1st order crossover experiments

The premise for this thread originated in Kenny's thread(tubes/horns...)

Anyways as seen in my avatar, I am running JBL 2226J 15" midbass driver with a 2 mh coil for shaving off rising response between 800hz and 1.6khz. I tried this without the coil, but way too boomy in the midrange. The coil actually acts as a filter, not a crossover due to the high inductance of this driver. (I believe near 3.5mh for the 16 ohm "J" version, from memory that is.
I also like using the 2226 as its response rolls off naturally between 1khz and 2khz. By 2khz, the driver has rolled off 20db and nicely quits trying to reproduce anything above that.

Several folks have asked me about crossover design involving such as the Altec 416 driver. The 416 driver does not like to roll off much until 3khz or 4khz(per my ears, no not precision measuring equipment). I have tried using 2nd order crossovers and zoebels on the 416 and don't like the way this driver keeps working beyond the crossover point. I have a crude/barely readable schematic for the Altec model 19 crossover and found they are crossing the 416 over at approx 500hz 2nd order, then crossing over the horn/driver at 1.5khz 3rd order. So I guess even Altec figured out their driver was a rebel that probably works better with bi-amping. (special note: with the 416 installed in a A7 VOTT cabinet, does perform much better, and integrated better with the horn/driver)

Back to 1st order business again. I am using a 4 ufd oil in metal cap(50VDC) from I think 1958(part of the package deal I bought several years ago, the previous owner built the entire system about that time frame). The caps did not sound great the 1st listening, but did mellow out after some playing time. I used these as they brought me close to my target crossover frequency range. I believe 6db/oct around 4.4khz(8 ohms) which should leave me at 550hz @ 24db. I am using Edgarhorns, tractix round wood. Bruce says 650hz flare, crossover at 800hz, So I feel close. I am more concerned with safety for my ALtec 902 drivers as I have nearly $400.00 invested in them.

I have added a powered sub(Titanic 12"/250watt amp/3 cu' cab) since a friend of mine sold it to me cheaply. Having to time/phase align the 3 drivers is painfully difficult. I may dump the sub, and just keep the 2-way system as it sounds better in the soundstage dept and I really don't play much music with much bottom octave bass below 45hz(bottom end of my 2226's). I have found myself playing music with heavy bass content since I have the sub, but find myself playing, not playing music. so I think the sub will go to alternate system soon.

When I removed my 3rd crossover on the horn/driver, I thought I would be lacking in the HF extension dept. I also did not know what attenuation I would want to use for my horns either, previously, i used 14db to pad down my horns. Well, here is a cheap trick I used to eliminate the need for attenuation. The Tulip 2A3 SET amp at 3.5 watts/chnl has ouput taps for 4/8/16 ohm drivers. I connected the JBL 2226J(16 ohm) to the 16 ohm tap. I then connected the horns to the 4 and 8 ohm tap for 4 ohm use. The 902's I have are 8 ohm, but I also knew that with the 2226 loading the output tranny, I could use the 8 ohm 902's on 4 ohms equivalant tap(between the 4 and 8 ohm taps on the output. This is rather unorthodox and I don't recommend anyone try this unless you are aware this is not a common connection. I just know it works for me and provides approx 6db attenuation to the horn driver. Now for the other 8db (I was originally using 14db for the same components and 3rd order crossover. Not meaning to get into a physics debate(I like being a "backyard mechanic" rather than letting a computer do my designs for me), but basically(not all-so don't shoot me :uzi, horns are basically most efficient from 1khz to 3khz, beyond that, efficiency falls and distortion rises. That refers to above, and below this bandwidth. If you wonder why I would use a device that drops off after 3khz to reach up to 15khz/18khz, you have to remember that it drops off, not cuts off. So by crossing over a driver at 4400hz, the efficiency and peaking says the sound below that will be more efficient anyways.

Anyways, I have moved my speaker system to the longer use of the room. As to the sound, the 1st order crossover is more laid back, the HF notes are not quite as crisp, but the sound is more even across the board. The powered sub created more problems for me than it solves, but not the fault of my speaker system or the sub woofer, I am playing in a 12' X 15' room with 8 ft ceilings. Even Dr Bruce Edgar does not recommend horns for rooms that don't let you listen being farther away than 15 ft. I would agree with him, I just take the compromises in stride and with my listening preference stay with the horns rather than direct drivers(Note I mentioned my preferences, not what is best. You must design a audio system that works for you, even if the experts say it is wrong, not if you like it!!!!!)

Okay, I have rambled on enough, Once the posts come to this thread, I will see about answering whatever questions I can.

GF: Sorry, I am lucky enough to have time to tweak my stereo and listen to tunes, so I never have enough time to actually measure my system. Just go with my ears, if it sounds okay, I just enjoy.

Lets hear others opinions on this subject of crossovers!!!
Cheers, Ron

Very interesting Ron as this thread comes at a great time for me. I’ve wanted to play with different crossover schemes on my Altec 604’s for a while now and 6db slopes might be a good place to start (end?). Since I’m about to start building a 45 SET that might just end up driving the horns in a bi-amp configuration, I’ll be adding a shunt bypassed Alps pot in front of the amp driver tube (417A) to allow horn level adjustment at line level instead of a pot or padding resistors after the amp.

My system is also jammed in a small room (11 x 15) but I have found that the co-incident design of the 604’s actually works ok in a small space. I have heard various Edgars and Altecs in smallish rooms (owned Altec 19s) but found as Edgar mentioned, that I only really “loved” separate mid/high horn designs if I could get far back from them.

I was going to wait till I have the 45 done but I might just try 6db slopes after my 2 existing 300B amps with some attenuation on the horn amp - not sure how I’ll do the padding yet. What intrigues me here is your idea of using different cross points on the woof & horn, especially your thoughts on using a much higher point on the horn to essentially lower its level in the 2K area – just where the 604 horn seems too hot. I’m not sure how the 604 woofer compares to your 416s but I believe they are very similar and good response up to 2K is probably a good thing with the little 604 horn. I believe the stock Altec crossover I use crosses at 1K5 with 12db lowpass and 18db highpass but I might be wrong. Based on your findings, what would you suggest I try first for high and low cross points?

Thanks, Billfort

spkrman57 said:

Anyways, I have moved my speaker system to the longer use of the room.

Click to expand...

So this begs the question; why? Did you find something wrong with a wide-way setup? <<sorry if I'm off topic here - I'll go away if you like and post this in a new thread>>

We will be constructing our retirement home this year. It should be ready for audio about September (whatever else is supposed to be there will probably follow shortly thereafter).

The house is a bungalow with a 'bonus' room on the 2nd level. My good wife has allocated this space to me :> The room is 14' 10" wide, 18' 6" long and has an 8' ceiling.

There is a double window on one short-wall end opposite the entrance from the stairs and the long walls are unbroken. The window has a view of the ocean so I don't really want to put audio gear there!

My question at this point is; should I consider setting up my equipment and speakers using a long wall so that the speakers are not close to either short wall on their sides? This will limiting my listening distance to about 10 feet but present a very wide sound stage. Or should I go for a length-wise installation with the
equipment and speakers on a short wall (probably the entrance wall) where the speakers will end up close to the side walls?

Paul

Ron...thanks for startin' the thread. I've been backyarding a couple different crossovers for my speakers and still haven't found anything that worked as well (in my situation...for me) as the 1st order(tweet)/(natural roll off of 2441) 1st order(mid)/2nd order(woofer)

Don't know...I haven't tried an active crossover yet...and perhaps should.

Reading some of your posts...I decided to take a few backyard measurements on my compression drivers. I just wanted to see how low they actually ended up responding with the 1st order crossover in place. These measurements were taken with the woofer amps turned off.

• 250Hz <60db did not measure on the RadioShack meters lowest setting
• 315Hz -28db
• 400Hz -18db
• 500Hz -12db
• 630Hz -6db
• 800Hz -4db
• 900Hz 1st order crossover point for midhorn!
• 1kHz +-0db
• 1.25kHz -1db
• 1.6kHz -1db
• 2kHz -1db
• I just figured I'd keep going
• 2.5kHz -1db
• 3.15kHz -1db
• 4kHz +2db
• 5kHz +4db
• 6.3kHz +4db
• 8kHz +-0db
• 10kHz +-0db
• 10kHz 1st order crossover point to horn tweeter
• 12.5kHz -1db
• 16kHz -1db
• 20kHz -3db

Keep in mind...I was simply wanting to check how low of a response I got from my compression drivers...but after I was already going...I figured I'd just keep going.


I still have to figure out how high up my woofer is responding...another day


that sure is one goofy set up to level match...I had to read that one twice! But hey...as long as it works for ya. I still can't believe the difference it made when I removed the attenuation from the mid/highs...making it a clear path from my new passive NOH preamp to the tube amp to the midhorn. Instead of attenuating...I needed 10db of gain on my woofer side...so I'm using a Foreplay as a tubed gain stage before the SS amp driving my woofers. Strange yes...a recommended way of doing things...no...but it seems to work pretty darn good! (I doubt I wouldn't have attempted this if I still had my active tube preamp in there. the preamp had a gain of 10db...I removed this by going with the passive transformer based pre...so I didn't feel to bad adding it back in...especially when I was clearing up everything from 900Hz on up)


keep in mind...that I am learning as I go here...just thought I'd jot down a few thoughts on what I've heard as I learn.


strange days indeed :

attempt to answer some questions here

BF,
If you are going with a 1st order crossover, you want to start around 3 octaves above your lower limit. Another words, I wanted adequate protection for my horn driver, so 4 ufd for approx 4400hz is down 12 db @ 2200hz, and 18 db at 1100hz, but horns are most efficient between 1khz and 3khz, so while the driver is down 18 db at 1100hz, the horn is more efficient at 1100hz than it is at 4400hz. so basically between around 800hz to 4khz, my horns are pretty flat(to my ears). I have used crossovers for my altec drivers that were at 500hz/800hz and have found that the horn honk is very much alive and well. Even Altec finally figured out not to crossover the horns on their systems below 1500hz. I know in my system, if the crossover I used did not work, it would be changed or reverted to the former 3rd order on the horns which is 3 years time proven to me. And for any duty other than flea powered 3.5 watts/chnl, i will change it back to the 3rd order crossover so i could use it up to 300 watts with no fear of destruction(this has been proven in my house already)
By the way, for comparison purposes, the 604 used a 515 cone and the 605 used the 416 cone, major differences between these are fs and efficiency. My luck with 1st order on 416's were a disaster for me, I recommend a wide spread crossover for that driver with 500hz 2nd ordeer low pass w/zoebel and the horn passed over w/3rd order 1.6khz. Believe it or not, the gap between 500hz and 1.6khz is not noticable since the 416 doesn't stop at the 500hz, so no hole in the sound(IMHO).

To Morden2004:

The reason I wanted to change to the long use of the room is that when you are only 6 ft away from horn-loaded speakers, it is very easy to overload your ears as I consider that "nearfield" listening with horns. I now sit 3 ft further back and it is a little better. With the wide listening position, I have to admit the soundstage was wide and big and wonderful, but sometimes a vocal soloist don't need to sound lik she is 12 ft wide. I wish i had a large room to listen to but alas, the price of being poor!!! I do say this to you set up your room so you can alternate the listening every so often and hear the benefits of both ways. And no I am glad you posted , don't you dare go anywhere else, crossovers started this thread, but it don't dominate it. and it originated in Kenny's thread and we took that all over the world and back. Glad you voiced your opinion and asked your questions, now if I only had the answers(BIG GRIN)

To GF:

On my amp to driver hookup. The reason I did not use the common and 4 ohm tap for the horns is that by using the 4 ohm and 8 ohm tap, I did not share wires on any of the taps. I could have used common and 4 ohms, but then when I change something out, would have to mess with more than one wire per terminal, so just to give each wire its own terminal.

I would like to add something to your measurement preocedure, while the effective output is handy to know, I am always worried about when the driver has no appreciable output, but is consuming current(above and below bandwidth that is) that there is enough protection for the horn driver. That is the main reason for me wanting to use a crossover point high enough so that when approaching the point of the driver passing current, but producing no output, there is little signal of that type to mess it up.

Too all:

Another reason for using that cap is that I had a pair of them, and the original plan for my 1st order crossover is using the 4 ufd cap for my Altec 288(16ohm) on 311-60 horn. The 16 ohm will bring me to 2.2khz at 6db/1100hz at 12db/550hz at 18db and 275hz at 24 db. The horn flare is 300hz, the driver is 500hz low end limit. I look at the flare frequency for the Edgar is 650hz, and the crossover for it is approx 24db @ 550hz, see how close things are. This is what I consider lowest crossover point for me, that is when the horn flare rate frequency is the same as the 24db point (using 1st order that is), plus there is peaking involved with 1st order also. GF you know this from the Pi speaker forum, I know Wayne has addressed it many times. That is why I will not use 1st order with amps over 10 watts/chnl.

Sorry to run here, I will be indisposed until Sunday or Monday, so not ignoring anyone, just not near a computer for a few days, My GF feels the need for quality time, crossovers and tube amps don't qualify in her mind.

Disclaimer: I am backyard mechanic/technician/hobbyist, so don't shoot me for my opinions, they are meant to promote involvement from others and give ideas to think about.


Regards, Ron

I understood everything up to the word "Kenny" and then it went right over my head so can you folk add a few basics that might help me and others reviewing this post.

Is octave what I think it is? is it doubling frequency? I keep trying to sing an octave higher at 6db down and just can't get it (too many cigarettes?)

Looking at GF's readings, the 1st order on the mid set at 900hz drops it down 6db at 630Hz which makes me think my octave reasoning is screwed up.

And another terminology clarification. The 2nd order on the woofer is a low pass meaning anything above the crossover point gets knocked down 12db per octave while the 1st order on the mid is high pass so anything below crossover (900Hz in GF's case) gets knocked down 6db per octave.

And why does it not affect it as much when frequency goes up, does that have to do with the 1st order on the tweet kicking in? And finally, what I'm not seeing is the 2nd order on the woofer, is that done at 12db to keep the lows at a safe distance from those compression drivers?

I realize there are capacitors and inductors that make this all happen but I'd like to grasp what work they are doing first.

TIA, learning how little I know everyday.

Ron...

Believe it or not, the gap between 500hz and 1.6khz is not noticable since the 416 doesn't stop at the 500hz, so no hole in the sound(IMHO).

Click to expand...

As far as I know...your supposed to be using this gap. Then, the summation of the to falling responses (woofer response and midrange response) have an effect together great than by themselves.

I am using Edgarhorns, tractix round wood. Bruce says 650hz flare, crossover at 800hz, So I feel close.

Click to expand...

Hmmm, you know...I just thought of something. The crossover I'm using right now is what he recommended for the midrange/horn combo and woofer/enclosure I've got. I've tried a few other things but keep coming back to what I'm using right now. Do you think that's a coincidence? Hmmm...

yep...leaving some variation for different drivers...crossover at 800 or higher for the 650Hz horns and just a bit lower for the 500Hz horns is what I've always heard.

I have used crossovers for my altec drivers that were at 500hz/800hz and have found that the horn honk is very much alive and well.

Click to expand...

I've heard that too with altec drivers when they're crossed to low.

am always worried about when the driver has no appreciable output, but is consuming current(above and below bandwidth that is) that there is enough protection for the horn driver.

Click to expand...

I basically knew that the TiTans were using a similar croosover with the same drivers...but with the 350Hz horn...they are crossed over a bit lower than I have mine. I figured I was safe going 300Hz higher.



DingusBoy...

Looking at GF's readings, the 1st order on the mid set at 900hz drops it down 6db at 630Hz which makes me think my octave reasoning is screwed up.

Click to expand...

Far as I know...your right on. But when I took the measurements, with radioshack corrected test tones...these are the exact numbers I read.

And why does it not affect it as much when frequency goes up, does that have to do with the 1st order on the tweet kicking in? And finally, what I'm not seeing is the 2nd order on the woofer, is that done at 12db to keep the lows at a safe distance from those compression drivers?

Click to expand...

I'm not sure if I understand what your saying. I think your asking why the midhorn doesn't have anything for it's top end? If so...the jbl 2441 naturally rolls off pretty decent by itself...for the tweeter coming in around 10k.

I've tried 1st order low pass line level filters on my woofers before...I had response well over 2K. I figured that was enough of that.

I realize there are capacitors and inductors that make this all happen but I'd like to grasp what work they are doing first.

Click to expand...

me too. I figure if I put my experiences down here to talk about...and read what others have to say...it may cause me to look into a direction that may benifit myself to learn. Who knows...hope it works

here's a link to some discussion on passive crossovers. There are tons out there...and just about everyone seems to have what they think works out best. I don't know...but I know what I've heard so far...and what I've liked the most so far...course, that ain't gonna stop me from looking or trying to learn.

There are I believe 3 different 416s. I have a pair with the metal dust covers that actually go up to 8K. I have these in a pair of Jensen cabinets hooked directly to a Fisher PP 7591 with no crossover of any kind on them. I do have to be very careful about not using the treble control because it will easily distort in the mids. I have isolated on top of the cabinets a pair of Heil dipoles with 5 uf 330VAC oil caps that are the size of a pack of cigs. That is 8 K with a 6 db slope so there is a bit of overlap there but the upper limit of the 416s is the weak point so it works out quite well. I sometimes biamp using SE 6BQ5s for the Heils and that works out well too.
I took this setup out to the Bottlehead meet last Saturday and we hooked up a lot of different amps to it and biamping with PP for the 15" Altecs and SE for the Heils worked out best. It didn't sound great with everything but when it clicked it worked great even at the loud volumes that tend to happen at meetings. For some reason things get turned up till distortion starts and then is backed off a bit.
I listen to this in the small bedroom where the computer is setup with the speakers facing each other and the puter desk in the middle. I believe that I typically listen at around 90 db and it is big, rich and very articulated. One of the reasons this is so is because the only parts used for crossovers are a single large capacitor per channel.
One thing I want to mention is that I was told by an old EE who was building custom tube amps and speakers in the 40s and 50s is that capacitors do have inductance. The smaller the physical size (I think) the higher the inductance is. Using 2nd and 3rd order crossovers that have small bodied capacitors like NP electrolitics could very well adversly effect the bottom end if used on a woofer. This is an impression that I got because he was explaining how the inductance could cause roll off when used in parrallell with a resister between the cathode and ground in an amp. I might be making more out of this than it merits but found it quite interesting when thinking about crossovers.
As we all know, everything in the signal path will effect the sound, so for me it is important to choose drivers that need no crossover parts or to be able to use a 6db passive. In my mains I have 6 db chokes on the subs, free running full range in TLs and a 6 db high pass on Heil dipoles in the form of high voltage oil in paper 4 uF that are 28 cu"!

GF - What are your crossover component values and the drivers you are using with them

I am curious of the drivers and crossover componenet values you are using. Could you post the drivers and the crossover componenets you are using??? Did you say you are running the Titans from Edgar???

DingusBoy, I know what we are talking about seems confusing, but in person would not be too hard to explain to you.

Ron

Just when I thought things were going good!

I had a hum problem in my system last night that would not go away. I did not have much time to research the problem, so I brought my Marantz 1060 integrated amp out to use temporarily.

Well, while what I had prior worked out pretty good, with the SS 1060, I had no way to lower the efficiency of the horns since I had no spearate taps to use.

Well, the horns were just like a runaway train, they were too hot compared to the midbass (2226). I had originally intended not to use resistors in this crossover. I figured if I had to use them, I would just go back to my 3rd order crossover on the horns since that crossover has the attenuation/HF compensation int it.

At the time being I am debating the use of the 1st order crossover, I will probably just go back to the 3rd order I had before. As for the difference between the 2 differenent crossovers, I did not find enough to justify keeping the 1st order.

I will build a new 3rd order tonight and revert to where I had it originally except for a change in the attenuation and compensation.

The one thing I lost when using the 1st order crossover was the ability to use amps with any power. With the 3rd order on the horn, I can bring the big amps back out, or play the flea powered ones also with no change.

I will try the 1st order project again, but it will be on a secondary system where I have time to play with it. With the 2A3 SET amp and the taps hookup like I had, it did play rather well, but I am limited with this amp and unable to swap out with other amps to get the same result. So my experiment worked out due to just "DUMB LUCK", I am not trying to discourage anyone, but there are still issues to be resolved for me. Biggest problem lately is just finding the time to do this.

GF: I am still curious of your drivers and crossover values you are using. Might help me figure out some of the issues I need to deal with yet.

Ron

Stay tuned for further adventures when the time permits.

Ron...I'm at work today...I'll get them to ya in the morning.

No, I don't have the TiTans. The TiTans I've listened to have used the 2440 and 2441 drivers on 350Hz round tractrix EdgarHorns, which I believe are crossed at 500Hz. My speakers are not exact...but close...I have JBL 2441 drivers on 500Hz Horns. The crossover I'm using is similar also...but crossed a bit higher.

I'll e-mail orPM ya the info

Will wait for the details

GF,
I will get back to the 1st order crossover in a week or two when the time permits. I am looking forward to your drivers/crossovers to see the design parameters, it will help me out with my situation.
Cheers, Ron

Another attempt at the 1st order fiasco

I am wondering what attenuation configuration to use to bring these back into the fold. On my 3rd crossover, I use 14db attenuation, but not sure for this setup. Also the 3rd order XO has HF compensation and I'm not sure how all this will fit in.

If the cold weather is not enough here in Ohio, my home phone is dead again(3rd time this year), My home PC is acting funny and pop server won't authorize(???). I had some e-mails w/virus attch's and I deleted my entire inbox without reading anything, but no response yet. (Iam at work now), I will try the internet explorer when I get home and see if my PC is hosed.

How about some good news from someone, I could use it about now.

Ron

Ron...that did seem kind of odd that it leveled matched everything for you. You should have played the lotto that day! But hey...if it works it works

You may need to take some measurments to get yourself level matched...get yourself the set point where everything is matched...then if you want to deviate from there fo taste...you can do so...but you'll always have your reference. I suppose you could make an l-pad or use an attenuator to get everything level matched.

If you don't have an spl meter...I could have Thor send you mine, when he gets done with it...then you could send it to me when your done.

I wonder if getting an active would just be easier!