What is the sonic result of this crossover design

[corElement]

Hi all

Could someone with some knowledge of crossovers take a look at this crossover design and tell me what thoughts occur to you.

Also what order of crossover is this?

I'm trying to understand which component on the diagram is responsible for the lean bass from the woofer for my own knowledge.
Not planning on doing any physical modification but rather just understand crossover.

Here is the woofer and tweeter connected to the crossover. It is crossed over at 2500hz I believe

http://www.seas.no/images/stories/vintage/pdfdataheet/e0003_w17ex001.pdf
http://www.seas.no/images/stories/excel/pdfdataheet/e0006_t25cf001_datasheet.pdf

Here are some photos of the actual crossover.

http://i.imgur.com/3XsJC.jpg
http://i.imgur.com/yYCqk.jpg

 

[Hari Iyer]

Looking at the diagram the woofer is first order and the tweeter is 3rd order. Zobel compensation is done for the woofer for constant impedance for the cross-over frequency. Compensation for tweeter for level matching is done. No series resonance componenets is used as tweeter may be ferro-fluid cooled. There is some kind of contour shaping also done for the woofer response using a inductor & capacitor.

 

[corElement]

Hmm, so R1 is the resistor for the woofer and l1 and l2 the inductor? How come there are two inductors and what purpose do they serve?

 

[Rajiv]

Hi

How come there are two inductors and what purpose do they serve?

,

One is the inductor in series for the crossover frequency and the second as Hari has answered

There is some kind of contour shaping also done for the woofer response using a inductor & capacitor.

Regards
Rajiv

 

[corElement]

Looking at the diagram the woofer is first order and the tweeter is 3rd order. Zobel compensation is done for the woofer for constant impedance for the cross-over frequency. Compensation for tweeter for level matching is done. No series resonance componenets is used as tweeter may be ferro-fluid cooled. There is some kind of contour shaping also done for the woofer response using a inductor & capacitor.

I see, yes the tweeter is ferrofluid.
What do you think in this situation could be causing a lean bass / mid-bass? and if one had to correct it or change it what would you do in my shoes?

 

[Hari Iyer]

I see, yes the tweeter is ferrofluid.
What do you think in this situation could be causing a lean bass / mid-bass? and if one had to correct it or change it what would you do in my shoes?

If you feel this speaker is a bit bright you can increase the mid-bass by adding an external baffle step compensator circuit in series with your speaker. This consist of a inductor and resistor in parallel which is connected in series with your speaker. This can be done in your speaker as its absent in your cross-over. The inductor decides at what frequency you need to apply the step and the resistor determines the amount of attenuation required. Theoretical value is -6dB but can be increased or decreased based on the balance required and the size of the room if you expect too much reflection. As this can be done externally without opening the box there is no harm of any damage to the speaker or its warranty. Always use wire guage of atleast 18SWG for this else it will reduce the SPL of your speaker due to loss in the inductor. A higher wattage resistor of 10watts will also benefit.

 

[doors666]

wow, cost cutting at this price level also. This is one of their top of the line products? but then again, I have seen much worse.
very light gauge inductor at L3, even L1 looks thin for woofer inductor. 100v caps. 5w resistors. I used better specced components in my overnight sensations which uses 20$ drivers.

The images dont match the schematics. there is an inductor and capacitor missing.

 

[corElement]

If you feel this speaker is a bit bright you can increase the mid-bass by adding an external baffle step compensator circuit in series with your speaker. This consist of a inductor and resistor in parallel which is connected in series with your speaker. This can be done in your speaker as its absent in your cross-over. The inductor decides at what frequency you need to apply the step and the resistor determines the amount of attenuation required. Theoretical value is -6dB but can be increased or decreased based on the balance required and the size of the room if you expect too much reflection. As this can be done externally without opening the box there is no harm of any damage to the speaker or its warranty. Always use wire guage of atleast 18SWG for this else it will reduce the SPL of your speaker due to loss in the inductor. A higher wattage resistor of 10watts will also benefit.

What I've noticed is the tweeter is actually not bright at all, infact it's just right for me, the mid freq's are quite excellent as well, however the LOWER mid frequencies of 350 and below are considerably lean.

At the moment I'm using the equalizer to peak 150hz +16db with slopes to compensate. I'm satisfied with it but still wanted to learn what is causing the lean lower mid frequencies. Looking at the data chart it seems the woofer is more of a mid and midbass woofer than a low one.

This inductor and resistor in series that you're talking about, is it for the tweeter or the woofer?

I can see now why people say this speaker performs very well with tube amps.

wow, cost cutting at this price level also. This is one of their top of the line products? but then again, I have seen much worse.
very light gauge inductor at L3, even L1 looks thin for woofer inductor. 100v caps. 5w resistors. I used better specced components in my overnight sensations which uses 20$ drivers.

The images dont match the schematics. there is an inductor and capacitor missing.

The foam is hiding the inductor and capacitor.

There has been no cost cutting done with the L1 inductor because logically a better coil/inductor with lower resistance (close to 0 ohm), is a better choice and many tweekers do this blindly. However it will reduce bass extension, and it will make the woofer sound even more lean than it currently is. A coil with resistance larger than 0 ohm is therefor a part of the design, and should not be replaced by a "better" one when the woofer is already weak on lower frequencies below 300hz.

However, ignoring the inductor, what kind of upgrades would you recommend for the woofers cap and resistor? I'm ignoring the tweeter side of the crossover because the tweeter is perfect to me. It's just the right amount of brightness I want.

 

[Hari Iyer]

What I've noticed is the tweeter is actually not bright at all, infact it's just right for me, the mid freq's are quite excellent as well, however the LOWER mid frequencies of 350 and below are considerably lean.

At the moment I'm using the equalizer to peak 150hz +16db with slopes to compensate. I'm satisfied with it but still wanted to learn what is causing the lean lower mid frequencies. Looking at the data chart it seems the woofer is more of a mid and midbass woofer than a low one.

This inductor and resistor in series that you're talking about, is it for the tweeter or the woofer?

I can see now why people say this speaker performs very well with tube amps.




The foam is hiding the inductor and capacitor.

There has been no cost cutting done with the L1 inductor because logically a better coil/inductor with lower resistance (close to 0 ohm), is a better choice and many tweekers do this blindly. However it will reduce bass extension, and it will make the woofer sound even more lean than it currently is. A coil with resistance larger than 0 ohm is therefor a part of the design, and should not be replaced by a "better" one when the woofer is already weak on lower frequencies below 300hz.

However, ignoring the inductor, what kind of upgrades would you recommend for the woofers cap and resistor? I'm ignoring the tweeter side of the crossover because the tweeter is perfect to me. It's just the right amount of brightness I want.

I think you should try and experiment with the inductor & resistor in parallel and this in series with the tweeter. I will PM you the values for a -6dB step. I will however need to know the width of the baffle for the step response. You can build them locally and place them in series with your +ve terminals of your speakers. You can experiment as there is nothing to loose here but only to gain. I have the data sheet for the drivers and can calculate the values for you. This will help to increase the mid-bass significantly and u will not have to use the EQ. Keep the EQ flat or better remove them after the BSC is installed.

 

[corElement]

I think you should try and experiment with the inductor & resistor in parallel and this in series with the tweeter. I will PM you the values for a -6dB step. I will however need to know the width of the baffle for the step response. You can build them locally and place them in series with your +ve terminals of your speakers. You can experiment as there is nothing to loose here but only to gain. I have the data sheet for the drivers and can calculate the values for you. This will help to increase the mid-bass significantly and u will not have to use the EQ. Keep the EQ flat or better remove them after the BSC is installed.

Hmm, so basically this will lower the volume of the tweeter and allow the woofer more gain to sound more balanced with tweeter once both are at a higher volume, is this what will happen? As for dimentions it's quite easy,

15x9.6x12.2 - 1" overall = 14x8.6x11 (thicker front baffle) so width is approx 8.6 inches.

Tell me one thing though, why do I need an inductor, why not directly a resistor to both? or just one post.

 

[Hari Iyer]

Hmm, so basically this will lower the volume of the tweeter and allow the woofer more gain to sound more balanced with tweeter once both are at a higher volume, is this what will happen? As for dimentions it's quite easy,

15x9.6x12.2 - 1" overall = 14x8.6x11 (thicker front baffle) so width is approx 8.6 inches.

Tell me one thing though, why do I need an inductor, why not directly a resistor to both? or just one post.

An inductor is a passive component whose impedance increases with the increase in frequency. At lower frequency the impedance of the inductor is low. Hence it allows low frequency and will block all high frequency due to its high impedance at that frequency (capacitor behaves reverse of inductor). The inductor + resistor in parallel will attenuate only the high frquencies due to the above behaviour and will allow the low frequency to pass which is determined by the value of the resistor.

The baffle step is required because when the mid-woofer is mounted on the baffle along with the tweeter, the low frequency has a longer wavelength (frequency = 1/ wavelength or wavelength = 1/frequency). Hence lower frequencies will have a higher wavelength and higher frequency will have a lower wavelength. Due to this the sound from the mid-woofer will not hit the baffle or in other terms escape from the baffle edge to the rear side of the speaker. But the sound from the tweeter due to the shorter wavelength will hit the baffle and get enhanced due to the reflection from the baffle. The theoretical enhancement is around 6dB and due to this the tweeter will sound brighter or will mask the low frequency that is determined by your baffle width. The BSC circuit will hence reduce the response of both the woofer and tweeter from this frequency to bring it in level to the loss of low frequency due to the baffle step. The width of the baffle will determine the frequency from which the step needs to be applied and the inductor will start attenuating the frequencies above that @ of -6dB. Hence the masking affect by the frequency will be balanced and you can hear the mid-bass well balanced with the rest of the frequencies. This theory is documented well by olson and you can find many white papers in the net.

 

[corElement]

Hmm, I see. ok I will look olson on google when I have have everythign setup.
BTW this circuit that you're talking about, how will it affect both woofer and tweeter if I'm using speaker zones A+B with entirely separate speaker wires coming from amp for tweeters and woofers. Because that's how I'm going to use them on my ca 840a, bi-wired from the amp directly.

 

[doors666]

Hi all

Here is the woofer and tweeter connected to the crossover. It is crossed over at 2500hz I believe

Please confirm this. 2500 is an extremely poor choice. The woofer cant handle that. the tweeter can go much lower than that.

Looking at the diagram the woofer is first order and the tweeter is 3rd order. Zobel compensation is done for the woofer for constant impedance for the cross-over frequency. Compensation for tweeter for level matching is done. No series resonance componenets is used as tweeter may be ferro-fluid cooled. There is some kind of contour shaping also done for the woofer response using a inductor & capacitor.

Its a metal cone woofer. has a huge breakup. see the graphs. if it has a first order crossover, that is also a very poor choice. you need to lower the level of those nasties. Crossing over a 6.5" metal cone woofer at 2500 hz with a first order crossover. It sure sounds like a recipe for disaster.

If you feel this speaker is a bit bright you can increase the mid-bass by adding an external baffle step compensator circuit in series with your speaker. This consist of a inductor and resistor in parallel which is connected in series with your speaker. This can be done in your speaker as its absent in your cross-over. The inductor decides at what frequency you need to apply the step and the resistor determines the amount of attenuation required. Theoretical value is -6dB but can be increased or decreased based on the balance required and the size of the room if you expect too much reflection. As this can be done externally without opening the box there is no harm of any damage to the speaker or its warranty. Always use wire guage of atleast 18SWG for this else it will reduce the SPL of your speaker due to loss in the inductor. A higher wattage resistor of 10watts will also benefit.

Wouldnt you need to pad the tweeter also a little more to introduce bsc.

At the moment I'm using the equalizer to peak 150hz +16db with slopes to compensate. I'm satisfied with it but still wanted to learn what is causing the lean lower mid frequencies. Looking at the data chart it seems the woofer is more of a mid and midbass woofer than a low one.

try pushing the speakers against the wall. that usually reduces the need for bsc.

I can see now why people say this speaker performs very well with tube amps.

And where do you see that

There has been no cost cutting done with the L1 inductor because logically a better coil/inductor with lower resistance (close to 0 ohm), is a better choice and many tweekers do this blindly. However it will reduce bass extension, and it will make the woofer sound even more lean than it currently is. A coil with resistance larger than 0 ohm is therefor a part of the design, and should not be replaced by a "better" one when the woofer is already weak on lower frequencies below 300hz.

Thats not correct. The resistance with thin coils could get significant. It will consume power, lowering the power going to the load. Thicker one would not reduce the bass but increase it.
The logic actually works with your tweeter as its already padded down. so to use thinner coil, just decrease the amount of padding appropriately.
Edit: Check here, the resistance of 1.0mh 20awg coill is 0.71 ohms, not close to zero.
http://www.parts-express.com/pe/showdetl.cfm?partnumber=255-048

However, ignoring the inductor, what kind of upgrades would you recommend for the woofers cap and resistor? I'm ignoring the tweeter side of the crossover because the tweeter is perfect to me. It's just the right amount of brightness I want.

redesign the crossover. lower the xo freq, use higher order xo for the woofer. implment bsc if needed.
You might want to remove the complete xo and create a new one, so that at the time of resale, the cost is not affected.

 

[corElement]

Man, this entire thread was to learn about crossovers and I feel like I've learnt a lot in just these few hours. I now understand what inductors do, what resistors do, what capacitors do and how capacitance works yet dosnt always work when one follows nothing but theory.
Thank you for having the patience to keep replying on this thread Hari.

I'm actually quite satisfied with the speaker, the sound signature, speed, transparency, grip, texture, everything about it I like, in-fact it's one of the best I've ever heard. The lean lows and lower mids is not actually that seriously bad.

I was just curious as to what causes the lean signature. And from what I've understood so far, it appears to be a minor lack of bafflestep compensation, not a very serious one for me as I more or less prefer lean bass.

 

[corElement]

Please confirm this. 2500 is an extremely poor choice. The woofer cant handle that. the tweeter can go much lower than that.

How does one notice woofer breakup? Is it when the cone starts wobbling instead of moving forward and back? Because I've see that happen on my Polk audio sub-woofer I had before. And that doesn't happen on the jamos even at high volumes.

Its a metal cone woofer. has a huge breakup. see the graphs. if it has a first order crossover, that is also a very poor choice. you need to lower the level of those nasties. Crossing over a 6.5" metal cone woofer at 2500 hz with a first order crossover. It sure sounds like a recipe for disaster.

It's possible it's a 2nd order with the capacitor attenuated + notch. The lowpass coil in the woofer section is the one which carries most of the
energy, so a correct dimensioning of it is necessary. But I feel this has already been taken into account by the designers at jamo, because the added resistance ( Ra ) represented by the (in series) coil and the wire , they influence other parameters, and eventually lowering of the Ra would bring to a bigger box?

Does that make any sense? I feel like a kid in physics class at school who just went through an entire book in less than afew hours because I couldn't even imagine writing what I just wrote 6 hours ago >.< >.<

try pushing the speakers against the wall. that usually reduces the need for bsc.

One of them was pushed against the wall 1 feet away and the other 2 feet away. Oddly enough I liked the one which was 2 feet away. *shrug* :lol:

And where do you see that

Other forums from other countries across various languages on the internet. Some say this particular custom woofer the 003 model that seas made for jamo has excellent synergy with tube amps because of their fat midrange and the woofers high quality midrange performance. Infact even the previous owner of this pair was using it on a SET tube amp from audio research. Unfortunately thats much beyond my reach haha.

Thats not correct. The resistance with thin coils could get significant. It will consume power, lowering the power going to the load. Thicker one would not reduce the bass but increase it.
The logic actually works with your tweeter as its already padded down. so to use thinner coil, just decrease the amount of padding appropriately.
Edit: Check here, the resistance of 1.0mh 20awg coill is 0.71 ohms, not close to zero.
Jantzen 1.0mH 20 AWG Air Core Inductor 255-048

Hmm, interesting. I see.

redesign the crossover. lower the xo freq, use higher order xo for the woofer. implment bsc if needed.
You might want to remove the complete xo and create a new one, so that at the time of resale, the cost is not affected.

Ok, will keep it in mind for the future when I get a chance to work on it. Most likely I wont be able to for a couple of years with masters course coming up.

 

[Hari Iyer]

I agree with doors666 that the cone break-up will result if the cross-over frequency is high. If i had to design this speaker then i would have crossed it around 1.7KHz to 1.8KHz to take the benefit of the acoustic roll-off of both the woofer & tweeter. Also would have added additional circuits for compensations that would have probably benefited. Both the electrical and acoustic roll-off together would result in 12dB/octave roll-off for a first order filter (electrical 6dB + acoustic 6dB). I wonder if the off-axis response of this speaker is good as the cross-over point selected is 2.5KHz and the off-axis response roll-off around 2KHz. Only the on-axis response of the woofer extends till 2.5KHz. Hence the imaging and sound stage theoretically would seem to be poor. A cross-over point of 1.7KHz would have benefited much better and allowed a first order filter for both woofer and tweeter without any phase issues. Also the resonating frequency of the tweeter is 700Hz ie. atleat 2.5 octaves above the cross-over point.

If you are using two separate cables for the woofer and tweeter from your amp then you will have to design 2BSC as the voice coil resistance is required for designing the values of the inductor and resistor which is different for both woofer and tweeter. Alternatively you can use a single wire and use a single BSC for both woofer & tweeter atleast for the testing.