time that cap is a very, very high resistance to these same frequencies, so it drains almost nothing in terms of current.

Now at high frequencies, the coil creates significant resistance. At the same time the cap is almost a dead short, so most of the high frequency voltage is dropped across the coil with nothing left for the woofer.

That explanation is somewhat simplistic, but you get the idea of how a low pass filter works. In reality, it's more complicated, because the woofer has to cross with the other drivers. That is, at the crossover frequency half of the sound comes from the woofer and half from the next drive up (could be a mid driver or a tweeter). The purpose of the resistor is to insure that the slope of the woofer cutoff matches well with the slope of the next higher driver.

In short, the speaker designers are attempting, to provide their customers with as "flat" a frequency response as they can with these specific drivers in this specific cabinet.

Hope this helps...
Jerry
 
Last edited:
Although it looks like a Zobel network, I think the resistor is to limit the high frequency roll-off and mostly to add some phase correction to the crossover. My 2 cents....
 
All of the info above explains the sonic differences heard, and felt, from the DX9 compared to many, many, many other CV speakers I have owned and heard. I have been a CV fan since junior high (1978) when I got my first pair since. This info explains why this speaker has much cleaner, tighter, and more accurate bass and mids as well as explaining why they blend better with the tweeter than many other CVs I have experienced.
 
A Zobel network would typically be used across all drivers to manipulate the load to the amplifier by minimizing the rate and increase of impedance above Fs. In your case, C1 and R1 affect the slope and shoulder of the frequency drop off dictated by L1.
 
baco99 said:
A Zobel network would typically be used across all drivers to manipulate the load to the amplifier by minimizing the rate and increase of impedance above Fs
Click to expand...
Some companies also did this specifically to better match the impedance of a woofer or midrange to the rest of the speaker system.
 
So I’ve been reading up on things because I have a pair of DX9’s. I reformed the woofs n found a crossover to be damaged. I am going to have them both rebuilt. I do have the schematic that was posted earlier in this thread, but do any of you have some clear, tested and proven upgrades to these crossovers that would add some clarity and presence to the mids/highs ? If so I will take these to the shop who is rebuilding my crossovers and see if he feels it beneficial. I’d like these speakers original otherwise, but a little cleaner filtering might help them in my 7.2+ home theater/music hall I am building. :)
Also do the CV vs150’s sound same/better/worse than DX9? About to buy a set of em.
Any help/suggestions appreciated.
 
blackie607 said:
So I’ve been reading up on things because I have a pair of DX9’s. I reformed the woofs n found a crossover to be damaged. I am going to have them both rebuilt. I do have the schematic that was posted earlier in this thread, but do any of you have some clear, tested and proven upgrades to these crossovers that would add some clarity and presence to the mids/highs ? If so I will take these to the shop who is rebuilding my crossovers and see if he feels it beneficial. I’d like these speakers original otherwise, but a little cleaner filtering might help them in my 7.2+ home theater/music hall I am building. :)
Also do the CV vs150’s sound same/better/worse than DX9? About to buy a set of em.
Any help/suggestions appreciated.
Click to expand...

The DX9 are a decent speaker. The bass is obviously the center of attention. The midrange and high range...Not so much.

Short of a complete redesign of the system, I would recommend replacing ALL of the crossover components with higher quality parts. This may require assembling new crossover boards to make room for larger components. The original crossovers used ferrite core inductors or of a rather thin gauge, and the cheapest possible electrolytic capacitors. A full recap using a good, entry level poly-film capacitor for all of the circuits (I prefer the Audyn Q4 line for clarity and price). And replacing those ferrite core inductors with air core. I would also replace the ceramic resistors with non-inductive Mills type.

You are looking at about $100-$150 in parts, plus whatever labor your tech charges.
 
The addition of the cap changes a 1st order network (-6 dB/octave) to a 2nd order network (-12 dB/octave). This further filters out remnents from the driver that you don't want because the other speaker(s) is supposed to be reproducing these frequencies and the combination creates an interference that can make that range "rough" or create a "suck-out."

The resistor reduces the rolloff created by the addition of the cap so that a bit less is filtered out.

I've been reworking my own design crossovers (JBL 2241 18" woofers / JBL 2251J 10" midranges/ Heil AMTs) and I use a 1 mH choke to limit the top end of the 10" (band-pass filter). But this lets it play too high up into the Heil range creating an interference and "glare," and a 6.8 mfd cap I recently added to ground creates additional rolloff to avoid this.

...But then there was now a "gap" between the midrange and the Heil, and the addition of a 20 ohm resistor between the cap and ground lets the 10" play up into this range filling the gap, but still cutting that area where it interfered with the Heil creating a suck-out.

Crossovers become a black art and you find that if you make you own, and are serious about getting the most out of it, you end up doing unconventional things to get the smoothest sound, especially because you are also (even if unknowingly) designing for the interaction of the speaker and the actual room itself. Factories don't do this.
 
You must log in or register to reply here.
Back
Top Bottom