"Coloration" of horns

[pangea]

Hi!

This is my first post here and I'm hoping to find some help regarding a suitable X-over for my JBL 2445J mounted on a short (13 cm) exponential horn.

Right now I'm experiencing some "coloration" from the horns as my x-over is 551 Hz. I have tried another x-over at 7000 Hz without any coloration what so ever. But I would like to be able to go as low as possible, but without the coloration of course.

I have noticed that the x-over on some large JBL monitors is set at 1.1 KHz, so I wonder if that x-over frequency is high enough to avoid coloration, or do I have to go even higher? :confused:

Otherwise my filter is a 12 dB/oct L/R, HP: 10uF (polypropylene) and 8 mH, L-Pad (-10 dB attenuation) 7,5 Ohm/11 Ohm with a 4,7 uF parallell over the 11 Ohm.

BR
Roland

[boputnam]

Hey, "pangea" :wave:

Great moniker, by the way (said the geologist... ;) )

Widget knows best on the 2445's, but note that the 4355 using the 2441 x's at 1,200 Hz, and the 4345 with the 2421B (smaller diaphragm...) x's at 1,300 Hz. Me thinks if you're able to do something in that range, you'd be best to do so.

What MF are you using, and is this a three- or four-way assemblage?

[pangea]

Hi!

Thanks for your advise!!!

I would say my set-up is a total hybrid, but I'm quite happy with it. I have a pair of speakers made by the former Swedish representative Tommy Jenving called "JBL KIT 65 BIG" they are somewhat upgraded with new 2402H, new 2105 although with Alnico magnets, but never the less brand new and at the bottom end a pair of brand new 2231H. On top of all that I have next to them a pair of newly re-coned 2220A in a 127 L custom build and extremely rigid enclosure made out of 25 mm MDF board with an extra panel on the baffle, with heavy bracings and all kinds of special damping materials inside!!! Rediciously well built, when come to think of it! :D
The MF sub chamber is still empty, but I hope to find a pair of reasonably priced 2123H to fill that void, in the near future! :p

On top of this speaker I have a custom build adjustable "top" for the 2445J with a pair of, dare I say this here, Selenium exponential horns from Brasil (HL 14-50). I bought them because I had the opportunity to listen to them some time ago and they sounded completly neutral then, so I thought they would be ideal for my 2445J, replacing my old 2425J with 2370 horns, which by the way also had some, not so nice "coloration" when x'ed 550 HZ. :(

All the low-end is then adequatly taken care of by a pair of (Swedish) Sentec mono blocks ACM 1 :smthsail:
The rest is driven by a H/K AVR 5000 (AVR 500, for you over there) :) The H/K sounds very nice, but has that annoying habit of suddenly turning up the volume, now and then, without even asking me.
Shame on you Harman!!!

But all in all, I'm quite happy with the lot, for the time being. :cool:

Oh yeah, I almost forgot to ask. Does anyone know if there is a "magic" line where the horns, I mean any horn becomes "colored", theoretically???

Thanks again for your help and thanks also for this great forum.

BR
Roland

[Robh3606]

Hello Pangea

I am wondering if you are crossing to low for the horns used. If you are below the cut off frequency it unloads and that could be what you are hearing. You should really run some what above the cutoff of the horn figure 1/2 octave. Depends on your crossover slopes and the type of horn flare used. The 2445 should OK at 500Hz. What are your slopes??

"2425J with 2370 horns, which by the way also had some, not so nice "coloration" when x'ed 550 HZ. "

Running 2425 at 500 hz is too low for the driver and the 2370 won't load the compression driver properly that low. Try up around 800Hz.

Oh yeah, I almost forgot to ask. Does anyone know if there is a "magic" line where the horns, I mean any horn becomes "colored", theoretically???

Yeah don't run them lower than they are designed to run. Horns are bandwidth sensitive. That is one potential but, avoidable pitfall of using them. Asside from that I have not played with too many horn types but the three I have tried were all different. You need to try a few out and decide what works for you. My favorite is still the 2344A for a 1 " driver. 2123 is one nice driver! Hey where are the pictures???!!!


Rob:)

[scott fitlin]

500hz is too low. 800hz would be much better.

[Alex Lancaster]

And if You use 2123H's, 1 to 1.2 KHz., even better.

Alex.

[pangea]

Thanks for all your help!!!

I think I'll play it safe and go for the 1.1 or thereabout, depending on which standard components come closest.

My filter slope is 12 dB/Octave, as suggested by the manual. I did chose the x-over because the manual suggested a minimum of 500Hz with a minimum 12dB slope.
What horns work that low? Does the tractrix go that low?
What about 6 dB slope, where would you set the x-over with a first order x-over and how would you design the filter and at the same time trying to make the frequency response as flat as possible up to 20 kHz?

Pictures, well I dont have any, yet. I havent got around to buying one of those digital cameras, but who knows, maybe Santa will bring me one this year. I'll see though if maybe I can squeeze some useful pictures out of my web camera. :)

By the way, does anyone have an opinion regarding which 2" driver has the best tonal qualities as well as punch?
A friend of mine says that the 2440 with a 2445 diphragm beats everything else! Any comments on that statement?

BR
Roland

[pangea]

Hi again!

Yesterday I received a somewhat unorthodox suggestion on how to solve my x-over problems. I got it from someone who supposedly is an expert at these things in Sweden. It may well be OK to use his suggestions, but I'm not sure it will give me a flat response in the high/ultra high region or even if it will work properly, since he left out the tweeter R1 9,82 Ohm, serial resistance and therefore I would appreciate your thoughts and opinion, whether you think it could do it for me.

I have attached the x-over as a JBL speaker shop file, if anyone would like to have a look at it.
The cross over is now made with the assumption that I hopefully will soon find a pair of 2123H, :) which would definitely take away my "coloration" problem at the same time. Abrakadabra!

BR
Roland

[pangea]

Here is the 2445J x-over suggestion, integrated with the rest

[pangea]

Once again the attachment was lost :(

I'll try once more to attach the picture of the x-over.

[boputnam]

Originally posted by pangea
I have a pair of speakers made by the former Swedish representative Tommy Jenving called "JBL KIT 65 BIG" they are somewhat upgraded with new 2402H... I don't see the UHF in the schematic...? :confused:

So, these are only three-way's, or am I missing something? I think you'd notice improvement with a 2405H on top. :yes:

[scott fitlin]

Originally posted by pangea
Thanks for all your help!!!

I think I'll play it safe and go for the 1.1 or thereabout, depending on which standard components come closest.

My filter slope is 12 dB/Octave, as suggested by the manual. I did chose the x-over because the manual suggested a minimum of 500Hz with a minimum 12dB slope.
What horns work that low? Does the tractrix go that low?
What about 6 dB slope, where would you set the x-over with a first order x-over and how would you design the filter and at the same time trying to make the frequency response as flat as possible up to 20 kHz?

Pictures, well I dont have any, yet. I havent got around to buying one of those digital cameras, but who knows, maybe Santa will bring me one this year. I'll see though if maybe I can squeeze some useful pictures out of my web camera. :)

By the way, does anyone have an opinion regarding which 2" driver has the best tonal qualities as well as punch?
A friend of mine says that the 2440 with a 2445 diphragm beats everything else! Any comments on that statement?

BR
Roland The 2440 is a great driver. BUT, you put the 2441 diaphragm in it. Aluminum sounds better than titanium. This will give you the tonality and crisp punch your seeking.

[pangea]

OK, my mistake that I wasn't clear enough, but I wasn't referring to the "JBL KIT 65 BIG", when I was describing the x-over. It's the other ones with the 2220A, (the soon to be incorporated 2123H) and the 2445J that have given me a few new gray strains and that is why I'm trying to find the best solution for the cross-over, whether it's a 6- or 12 dB x-over.

I have also been playing with the thaught of putting a 2405 on top of the 2445J, but if a first order (6 dB/oct) x-over can even out the frequency response, then I would think the 2405 can't add all that much, or can it!? But that's never the less a secondary issue right now. My main focus is trying to build the best possible x-over for the JBL 2445J and I'm thinking that I would prefer a first order x-over if you or anyone else, think it will work as suggested by this expert.

BR
Roland

[pangea]

Originally posted by scott fitlin
The 2440 is a great driver. BUT, you put the 2441 diaphragm in it. Aluminum sounds better than titanium. This will give you the tonality and crisp punch your seeking.

OK, thanks for your input, I'll be sure to take it to the source!!!:D

By the way, what's the difference between the 2440 and the 2441 diaphragms? Why didn't JBL do this instead of putting a weaker magnet in the 2441?
I'm also thinking, why isn't the titanium diaphragm better? Shouldn't it be better? :confused:

BR
Roland

[scott fitlin]

The 2441 diaphragm has a diamond pattern surround instead of the rolled surround of the older 40 diaphragm. The 41 has cleaner, more extended top end response.

Aluminum has a sweeter sound than titanium, which tends to sometimes be hard sounding!

The 2441 diaphragm sounds really good in a 2440 driver, I know, because I have done it.

[pangea]

OK and they fit just like that?

By the way, what should I be looking out for if I come a cross a pair? Any pit falls?

BR
Roland

[scott fitlin]

Originally posted by pangea
OK and they fit just like that?

By the way, what should I be looking out for if I come a cross a pair? Any pit falls?

BR
Roland Yes the 41 phragm fits perfectly inside the 40 driver!

Look for drivers that havent had a hard life, look good cosmetically. I mean if you see a picture of a driver thats all dented, stay away as it may have been dropped and thats not a good thing for the magnet or the gap. If they have come from a touring company you know they were used hard. But if you can find a pair from a recording studio or something they lived an easier life.

As for sellers of compression drivers, as long as they are packed properly for shipment they should arrive fine! But if they are not properly packed they can be damaged during shipping, and this has happened. Improper packing and throats can crack, etc.

They are out there and many are good. Most people are honest about what they are selling. Some tell stories, but most people are good, at least in all my buying experiences.

[pangea]

Thanks I appreciate your input!:)

BR
Roland

[boputnam]

Originally posted by pangea
I would think the 2405 can't add all that much, or can it!? :yes:

And, do go for Aluminum if you are inclined to replace those diphragms...

[pangea]

Hmmm...

Would it be OK or even an improvement to put the 41 dia in to my 2445J?

I was thinking also, if the alus are better than titans, are they also better than beryllium as well???

BR
Roland

[scott fitlin]

well, if the 45 diaphragm fits in a 40-41 I would think the 41 phragm will also fit in a 45 assembly! Not sure what youll get sonically, thats one I havent done. For sure, I like the sound of drivers with Alnico and Neodymium magnets over Ferrite ones.

Aluminum sounds really good. Beryllium is even better, but its hellaciously expensive, and JBL never made a beryllium diaphragm for the drivers your talking about.

TAD has been doing beryllium for years, and they are clean sounding, and have great transient response, but MAJOR $$$$$.

You put the 2441 phragm in your 40 driver, and as long as you have a proper horn, proper crossover point, and the right power, youll really like what you hear!

[pangea]

hmmm...

I do too like the Alnico, but I got scared when so many were talking about Alnicos loosing magnetism over the years and since they're not even from yesterday, I thought I'll play it safe. Perhaps I was wrong.

Regarding Beryllium I thought they're closer to titanium than Aluminum and therefor should be closer sonically, but hey I'm no expert on metals.

Regarding Neodymium, I'm getting mixed signals, some say they are better, some say they sound harsh, so I wount be able to tell the differenc until I can make an A/B test, will I.

BR
Roland

[scott fitlin]

beryllium has the highest stiffness to weight ratio, and because its so rigid yet so light, it makes very clean, very transparent sound!

Alnicos last and unless they have really overdriven they are usually good. And even if you do get Alnico drivers, the magnets can be recharged by JBL, or any speaker company that has the machine to charge magnets!

Neodymiums do not sound harsh, rather I find ferrite magnet compression drivers to be somewhat harsh sounding. Personally, I find neodymium magnet drivers to be a bit softer sounding than either ferrite or Alnico, and with digital sources at high playback levels, I sometimes find this to be prefferable. My all time favorite is Alnico. However, I have and am using TAD 4002 neodymium drivers with beryillium diaphragms and they sound beautiful!

But I also have JBL 2441,s and will always like them too!

[pangea]

Originally posted by scott fitlin
beryllium has the highest stiffness to weight ratio, and because its so rigid yet so light, it makes very clean, very transparent sound!

Alnicos last and unless they have really overdriven they are usually good. And even if you do get Alnico drivers, the magnets can be recharged by JBL, or any speaker company that has the machine to charge magnets!

Neodymiums do not sound harsh, rather I find ferrite magnet compression drivers to be somewhat harsh sounding. Personally, I find neodymium magnet drivers to be a bit softer sounding than either ferrite or Alnico, and with digital sources at high playback levels, I sometimes find this to be prefferable. My all time favorite is Alnico. However, I have and am using TAD 4002 neodymium drivers with beryillium diaphragms and they sound beautiful!

But I also have JBL 2441,s and will always like them too!


Yes, but isn't Titanium the second best metal, with properties much closer to Beryllium than Aluminum and therefore, shouldn't Tatanium's be closer in all aspects? Just wondering.:confused:

Good to know they can be re-charged! But how does one know when to do this?

BR Roland

[Alex Lancaster]

Actually, Aluminum has the highest stiffness to weight ratio; Ti and Be, (depending if they are alloyed, in the 3 cases), have a higher strength to weight ratio, and Al-Be can be the strongest.

Alex.

[scott fitlin]

well, regardless, both Aluminum and beryllium make great sounding diaphragms!

For the record, I dont care for the way Titanium diaphragms sound. Of course it is a matter of user preferance.

Most seem to agree Aluminum sounds better than Titanium.

[pangea]

Originally posted by Alex Lancaster
Actually, Aluminum has the highest stiffness to weight ratio; Ti and Be, (depending if they are alloyed, in the 3 cases), have a higher strength to weight ratio, and Al-Be can be the strongest.

Alex.


Thanks for the lesson! :) This I didn't know.

BR
Roland

[pangea]

Originally posted by scott fitlin
well, regardless, both Aluminum and beryllium make great sounding diaphragms!

For the record, I dont care for the way Titanium diaphragms sound. Of course it is a matter of user preferance.

Most seem to agree Aluminum sounds better than Titanium.

Could it be that preferences vary between Sweden and the US, just like the Japanees have their own preferences, different from anyone elses. I may ofcourse be wrong here, but it is my impression most people here seem to prefer Titanium.

BR
Roland

[scott fitlin]

I dont know anymore what is and isnt possible!

It is possible that some prefer the " Bite " from 2445,s with their Titanium diaphragm.

Its also possible you may have never heard a " properly " set up 2441 with aluminum diaphragm! do not take offense to this statement, but you would be surprised at how many people never heard certain things, and dont like them because they are old, or from things they have been told about them as well! If you talk to sound people and companies selling sound gear, of course they will tell you "Oh those old alnicos with aluminum phragms. No one uses that anymore!" They want to sell you what they have to sell. No one uses the old stuff, that doesnt have it, because they cant get it. Or have to scour the used market.

As for Beryllium, well, they really do sound terrific, and my TAD drivers are among the most fatigue free I have ever heard! But, units that use Beryllium are hellaciously expensive, and not many consumer grade speakers use beryllium diaphragms! The French company JM Labs is now featuring a beryllium diaphragm in their flagship speaker model.

Personally I dont care for the 2445 or the 2446, and would not have used these in the first place.

[pangea]

Originally posted by scott fitlin
I dont know anymore what is and isnt possible!

It is possible that some prefer the " Bite " from 2445,s with their Titanium diaphragm.

Its also possible you may have never heard a " properly " set up 2441 with aluminum diaphragm! do not take offense to this statement, but you would be surprised at how many people never heard certain things, and dont like them because they are old, or from things they have been told about them as well! If you talk to sound people and companies selling sound gear, of course they will tell you "Oh those old alnicos with aluminum phragms. No one uses that anymore!" They want to sell you what they have to sell. No one uses the old stuff, that doesnt have it, because they cant get it. Or have to scour the used market.

As for Beryllium, well, they really do sound terrific, and my TAD drivers are among the most fatigue free I have ever heard! But, units that use Beryllium are hellaciously expensive, and not many consumer grade speakers use beryllium diaphragms! The French company JM Labs is now featuring a beryllium diaphragm in their flagship speaker model.

Personally I dont care for the 2445 or the 2446, and would not have used these in the first place.

A close friend of mine has a pair of 2440, which he says sound better than the 2441, due to stronger magnet. He uses the 2445 diaphragms without the radial ribs, which he says are a big no no (Something with the ribs). The same ribs that are on the diaphragms in the neodymium 2451 by the way.

In my personal oppinion, I think his 2440's sound great by the way.

I don't think I have listened to any Be phragms yet, so I have to go with what others are saying.

For the time being I have no choice but to use the 2445's until I can find a pair of 2440 in good condition.

BR
Roland

[scott fitlin]

I never said the 40,s werent good, just that I like the better top end response of the 41 diaphragm. I also have several 40,s with 41 phragms and to me that sounds really good. I have stated this in other posts as well. The 41 phragm in a 40 driver just sounds really good.

I have heard the 2450J and 2451J and it is smoother than a 2446. But this particular neodymium is not my favorite, The TAD 4002 is. neodymium with Beryllium diaphragm! The beryllium diaphragm inside the TAD driver has no ribs, its a smooth surface dome.

Crossover has much to do with everything, as Robh already said, and your expert is telling you.

My guess is that for the definitive answer, why dont you call JBL, explain to them which horn/driver combination your using and ask what they recommend!

[pangea]

Originally posted by scott fitlin
I never said the 40,s werent good, just that I like the better top end response of the 41 diaphragm. I also have several 40,s with 41 phragms and to me that sounds really good. I have stated this in other posts as well. The 41 phragm in a 40 driver just sounds really good.

I have heard the 2450J and 2451J and it is smoother than a 2446. But this particular neodymium is not my favorite, The TAD 4002 is. neodymium with Beryllium diaphragm! The beryllium diaphragm inside the TAD driver has no ribs, its a smooth surface dome.

Crossover has much to do with everything, as Robh already said, and your expert is telling you.

My guess is that for the definitive answer, why dont you call JBL, explain to them which horn/driver combination your using and what they recommend!

I have tried to contact JBL in Sweden several times, but most of the time they haven't even replied.

Last week I managed to talk with the head there and briefly he gave me the values on a 12 dB x-over that would lower the hump 10 dB, thus flattening the total response. Unfortunately he did not have the time to discuss the 6 dB alternative, which I would prefer if only it works as intended.

BR
Roland

[Guenter]

Roland, do you mean 700Hz rather than the 7000Hz you typed? I mean, crossing over at 7000??? does the horn play at all?? If it is 700Hz, then why not simply leave it there, it's plenty low already.

regards

[Mr. Widget]

Originally posted by pangea
A close friend of mine has a pair of 2440, which he says sound better than the 2441, due to stronger magnet. He uses the 2445 diaphragms without the radial ribs, which he says are a big no no (Something with the ribs). The same ribs that are on the diaphragms in the neodymium 2451 by the way.

In my personal oppinion, I think his 2440's sound great by the way.

I don't think I have listened to any Be phragms yet, so I have to go with what others are saying.

For the time being I have no choice but to use the 2445's until I can find a pair of 2440 in good condition.

BR
Roland

Your friend may prefer the 2440, but it is probably due to the response peak above 8KHz when compared to the 2441s slight dip above that frequency. According to the published specifications, the 2440 does have a higher flux density than the 2441, but they use identical motors. I don't know why the numbers are different, I would guess it is due to different measuring techniques, but I am not sure. In any event what Scott has been saying is true. The aluminum diaphragms do have lower distortion and I and most people I have talked to prefer their sound quality over the titanium diaphragms. You certainly can interchange the 2440, 2441, and 2445 diaphragms in all three motors. I am not sure, but I would guess that the 2441 or 2440 diaphragm in the 2445 driver would probably sound virtually identical to a stock 2441 or 2440 despite it's use of ferrite instead of alnico. If you use the 2445 diaphragm in the 2440/2441 motor you will get an extended HF output and different sound character, one that you may prefer, but for accuracy and lower distortion I would recommend the 2441 diaphragm. On a related note as was pointed out by Bo on another thread changing his older aluminum diaphragms for new ones of the same design made a huge improvement due to the aging of the diaphragm's surrounds. This is also a possible explanation for your friend's preference. He may have compared old 2440 diaphragms to new Ti diaphragms.

As far as Be is concerned. TAD has always used it as they have always produced a line of cost no object drivers. Yes JM labs is now using it in their extremely expensive speakers and even our own JBL is now using it in their flagship K2-S9800s. It seems that the material is universally accepted as the superior material if cost is no object.

[pangea]

Originally posted by Guenter
Roland, do you mean 700Hz rather than the 7000Hz you typed? I mean, crossing over at 7000??? does the horn play at all?? If it is 700Hz, then why not simply leave it there, it's plenty low already.

regards

Hi Guenter!

Which post are you referring to?

BR
Roland

[pangea]

Originally posted by Mr. Widget
Your friend may prefer the 2440, but it is probably due to the response peak above 8KHz when compared to the 2441s slight dip above that frequency. According to the published specifications, the 2440 does have a higher flux density than the 2441, but they use identical motors. I don't know why the numbers are different, I would guess it is due to different measuring techniques, but I am not sure. In any event what Scott has been saying is true. The aluminum diaphragms do have lower distortion and I and most people I have talked to prefer their sound quality over the titanium diaphragms. You certainly can interchange the 2440, 2441, and 2445 diaphragms in all three motors. I am not sure, but I would guess that the 2441 or 2440 diaphragm in the 2445 driver would probably sound virtually identical to a stock 2441 or 2440 despite it's use of ferrite instead of alnico. If you use the 2445 diaphragm in the 2440/2441 motor you will get an extended HF output and different sound character, one that you may prefer, but for accuracy and lower distortion I would recommend the 2441 diaphragm. On a related note as was pointed out by Bo on another thread changing his older aluminum diaphragms for new ones of the same design made a huge improvement due to the aging of the diaphragm's surrounds. This is also a possible explanation for your friend's preference. He may have compared old 2440 diaphragms to new Ti diaphragms.

As far as Be is concerned. TAD has always used it as they have always produced a line of cost no object drivers. Yes JM labs is now using it in their extremely expensive speakers and even our own JBL is now using it in their flagship K2-S9800s. It seems that the material is universally accepted as the superior material if cost is no object.

Hmmm...

You may well be right there.
Interesting that the 2440 and 2441 are using the same motor and therefore should have the same flux.

I'll try to get hold of a pair of 2441 diaphragms, I know a person who's got a few from a bankruptcy auction.:D

The TAD's are way out of my league. :(

BR
Roland

[Guenter]

Roland, your very first post in this thread.

-regards

[PSS AUDIO]

Originally posted by scott fitlin
The 2441 diaphragm sounds really good in a 2440 driver, I know, because I have done it.

Hello Scott,

I was said that the only difference between a 2440 and a 2441 was the diaphragm?

Is it right?

How can I recognize a 2440 than a 2441 as their is no more rear plates?

[pangea]

Originally posted by Guenter
Roland, your very first post in this thread.

-regards

OK, I had a pair of 16Ohm HP 7000Hz x-overs on the shelf, so I simply tried them out for comparison, just to see if there was any coloration left, but there was none, so I figured there maybe was i magical line which one shouldn't cross.
Little did I know that it's when you go to low for the intended horn and tap in to the unloaded frequencies, when the problem with coloration occurs.

But thanks to you guys on this great forum, I have learned alot already. :)

THANKS!!! :thmbsup:

BR
Roland

[Guenter]

[QUOTE]Originally posted by pangea
[Little did I know that it's when you go to low for the intended horn and tap in to the unloaded frequencies, when the problem with coloration occurs.

It is not the low, unloaded, frequencies that are the problem. As I understand it form Drew Daniels, see his comments yourself elsewhere on this site, the problem mainly arises because of the possibly mismatched off-axis response near the xover frequency
of the two drivers. Case in point is the typically 500Hz crossover point used between the 15" base driver and the horn. While the former is starting to beam, the latter is essentially omnidirectional. Makes much sense and, aferall, he should know.

-regards

[Mr. Widget]

Originally posted by PSS AUDIO
Hello Scott,

I was said that the only difference between a 2440 and a 2441 was the diaphragm?

Is it right?

How can I recognize a 2440 than a 2441 as their is no more rear plates?

With the newer 2440/2441 there is no physical difference between them. If you open the back and it has a 2440 diaphragm it is a 2440, conversely if the diaphragm is a 2441... You can tell the diaphragms apart by two things, first the surround is a simple concentric crimp in the 2440 and a series of diamond shaped radial dimples in the 2441. Also the 2441 has a small dot applied to the center of the diaphragm.

If you have an older driver with the flat back it would have to be a 2440/375 as all of the 2441/376s were done with the chamfered back cap.

[TimG]

There are advantages and disadvantages to all of the diaphragm materials being discussed. If you already have 2445 drivers your choices for "new" diaphragms are limited to 2441 aluminum, 2445 bare titanium and 2450SL coated titanium. I have heard from 3 sources that JBL has stopped making the 2441 domes, so buy them fast if you want them. I would expect the titanium domes to be available well into the future, unless JBL starts phasing out all of the 4" drivers and uses only 3" and 1.75" drivers in their newer products. Since titanium has much better fatigue resistance than aluminum, they would last longer if used with the same amount of power. This may be something to think about if you buy 2441 domes now and need to replace 1 in the future.

If you want berrylium diaphrams now, you will have to buy TAD drivers. Cost is the big issue with these drivers. You can get JBL aluminum or ti domes for $150 now but I the TAD replacement domes sell for around $650 each.

With any of these compression drivers, I would stay away from 6dB per octave crossovers unless you want to spend a lot of money on replacement diaphragms.

People have previously mentioned the improved transient response of first order networks, but the schematic you posted is not a first order acoustic network, and first order networks need to meet special criteria to have the advantages of first order acoustic networks. The theoretical advantages of first order acoustic networks only apply when the driver in question, in combination with the cabinet and network, produces an acoustic response that matches the first order response curve. Achieving this is a lot more complicated than it sounds, and will definitely not be achieved with a single element (capacitor or inductor) crossover. A single element crossover will create a first order electrical network, but a first order electrical network does not give you a first order acoustical response curve. To build a true first order acoustic crossover normally requrires mutiple components and equilization circuits to approach a first order acoustical response, and it will almost always require more components than say a 4th order LR crossover. When trying to reach a crossover acoustical target one has to consider both the inherent driver response mounted in a specific horn or baffle, as well as the transfer function of the electrical components of the crossover.
From this site http://www.ijdata.com/ you can download a demo version of the LspCAD crossover design software and experiment with crossover design with some sample drivers. You can experiment with passive and active crossovers and target different acoustic slopes using the optimizer, such as first order, and 4th order LR. Being able to see graphically the changes that different components makes in the crossover makes it easier to see why true 1st order acoustic crossovers are difficult to build.
There is also a small file that you can download at this site to see if you can hear the difference between a not distorted sqaure wave at 1000Hz and a wave distorted by a 4th order acoustic network.

[scott fitlin]

But as Tim says, and I said already, I would not use a first order network with compression drivers, and especially crossing over at 500hz!

[pangea]

Originally posted by TimG
There are advantages and disadvantages to all of the diaphragm materials being discussed. If you already have 2445 drivers your choices for "new" diaphragms are limited to 2441 aluminum, 2445 bare titanium and 2450SL coated titanium. I have heard from 3 sources that JBL has stopped making the 2441 domes, so buy them fast if you want them. I would expect the titanium domes to be available well into the future, unless JBL starts phasing out all of the 4" drivers and uses only 3" and 1.75" drivers in their newer products. Since titanium has much better fatigue resistance than aluminum, they would last longer if used with the same amount of power. This may be something to think about if you buy 2441 domes now and need to replace 1 in the future.

If you want berrylium diaphrams now, you will have to buy TAD drivers. Cost is the big issue with these drivers. You can get JBL aluminum or ti domes for $150 now but I the TAD replacement domes sell for around $650 each.

With any of these compression drivers, I would stay away from 6dB per octave crossovers unless you want to spend a lot of money on replacement diaphragms.

People have previously mentioned the improved transient response of first order networks, but the schematic you posted is not a first order acoustic network, and first order networks need to meet special criteria to have the advantages of first order acoustic networks. The theoretical advantages of first order acoustic networks only apply when the driver in question, in combination with the cabinet and network, produces an acoustic response that matches the first order response curve. Achieving this is a lot more complicated than it sounds, and will definitely not be achieved with a single element (capacitor or inductor) crossover. A single element crossover will create a first order electrical network, but a first order electrical network does not give you a first order acoustical response curve. To build a true first order acoustic crossover normally requrires mutiple components and equilization circuits to approach a first order acoustical response, and it will almost always require more components than say a 4th order LR crossover. When trying to reach a crossover acoustical target one has to consider both the inherent driver response mounted in a specific horn or baffle, as well as the transfer function of the electrical components of the crossover.
From this site http://www.ijdata.com/ you can download a demo version of the LspCAD crossover design software and experiment with crossover design with some sample drivers. You can experiment with passive and active crossovers and target different acoustic slopes using the optimizer, such as first order, and 4th order LR. Being able to see graphically the changes that different components makes in the crossover makes it easier to see why true 1st order acoustic crossovers are difficult to build.
There is also a small file that you can download at this site to see if you can hear the difference between a not distorted sqaure wave at 1000Hz and a wave distorted by a 4th order acoustic network.

What about Diaphragms made by for instance Radian, which supposedly are made at the same factory? Wouldn't that work for the time being, or is the quality inferior on those phragms?

Where would you set the crossover point IF you would have to go for the 6 dB slope for any reason?


I think I've managed to go around some of the problems you mentioned though, by putting the driver in a loose top cabinet, which is adjustable in all directions, except up and down.

I tried the demo version of the (Swedish) LspCAD but unfortunately it was restricted to 4th order calculations only. :(

Also I listened to the sound file through my computer, but wasn't able to hear any differences apart from a faint "clicking" sound in the middle. But then again, maybe I wasn't supposed to even? ;)

BR
Roland

[TimG]

I have no personal experience with the Radian replacement diaphragms, but other people have had good experiences with them.

Are you asking about a 6dB electrical or acoustic crossover? I would not recommend a 6dB acoustic crossover on a 2" throat compression driver. Compression drivers just don't have the excursion to handle 6dB acoustic slopes. A 6dB electrical network is not going to give you a 6dB acoustic slope, so without modelling software you can't know exactly what you are going to end up with.

I would contact Ingemar about that issue. I have the full version of the program and it was a bargain if you are going to be doing any serious speaker building. LspCAD basic includes the modelling software and a basic MLS measurement system. All you need is a sound card like the Soundblaster Live and a Panasonic microphone capsule and you can get actual measurements of the drivers you are using. Even without calibration files the Panasonic capsules are accurate, only rolling off in the top octave. The program includes instructions on making a simple system with some resistors to obtain impedance readings on your driver, which is critical for developing accurate crossovers.

If this seems like too much work or expense, here is my suggestion for the safest solution for the drivers you are working with. Just buy the JBL 3107 crossover from the tent sale. I'm sure someone here would help you with shipping the crossover to Sweden for you. This is the crossover for the upper 3 drivers from the 4350 and the 2311 horn used in this model is very similar to the horn you are using. You could probably even add a 2308 lens to your horns and get even closer response. The 3107 will give you adequate protection so you don't blow your compression driver and you could add a 2202H and a 2405 and be set for the top 3 drivers of a 4 way system.

The point of the sound file is to see whether you can hear the impulse distortion caused by a 4th order crossover (the sound after the click) versus the lack of impulse distortion from a transient perfect crossover (the sound before the click). If the difference isn't offensive to your ears, that implies that you won't find a 2nd, 3rd, or 4th order acoustic crossover offensive.

Here is an excellent page link that shows the impluse response of some different crossover slopes and the necessary conditions for a 6dB acoustic crossover to have perfect impulse response. If there is too much traffic at the site you may not be able to link to it so try again in an hour. http://www.geocities.com/kreskovs/TimeAligned1.html
http://www.geocities.com/kreskovs/Phase-B.html
Main site http://www.geocities.com/kreskovs/John1.html

There are many disadvantage to 1st order acoustic crossovers that sometimes don't get mentioned. I will admit my bias that I don't believe they are the ultimate crossover. One reason is that the severly reduced protection of the drivers from overexcursion will increase distortion dramatically, and greatly increase the risk of damaging your expensive compression drivers. Isn't low distortion and wide dynamic range one of the reason you went with compression drivers in the first place?

If anyone can take measurements of amplitude and impedance on your driver on your horn I can help you design a crossover. I have both LspCAD and Soundeasy modelling software.

[pangea]

Hi Tim!

Thanks for your input! Feels good to know it wont be impossible to get some phragms in the forseeable future.

I noticed the JBL 3107 is for a 8 Ohm driver. Is there an easy way to convert the network to a 16 Ohm driver?

OK, it seems I wont be able to go on building speakers without some special tools like the LspCAD, tone generator, mic, SPL meter and soforth. What about a scope, do I need one of those, or is that only to show off among my friends?

Any particular types of equippment you could recommend?
What to think about and what to avoid?
What does the pink noice generator do, that a ordinary tone generator doesn't do?
What's the differens between a SPL-meter and a dB-meter?

BR
Roland

[Mr. Widget]

Originally posted by pangea
Hi Tim!

Thanks for your input! Feels good to know it wont be impossible to get some phragms in the forseeable future.

I noticed the JBL 3107 is for a 8 Ohm driver. Is there an easy way to convert the network to a 16 Ohm driver?

OK, it seems I wont be able to go on building speakers without some special tools like the LspCAD, tone generator, mic, SPL meter and soforth. What about a scope, do I need one of those, or is that only to show off among my friends?

Any particular types of equippment you could recommend?
What to think about and what to avoid?
What does the pink noice generator do, that a ordinary tone generator doesn't do?
What's the differens between a SPL-meter and a dB-meter?

BR
Roland

Hmmmmmm.

[pangea]

I know, I'm sorry, but the questions never seem to end. They just keep on popping up from out of nowhere in my head. :p

BR
Roland

[TimG]

The 3107 was designed for a 2440 (16ohm) midrange mounted on a 2311 horn with 2308 lens, a 2405 (16ohm) supertweeter and a 2202A or 2202H (8ohm) midwoofer. The 2440 was always a 16 ohm driver and according to Giskard the 2405 was always a 16 ohm product as well since only one type of diaphragm was ever made for it. It seems that all the 3107s are gone from the tent sale, they will probably show up next week on ebay at an inflated price. If you search the forum you should be able to find the updated schematic for the 3155 crossover that you can build yourself. It was discussed a few weeks ago. The 3155 is the crossover from the 4355 monitor. The crossover points are at 290, 1200, and 10kHz. http://www.jblpro.com/pub/obsolete/4355.pdf

All you need to get started measuring speakers is the software, LspCAD and Soundeasy are the most popular, a soundcard that can be used in duplex mode (I just used the SB Live as an example) and a panasonic microphone capsule mounted on a stick. The tone generators and measurement ability are part of the software packages. Both software packages include basic measurement systems. You don't need a scope or any other fancy laboratory equipment to take basic measurements. Both software packages also have user groups on Yahoo if you have questions. To make impedance measures you just need some resistors and some test leads. You can get started measuring speakers and desinging crossovers, assuming you have the ability to build boxes, for a little bit more that you would spend on one 2440 replacement diaphragm. If you decide you want to buy a software package, I can provide you with links to user guides for each program that walk you through the entire process.

[pangea]

OK, Thanks again for your help and yes please I would love to have those links! :D

I've tried to read the 3155 schematic pdf document, but it is very hard if not impossible to make out the details.

Does anyone have an edited and perhaps also a "stripped" version of that particular network, with only the essentals in it?

BR
Roland

[TimG]

http://www.audioheritage.org/vbulletin/showthread.php?s=&threadid=774&highlight=3155
This is the link to the modern version of the crossover. You can buy all the parts you need for this from a variety of vendors. I would recommend madisound or partsexpress in the US, but I'm sure there is some place closer to you.

Also remember, the 4350/55 is a 4 way design with an active crossover between the woofers and midbass. The link above explains how some other people have dealt with the crossover below 290Hz.

I will look up those links for you later tonight.

[PSS AUDIO]

Hi,

Have a look at it, its quite easy to build, I may have the parts.

If you need help or parts ask ALLBA, our distributor in Sweden for the parts, we can add them while their next shipment!

[pangea]

Why are all the x-overs, mentioned in regard to the 2445, in their schematics, drawn as 8 Ohm drivers, if and when they almost always have 16 Ohm impedance, for instance the 3155, 3107 and others? :confused:

BR
Roland

[4313B]

I've explained this before, here it goes again -

This is the link to the Standard Test Fixture (http://www.jblproservice.com/pdf/Network%20Schematics/Standard%20Test%20Fixture.pdf). You take a network to be tested and hook it up to this text fixture. You run a voltage drive on the network and check the results against the specifications laid down in that particular network's matrix.

So for the 3107 at 800 Hz with the test fixture set to 8 ohms the MB (midbass) tap should read -4 dB and the MR (midrange) tap should read - 17 dB. This has nothing to do with the actual electrical "crossover" with driver loads and it has nothing to do with the actual acoustical "crossover" of the system. It's simply a standard, consistent way of measuring all the various networks.

[pangea]

Originally posted by Giskard
I've explained this before, here it goes again -

This is the link to the Standard Test Fixture (http://www.jblproservice.com/pdf/Network%20Schematics/Standard%20Test%20Fixture.pdf). You take a network to be tested and hook it up to this text fixture. You run a voltage drive on the network and check the results against the specifications laid down in that particular network's matrix.

So for the 3107 at 800 Hz with the test fixture set to 8 ohms the MB (midbass) tap should read -4 dB and the MR (midrange) tap should read - 17 dB. This has nothing to do with the actual electrical "crossover" with driver loads and it has nothing to do with the actual acoustical "crossover" of the system. It's simply a standard, consistent way of measuring all the various networks.

This is getting more and more complicated and possibly even too complicated for my head.

Does this mean it's OK to just use the component values in the "8 Ohm" drawings when using 16 Ohm drivers, Or should I simply give it up and bang my head against the wall? :banghead: :confused:

BR
Roland

[4313B]

Originally posted by pangea
Does this mean it's OK to just use the component values in the "8 Ohm" drawings when using 16 Ohm drivers, Or should I simply give it up and bang my head against the wall? :banghead: :confused:

BR
Roland

It isn't an "8 ohm" drawing. The 3107 network is designed for an "8 ohm" 2202 load, a "16 ohm" 375/2440/HL93/2311/2308 load, and a "16 ohm" 2405 load. The 8 ohm dummy loads are simply to give an idea what the voltage drive should look like for service purposes only.

If it seems confusing then just block out everything else on the network service bulletin and concentrate only on the schematic. Keep in mind that these things aren't meant for public consumption. If you don't fully understand them don't sweat it. All you care about is the schematic itself and the values of the capacitors, chokes, resistors and how they are wired together. You already know what the originally intended drivers are and what their load ratings are.

Now, if you decide to pop an "8 ohm" diaphragm into a 2440, THEN you will need to make adjustments accordingly. Different horns also present different impedence loads so if you are using a different horn you will need to adjust the impedance accordingly. Chances are the acoustical properties will be so hosed with a different horn that the impedance load won't matter anyway, but maybe you will get lucky. It does happen.