Truextent Diaphragms in JBL four-inch compression drivers

[4313B]

These are frequency response and impedance plots. They don't convey how the various diaphragms actually sound on the various
horns. They are used to aid in the development of passive filters.

Yuichi A-290, a 290hz with a hyperbolic flare - it's a clone of the TAD TH-4001

These are some of the last measurements I took. The impedance curve of the 2445 core just looked weird. I thought it might be the felt instead of the foam in the back cap so I swapped the felt 2445 back cap with the foam 2446 back cap and ended up with pretty much the same deal above ~ 500 Hz.

The 2445 and 2446 have different phase plugs but it doesn't appear that the 2446 has any better top end response.

The plots are 1/2 octave.

[4313B]

The 2446H on the three different horns.

[4313B]

2446H and Truextent on 2311

[4313B]

2446H and Truextent on the 350hz Edgarhorn (tractrix flare).

[4313B]

2446H and Truextent on the Yuichi A-290, a 290hz with a hyperbolic flare - it's a clone of the TAD TH-4001

[4313B]

2445 core
2311 /2308 / baffled
D16R2441
D16R2445
Truextent

[4313B]

2445 core
350hz Edgarhorn (tractrix flare)
D16R2441
D16R2445
Truextent

[4313B]

2445 core
Yuichi A-290, a 290hz with a hyperbolic flare - it's a clone of the TAD TH-4001
D16R2441
D16R2445
Truextent

[4313B]

These were taken a few nights earlier using a 2441 core with a D16R2441 diaphragm. The 2446H plot shown here was taken at that time too. That was when I ran into trouble. The D16R2445 diaphragm wouldn't fit right in the 2441 core and when I went to take a Truextent diaphragm out of a 2445 to put in the 2441 there were mold issues (#51 (http://audioheritage.org/vbulletin/showpost.php?p=283024&postcount=51)) in the 2445 due to the Arlene's glue that was used to adhere the felt pad to the back cap. I called it a night.

[originaltubino]

Compared to the others, the Truextent diaphragms do not appear to have any special FR extension in the 14K to 20K range.

The Yuichi and Edgar do not differ as much as I expected. This is all on-axis, right? Both are pretty well-behaved.

THANKS for doing this and sharing it!

[timc]

I'm surprised that there are acutually less HF extension with the Be diaphragm than with the original Ti's.

Any idea why?

[Ian Mackenzie]

Here is a link to Yuichi's own measurement.

http://www.geocities.jp/arai401204/Horn/A290S/A290-S90.html

I dont have a 2445 / 2446 core to compare but my own measurements of the 4001 confirm the HF extension.

With a simple two stage passive EQ / impedance circuit per Tads design its quite easy to get a flat response and impedance from 600 hertz response to 18 khertz.

Controlling or should I say damping the impedance has a marked effect on sound quality.

I expect the Tad diaphgram surround and the 4001 phase plug play a big part on how it performs on paper. As to subjective impressions its a matter of taste.

[Zilch]

I'm surprised that there are acutually less HF extension with the Be diaphragm than with the original Ti's.

Any idea why?Absence of the diamond surround, perhaps.... :blink:

[timc]

Could be. Softer surround (higher compliance) should reduce HF extension. Good call Zilch.

[4313B]

Compared to the others, the Truextent diaphragms do not appear to have any special FR extension in the 14K to 20K range.

The Yuichi and Edgar do not differ as much as I expected. This is all on-axis, right? Both are pretty well-behaved.

THANKS for doing this and sharing it!They are nice horns. Yes, this is all on-axis. I did a few off-axis shots just to see where the horns started dropping off. I was especially curious as to how tight a sweet spot the Edgar might have.

Here is a link to Yuichi's own measurement.

http://www.geocities.jp/arai401204/Horn/A290S/A290-S90.html

I dont have a 2445 / 2446 core to compare but my own measurements of the 4001 confirm the HF extension.

With a simple two stage passive EQ / impedance circuit per Tads design its quite easy to get a flat response and impedance from 600 hertz response to 18 khertz.

Controlling or should I say damping the impedance has a marked effect on sound quality.

I expect the Tad diaphgram surround and the 4001 phase plug play a big part on how it performs on paper. As to subjective impressions its a matter of taste.Thanks Ian. That link was helpful. I know nothing about these Yuichi horns. They look and measure nice. I'll have to hear a pair someday.

It looks like he got a bit more extension from his 2441 than I get from either of mine. I tried both on the Yuichi and determined that they were probably a wee bit low on flux density. :p That's why I pulled their diaphragms and put them in the 2445 core.

I'd like to see a schematic of the Tad circuit.

Softer surround (higher compliance) should reduce HF extension.It does. The Truextent uses less beryllium as well as the soft surround (as compared to the 476Be diaphragm) with the result being reduced extension. The Truextent also has a bit higher distortion according to JBL's measurements.



The titanium and aluminum diaphragms can't really compete with the sound quality of the berylliums. The Tad and JBL four-inch beryllium compression drivers are very expensive (and in the case of the JBL, nearly impossible to get). These aftermarket beryllium diaphragms offer excellent cost/performance.

THANKS for doing this and sharing it!You're welcome. I really don't enjoy handling other people's beryllium diaphragms and swapping them around in various cores. Everytime one is exposed I get real nervous.

When I ran the impedance curve on the second 2445 with the Truextent installed I didn't like the way it sounded. There were unwanted resonances. I opened it up, cleaned out the little bit of mold on the felt in the backcap and then cleaned the gap out (there were a few metal particles in the gap) and then reassembled it. That fixed the problem. I'll check the first 2445 (the one with the mold) today so Dan can get two fully functional drivers back. Don't use Arlene's glue in these back caps guys. ;) I used the Moyen recone glue in my 2445J's and there weren't any issues.

[Ian Mackenzie]

I will try and find the schematic. Its a lot simpler then you might imagine.;)

Yuichi also has a 1.5 to 2 inch adapter for using 1.5 inch drivers (people with Tad 4003s are not complaining..except for the price):bouncy:

What you have measured about these alternative Be diapgragms is consistant with what I have been advise by company that uses Truextent diaphragms in field coil compression drivers with a Tractrix horn, hence they use a Be super horn tweeter above 16 khertz. The tractrix horns are apparently subjectively better then the Tad 4001 horn but you need a super tweeter.:blah:

So don't throw out your 2405s yet.

I am inclined to think the 0476 with the bi radial is still the winner:).

What this means is as usual there is no free lunch but you can have fun making it work.

The notion of a flat response to 20 khertz is more important on paper to the Japanese than it is subjectively anyways provided the diaphragm is not in breakup mode too early.

A gentle tappering off above 10 khertz can be a good thing in some situations;)

I think Greg has mentioned this before.

By the way I got some complaints from the local natives attempting to run a series external ground plane measurements with LMS. 2.83 volts at 110db sensitivity above 10 khertz is freakin loud if you are a dog I guess.:rotfl:

Ian

[cooky1257]

That's a shed load of work for the benefit and enlightenment of us all.
Thank you.
Anyone tried a Yuichi 290 with a 3" Be driver?.....Doh! thanks Ian.

[Robh3606]

Hello 4313B

Thanks for measuring and posting all those measurements.

Rob:)

[cosmos]

Thanks 4313B for doing these measurements. As it turns out, I am even more glad as you uncovered and resolved a couple problems as well. I can't thank you enough for that! :applaud:

If I am interpreting this correctly, the FR looks very similar to the stock Titanium diaphragms they are meant to replace. That is probably a good thing in some respects as it allows a similar compensation scheme and at the same time offers significantly improved quality of sound.

Ian, I am looking forward to seeing the schematic of the TAD compensation circuit. In addition, I'd really like to see any other adaptable compensation networks. Getting reasonably flat to 18KHz without a tweeter and with this quality sound would be a dream come true for me. I've heard these drivers with the TrueXtents and they sound great to me.. besides, I already own them.. :)

476Be would very likely be a better driver, (I have never heard them) but these are reasonably priced and readily available.

Now, I really wonder how TAD TD-4001 would measure in the same setup.

In any case, Thank you again for your time and help!

[originaltubino]

476Be would very likely be a better driver, (I have never heard them) but these are reasonably priced and readily available.

Claude Rains(Louis):The waters? What waters? We're in the desert.
Humphrey Bogart (Rick): I was misinformed.


Now, I really wonder how TAD TD-4001 would measure in the same setup.

Me too, and with the compensation scheme Ian mentions.

That's the problem with generous competence. It only gets you more work requests! :)

[spkrman57]

First off I would like to point out that the 350hz Edgarhorns are mine and are the ones in my avatar. :D

I had tried some EV DH1A compression drivers on them in the past. Dan (Cosmos) also heard them when he visited me before. We both agreed the HF extension was decent enough(passive crossover @ 800hz using passive attenuation/EQ), but the "simbilance" or what I call the "titanium sound" was readily evident.:banghead:

I plan to use a 4-Pi (Wayne Parham) 1.6khz crossover w/JBL 2226J's on the bottom end w/JBL 2446 compression driver equipped with Be diaphrams as soon as I can afford to make that happen!:)

I see with the 2446/Be/Edgarhorn plots that the 1.6 khz region looks to be fairly flat response wise. I don't know enough to extrapolate all the info here so I'm actually speculating!:blink:

Thanks a million for the help 4313B!!!:applaud::applaud::applaud:

Regards, Ron

PS: Did my e-mail with the att/EQ chart help show what Wayne is trying to do with his crossovers?

[4313B]

If I am interpreting this correctly, the FR looks very similar to the stock Titanium diaphragms they are meant to replace. That is probably a good thing in some respects as it allows a similar compensation scheme and at the same time offers significantly improved quality of sound.Yeah. I did measure the 2421B, 2441, 2445 with Truextent, and the 2446H (before I started swapping diaphragms and cores) with the stock 4344/4345 network last Sunday and none of the 4-inch drivers are a bolt-in without tweaking the network (or using EQ like Bo does).

As far as compensation schemes, I was thinking about trying to put together one of Wayne Parham's real quick as per Ron's email to everyone. I'm not sure I will get to it today though.

First off I would like to point out that the 350hz Edgarhorns are mine and are the ones in my avatar. :D:D
I plan to use a 4-Pi (Wayne Parham) 1.6khz crossover w/JBL 2226J's on the bottom end w/JBL 2446 compression driver equipped with Be diaphrams as soon as I can afford to make that happen!:)Send me some values so I can see what I have around and I'll measure the results if possible before all this has to be returned.
I see with the 2446/Be/Edgarhorn plots that the 1.6 khz region looks to be fairly flat response wise.I'd have to look at the raw data again to confirm that. These are half octave plots without all the warts. They're good for trend analysis. We'd have to look at the rough stuff while actually building any networks.

PS: Did my e-mail with the att/EQ chart help show what Wayne is trying to do with his crossovers?Yes. But give me some values for your application. I don't want to have to figure anything out right now. Just toss me some values.

[eso]

First off I would like to point out that the 350hz Edgarhorns are mine ..

I'm curious about where the Yuichi's came from. There's not many of those floating around and I just sold my old pair to a buyer in Ohio...

Do they look like this:

http://i119.photobucket.com/albums/o157/esodesign/Yuichi%20Crate/IMG_0101.jpg?t=1269187962



eso

[spkrman57]

Using my chart on the e-mail I sent out:

R2 20 ohms
R1 60 ohms
C1 .2 ufd

This is for approx 15 db attenuation/EQ on a 1.6 khz crossover which is the closest I can find for the difference between the JBL 2226 @ 97.4 db/watt and the JBL 2446 @ 112 db/watt if I have the information correct.

The crossover I plan to use is the one I sent you the schematic of. I adjusted the values to reflect the 16 ohm LF and HF drivers I'm using (ie: 2 x coil/resistor values and .5 x cap values).

My application is for passive crossover vs the active bi-amping that Rick, Brad and Dan are going to use.

Thanks again 4313B!

Regards, Ron

[Dingusboy]

I'm curious about where the Yuichi's came from. There's not many of those floating around and I just sold my old pair to a buyer in Ohio...

Do they look like this: (see pic in post 23)

eso

They look exactly like those - They are in fact - Those!!!
Brad

[Mr. Widget]

I'd have to look at the raw data again to confirm that. These are half octave plots without all the warts. They're good for trend analysis.I was curious about the mass break point of the Truextent diaphragms... I assume it was clearly visible in the unfiltered plots? With the mylar surrounds I expect the resonant peak is better damped than it would be with a simple roll suspension.


As others have posted... thanks for slogging through this. It was interesting and helpful. I have done a few projects like this and it is surprisingly more effort than most would expect.


Widget

[eso]

They look exactly like those - They are in fact - Those!!!
Brad

It's good to know they are being well used... I wanted them to go to a good home.

Here's one of my new little experiments for Urei duplex monitors:


http://i119.photobucket.com/albums/o157/esodesign/audio/Ureiyuichi.jpg?t=1269190033

Aren't those just the cutest little radial horns ever?

Brad, do you know if the TrueXtent diaphrams are available for 2425H drivers (8ohm)?

BTW, The plots do show how much nice the Yuichi response is down at 500Hz. The 350 Tracs begin rolling off above that. The trade-off is that the Tracs produce a more 3 dimensional image. The Yuichi's image well left to right, but they do not reproduce the depth of the soundstage as well.

Lots of work in those tests. Thanks.

eso

[Mr. Widget]

BTW, The plots do show how much nice the Yuichi response is down at 500Hz. The 350 Tracs begin rolling off above that. The trade-off is that the Tracs produce a more 3 dimensional image. The Yuichi's image well left to right, but they do not reproduce the depth of the soundstage as well. I've mentioned this before... some consider it a Hi-Fi effect, but I certainly enjoy the effect when well done.

Did you consider making mini tractrix horns for the Ureis?


Widget

[eso]

I've considered it, but the image issue between the Yuichis and the salad bowls is more due to the vanes in those radials, not unlike the effect of multicells.

With the Ureis I'm starting with pretty straightforward reverse engineering before making global changes. The crossovers are designed with correction for the horn and driver. Changing either will mean a major crossover redesign and I'm a wood butcher, not an electronics engineer...


eso

[originaltubino]

I've considered it, but the image issue between the Yuichis and the salad bowls is more due to the vanes in those radials, not unlike the effect of multicells.

I noticed that Mr. Yuichi has published plans for a vaneless or finless 290hz horn, called A-290FL, linked from this page:

http://www.geocities.jp/arai401204/Horn_Page/Horn_Speakers.html

I have the A-290S and love it, but I have heard Cosmos' new tractrix and been impressed with the 3D imaging. I think it's a real retrieval of recorded information -- on good recordings. It sure SEEMS real on live-recorded music, e.g. Mapleshade stuff.

[Mr. Widget]

I've considered it, but the image issue between the Yuichis and the salad bowls is more due to the vanes in those radials, not unlike the effect of multicells. Is this statement based on personal experience, i.e. comparing a pair of radial horns with an without vanes for example... or you read it somewhere... or gut instinct?


...I'm a wood butcher...I've noticed.

Nice butchering. ;)


Widget

[scott fitlin]

Thanks 4313, I really wanted to see measurements and hear what you have to say. I am also more than interested in hearing even more of your listening impressions of the True Extent diaphragms.

This looks promising, and if it is as good as it looks, it, to me, offers a quite reasonable way to upgrade performance.

Thanks for taking the time. :D

[eso]

Is this statement based on personal experience, i.e. comparing a pair of radial horns with an without vanes for example... or you read it somewhere... or gut instinct?

I've noticed.

Nice butchering. ;)


Widget

When I first made the Yuichis I wanted them Primarily to do a side-by-side comparison with the Edgar 350 Tractrix horns. Dr. Edgar and I measured the Yuichis and one of his 350 Tracs with the JBL 2440 drivers I own.

At a So Cal Horn Group meeting we installed the Yuichis in a set of Edgar Titans with the Seismic sub and a whole group of people listened to the system both ways.

The Yuichis are the best sounding Radial design I've heard, and they sound much better off-axis than the tracs do. OTOH the Tracs produce are far more 3-D soundstage but for a much smaller sweet spot. You pick your poison and we all know there's many trade-offs to be made.

I really think the vanes are a big part of the reason, but also the dispersion pattern has something to do with it too. There's only so much information, and if you spread it over a wider area you are going to be sacrificing somewhere to get that extra width. The initial flare outside of the phaseplug in the JBL and Altec drivers something like 12 dregree conic.

I think of it like looking at a foreshortened hand. The sound from the driver is like that hand with the fingers all together. Spread the fingers apart and what happens? Multicells do this to an even larger extent (I still own my very old tar-filled altec 1005A horns and when I was trying to hifi them I had blocked the outer cells to narrow the pattern and got an improvement in imaging for the effort).

Enough. My 2¢


eso

[spkrman57]

Everytime I switch to a other than round horn, I can only be happy for a short time. While I'm seduced by listening to something different, in the long run I live in that 1" sweet-spot" that images like no other horn.

The closest approximation is how headphones give you the 3D imaging effect.

Regards, Ron





The Yuichis are the best sounding Radial design I've heard, and they sound much better off-axis than the tracs do. OTOH the Tracs produce are far more 3-D soundstage but for a much smaller sweet spot. You pick your poison and we all know there's many trade-offs to be made.

[Mr. Widget]

The closest approximation is how headphones give you the 3D imaging effect.Well, kinda... the headphone experience places the musician or band in your head instead of in front of you. I prefer keeping those drug using creative types out in front where I can see them. ;)


Widget

[4313B]

R2 20 ohms
R1 60 ohms
C1 .2 ufd

This is for approx 15 db attenuation/EQ on a 1.6 khz crossover which is the closest I can find for the difference between the JBL 2226 @ 97.4 db/watt and the JBL 2446 @ 112 db/watt if I have the information correct.Here you go:

16 dB added to normalize (grey) the filter plot. If that 3.2 kHz bump sounds objectionable you can try a little notch filter. The 1.8 kHz bump should disappear with your high pass filter.

[4313B]

Aren't those just the cutest little radial horns ever?Yep!

[spkrman57]

It looks like really great response up to 16 khz.

My hearing drops above 14 khz, so I can get away without a tweeter easily.

I'm already working hard to get some 2446 drivers and the Be diaphrams so I can hear what the actual results in my system!

Regards, Ron

[cosmos]

Help me out here a little please.. To me it appears from the above curves that:

Without Network: (nominal output approx 100 db)
1. The 2445/Be goes to about 14KHz before it drops to roughly -5db
2. The 2445/Be goes to about 17KHz before within -8db
3. At 20KHz the 2445/Be is -15 db

With Network: (Nominal output about 84 db)
1. The 2445/Be goes up to 17KHz within -2db
2. the 2445/Be is -7db at 20KHz

So it seems the network works really well at helping HF extension.

Here's my question: The readings and SPL on the curve is not representative of output at 1W, right? However, if there is correlation to the readings pre and post network addition, it appears the network drops overall output about 15db. If so, it'll have a hard time keeping up with a E145 which is about 98-99db or a 2226 which is about 97db/1W/1M, right?

With my active setup, that isn't a concern, but with Ron's passive setup it is..

[Zilch]

2452 is conspicuously MIA here. Not possible?

********

Wayne's selection of HF comp components combines comp AND attenuation to achieve the optimum balance between drivers using a passive crossover:

http://www.audioroundtable.com/PiSpeakers/messages/1278.html

The basic design document is here:

http://www.pispeakers.com/Speaker_Crossover.doc

To understand how it operates, see the plots below of EconoWave using the values shown below. Passively, we combined two of them with L-pads to cover the full range. Actively, it doesn't much matter, but there's purpose in using the minimum workable amount of passive attenuation to accomplish the requisite comp with flea-power amps.... :)

[Dingusboy]

It's good to know they (Eso built Yuichi horns) are being well used... I wanted them to go to a good home.

Aren't those just the cutest little radial horns ever?

Brad, do you know if the TrueXtent diaphrams are available for 2425H drivers (8ohm)?

eso

Eso,

Thanks for the horns - they are getting very good use. I like them A lot. :applaud:

As others have posted, they are cute radials. As others don't know as well as I, your woodworking skills are even more stunning to hold then behold. Great work.

TruExtent does have plans to release diaphragms for JBL 1" drivers but there is no definitive date set for them or any other variant.




I know nothing about these Yuichi horns. They look and measure nice. I'll have to hear a pair someday.

That is a return favor I would be delighted to provide.


I really don't enjoy handling other people's beryllium diaphragms and swapping them around in various cores. Everytime one is exposed I get real nervous.

Understandable - thanks for running them though the paces in spite. It's really appreciated. We simply don't have the appropriate test gear and experience to give them the testing they deserve. Your assistance has been invaluable.


First off I would like to point out that the 350hz Edgarhorns are mine and are the ones in my avatar. :D

If you are going to be that way about it...I'D like to point out the Yuichi's are MINE and not currently in my avatar.

But don't you worry, another couple years of tests and we should be ready to return the 350hz Edgars to you. :D

I (and a few others) have had the good fortune to have heard the TruExtents (on 2445's) on both aforementioned horns and in very unscientific comparison to TAD TD-4001's. The TAD's are a great sounding compression driver. The TruExtent 2245 is as well. I drew no definitive conclusions aside from knowing both outperform any other compressions driver I have heard. I've never heard the 476be. Nor have I heard TAD's newer offerings.

As far as the horns go, my ears seemingly confirm the comments by Ron & Eso regarding the horns - the farther the listener is from the horns the less intense these differences. I could live happily with either but prefer the Yuichi design - the dispersion sounds more natural to my ears. I've picked the other poison.

[4313B]

Help me out here a little please.. To me it appears from the above curves that:

Without Network: (nominal output approx 100 db)
1. The 2445/Be goes to about 14KHz before it drops to roughly -5db
2. The 2445/Be goes to about 17KHz before within -8db
3. At 20KHz the 2445/Be is -15 db

With Network: (Nominal output about 84 db)
1. The 2445/Be goes up to 17KHz within -2db
2. the 2445/Be is -7db at 20KHz

So it seems the network works really well at helping HF extension.Yes, these are the same compensation type of filters we've been discussing since people started showing interest in the 4430 back when the forum first started.


Here's my question: The readings and SPL on the curve is not representative of output at 1W, right?No. 1W input is hella loud.
However, if there is correlation to the readings pre and post network addition, it appears the network drops overall output about 15db.Right. 15 dB by math, 16 dB by measurement.
If so, it'll have a hard time keeping up with a E145 which is about 98-99db or a 2226 which is about 97db/1W/1M, right?Probably not, the Edgar horn is quite efficient as shown in #2 (http://audioheritage.org/vbulletin/showpost.php?p=283237&postcount=2) where the input levels are equal. An actual network is worked up as a system where all the drivers are measured in their intended mounted positions. The actual corrected attenuation would be determined at that time.

With my active setup, that isn't a concern, but with Ron's passive setup it is..It depends on how you are going to go active. If you use something like a DEQX then no it wouldn't matter. If you are just going with a simple active filter without any EQ then you will need some kind of passive compensation. Not a 16 dB drop in your case, something more along the lines of a 6 dB drop.

Bascially you are looking at the response to see how much you have to work with and in this case "flat to 16 kHz" appears to be a reasonable goal. Since the horn/c.d. combination is ~ 6 dB down at ~ 16 kHz you know that you need to do a broadband reduction of about 6 dB to get that "flat" response to ~ 16 kHz.

2452 is conspicuously MIA here. Not possible?Not really necessary and it is a 1.5" exit compression driver. Additionally it has an MSRP of $1,430 and one would still be stuck with a titanium diaphragm. JBL Pro kind of missed the boat. They got on the boat that was moving all of manufacturing south instead...

Wayne's selection of HF comp components combines comp AND attenuation to achieve the optimum balance between drivers using a passive crossover:

http://www.audioroundtable.com/PiSpeakers/messages/1278.html

The basic design document is here:

http://www.pispeakers.com/Speaker_Crossover.doc

Thanks for posting those links. Ron sent them to cosmos and a bunch of other people last Friday via email but that doesn't do anyone any good reading the forum. :p

[Wilhelm P.]

4313B:

Thanks for sharing Greg's raw frequency response measurement of the 476Be on the Everest horn.

Could we also have the raw impedance response measurement?

[cosmos]

Probably not, the Edgar horn is quite efficient as shown in #2 (http://audioheritage.org/vbulletin/showpost.php?p=283237&postcount=2) where the input levels are equal. An actual network is worked up as a system where all the drivers are measured in their intended mounted positions. The actual corrected attenuation would be determined at that time.It depends on how you are going to go active. If you use something like a DEQX then no it wouldn't matter. If you are just going with a simple active filter without any EQ then you will need some kind of passive compensation. Not a 16 dB drop in your case, something more along the lines of a 6 dB drop.
Thank you. That's exactly what I thought. Currently, I am using a Behringer DCX2496, but intend to go to a fixed tube 18db crossover (Steve Bench design) shortly. However, my intention was to add a passive EQ, like discussed, to either system once I have some idea how much compensation is needed.

Of interest, is the change not only in output, but in slope on the yuichi in relation to the other horns. When we listened to both Yuichi horns (Ricks' A290S and Brads' A290) and Ron's Edgars, everyone there perceived lower HF output with Brads' (the ones you tested). While the two Yuichi types looked very much the same, there are some minor differences. Therefore, it's possible that the curve shown in #2 (http://audioheritage.org/vbulletin/showpost.php?p=283237&postcount=2) is not representative of all yuichi horns. I'd be willing to bet Ricks' A290S have a flatter curve in respect to HF extension. Of course, this is something that should be able to be remedied with the right network approach.


Bascially you are looking at the response to see how much you have to work with and in this case "flat to 16 kHz" appears to be a reasonable goal. Since the horn/c.d. combination is ~ 6 dB down at ~ 16 kHz you know that you need to do a broadband reduction of about 6 dB to get that "flat" response to ~ 16 kHz.

No doubt, based upon looking at those curves, the horns I built will be different and will require a different circuit for compensation. Just like I suspect Ricks and Brads horns require different compensation values.

[4313B]

4313B:

Thanks for sharing Greg's raw frequency response measurement of the 476Be on the Everest horn.

Could we also have the raw impedance response measurement?Sure... but... I lied, that wasn't the Everest II horn, it was the 1400 Array Horn. :D

This is for the 1200FE and 1400 Array Horn with 476Be system. Call it the 1200 Array II if you want. :)

The Everest II data is the second two plots. Raw data and data with actual network.

I changed the scales to match the other plots in this thread. Also note that the Everest II raw data is higher resolution.

[cosmos]

http://i119.photobucket.com/albums/o157/esodesign/audio/Ureiyuichi.jpg?t=1269190033

Aren't those just the cutest little radial horns ever?


Very nice looking horns. What are the two notches on the top and bottom for?

[yggdrasil]

Those 476Be curves are looking insanely good!

[4313B]

Those 476Be curves are looking insanely good!They sound alot better than they look but you should see a waveguide/476Be curve. :p

JBL has stated that the only way they might be able to make the 476Be any better would be to go to a five slot phase plug but the cost/performance probably wouldn't be justifiable. The JBL curves were done in the anechoic chamber with the usual 2.83 volts input. That kind of input is deafening outside a chamber.

[eso]

Very nice looking horns. What are the two notches on the top and bottom for?

According the Steve Schell, who is a walking encyclopedia of things Lansing, JBL, Altec and RCA among others, the designer of these was never fully satisfied with anything and continually sought refinement. The slots were introduced in the "A" series, I believe, as a sort of mechanical notch filter to alleviate some artifact in their sound.

As the slots remained in the horns through the final "C" series I suspect they have a positive effect. I went to pretty good lengths to make sure they were as identical in size, shape and location as possible so my wooden versions will have comparable performance to the originals.

eso

[4313B]

I was curious about the mass break point of the Truextent diaphragms... I assume it was clearly visible in the unfiltered plots?I'm just not seeing it. I guess we'd need a PWT plot. Jerry calculated the breakpoint of one 476Be at 4733 Hz this morning. I would imagine it is between 3 kHz and 4.5 kHz depending on the core. :dont-know

[doucanoe]

Wow, very interesting stuff guys! Thanks for taking the time to do the analysis, 4313B :)

I guess this confirms what I find so involving about my own 350Hz Edgars. Also, that I really need to hear the Yuichi horn and the TruExtent diaphragm. Unfortunately, thats probably going to require a trip east for me.

RC

[Mr. Widget]

I'm just not seeing it. I guess we'd need a PWT plot. Jerry calculated the breakpoint of one 476Be at 4733 Hz this morning. I would imagine it is between 3 kHz and 4.5 kHz depending on the core. :dont-knowYou know, I mis-posted... what I was looking for was a secondary resonance peak. But, I can't see that from your plots either. :(

Yes, for the mass break point to be clearly defined we would need a plane wave tube... these horns all add too much coloration to really see what is going on. As good old Al Klappenberger posted back during Project May, without a PWT, you really don't know what you have. :banghead:


Widget

[cosmos]

I guess this confirms what I find so involving about my own 350Hz Edgars. Also, that I really need to hear the Yuichi horn and the TruExtent diaphragm. Unfortunately, thats probably going to require a trip east for me.

RC
If anyone is interested, both will be playing at AKFest 2010 in Novi, MI on May 1st and May2nd. In addition, in the listening room I am Co-Hosting, I will have my new "Cosmos" horns, which are really good. :D AKFest is a really neat gathering of new and vintage gear and down to earth, relaxed atmosphere.. Frankly, it's a blast!

[Ian Mackenzie]

Details of crossover network as requested.

The actual resister values are quite flexible to tailor the response and this is only an example.

The topology is what is important.

Response image is the raw driver and horn.

The actual values may need to be modified as there are differences in the throat and flare geometry of the Yuichi horn.

This won't be up here indefinately.

[doucanoe]

If anyone is interested, both will be playing at AKFest 2010 in Novi, MI on May 1st and May2nd. In addition, in the listening room I am Co-Hosting, I will have my new "Cosmos" horns, which are really good. :D AKFest is a really neat gathering of new and vintage gear and down to earth, relaxed atmosphere.. Frankly, it's a blast!


Yes, I know. I listened to your jbl ewaves last year in your room. I think that you may have visited ours also. I was rooming with Retro Stereo (Greg). Your new horn creations are high on my "wish to hear" list also!

I apparently don't make long lasting first impressions, Dan ;)


RC

[4313B]

Details of crossover network as requested.Thank you!

[4313B]

You know, I mis-posted... what I was looking for was a secondary resonance peak. But, I can't see that from your plots either. :(You can see one in #7 (http://audioheritage.org/vbulletin/showpost.php?p=283242&postcount=7) which is right where JBL said I would find it, around 18 kHz. The 476Be occurs at around 21 kHz.
Yes, for the mass break point to be clearly defined we would need a plane wave tube... these horns all add too much coloration to really see what is going on.Yep, it's easier to just figure it out mathematically if you know the Bl factor and the moving mass.
As good old Al Klappenberger posted back during Project May, without a PWT, you really don't know what you have. :banghead:Yeah but do we really care since we aren't designing drivers? We're playing with the cards delt us.

[cosmos]

Yes, I know. I listened to your jbl ewaves last year in your room. I think that you may have visited ours also. I was rooming with Retro Stereo (Greg). Your new horn creations are high on my "wish to hear" list also!
Ron,

It sounded like you were not planning on coming to AKfest. I was attempting to entice you a little.. :)

For me, last year was more about just having fun at the Fest and less about higher end, as it was in the years that I brought KLH Nines, JBL 250Ti-LE and Yamaha NS1000. This year, however, I am teaming up with Kegger in a room. He's providing the chrome top stuff and I am providing the speakers. They'll be the new horns with 2445/TrueXtents and either E-145 or Lambda TD-15 woofers in a 2401 style cabinet. There may be a surprise or two.. Time will tell.. I always try to get more ready than I am able. ;)

I think it'll sound better than the 4648/E-Wave combo.

[cosmos]

Details of crossover network as requested.

The actual resister values are quite flexible to tailor the response and this is only an example.

The topology is what is important.

Response image is the raw driver and horn.

The actual values may need to be modified as there are differences in the throat and flare geometry of the Yuichi horn.

This won't be up here indefinately.

Thank you very much, Ian! :applaud:

I almost missed it.. (Whew)

[doucanoe]

Ron,

It sounded like you were not planning on coming to AKfest. I was attempting to entice you a little.. :)

For me, last year was more about just having fun at the Fest and less about higher end, as it was in the years that I brought KLH Nines, JBL 250Ti-LE and Yamaha NS1000. This year, however, I am teaming up with Kegger in a room. He's providing the chrome top stuff and I am providing the speakers. They'll be the new horns with 2445/TrueXtents and either E-145 or Lambda TD-15 woofers in a 2401 style cabinet. There may be a surprise or two.. Time will tell.. I always try to get more ready than I am able. ;)

I think it'll sound better than the 4648/E-Wave combo.


I really would like to be able to make it again this year, Dan, but unfortunately it's not in the cards for me. Greg has been doing his best to "entice" me also :D

Always something interesting going on in your room. With the teaming up with Kegger and his tubeware, I'm sure this year will be no exception.

RC

[4313B]

The TAD schematic Ian provided with the data I measured:

[spkrman57]

I will have my new "Cosmos" horns, which are really good.

Dan,

Pics of the "Cosmic" Tractix horns are needed to be added to this thread immediately!!!;)

Ron:D

[Dingusboy]

The TAD schematic Ian provided with the data I measured:

Thank you very much for posting this information.

Thanks also to Ian for providing the time clocked schematic.

[cosmos]

Dan,

Pics of the "Cosmic" Tractix horns are needed to be added to this thread immediately!!!;)

Ron:D

For reference, here they are resting on 4507 cabs with E145 and zilch plugs..

Made from compressed cosmic dust (Poplar) and a right handed, smoke shifting calavicator (secret hammer):

[Zilch]

They sound alot better than they look but you should see a waveguide/476Be curve. :pLet's SEE it! :)

[And how a 2452H with TruExtent diaphragm performs on the same waveguide.... :thmbsup: ]

[Ian Mackenzie]

Huh,

Unless its dead flat its useless!! Throw it out.

(the 2446 is obviously different to the 4001)

Might need some optimisation.

[4313B]

Let's SEE it! :)

[And how a 2452H with TruExtent diaphragm performs on the same waveguide.... :thmbsup: ]I sold all my waveguides to guys that couldn't buy them from JBL Pro directly (overseas). And I don't have any 2452H's anymore either.
Huh,

Unless its dead flat its useless!! Throw it out.

(the 2446 is obviously different to the 4001)

Might need some optimisation.I thought it looked quite good. It's perfect for someone with a 3 dB peak in their hearing at 2.2 kHz! :p And yes, it can probably be optimized for the 2446.

[Mr. Widget]

I thought it looked quite good. It's perfect for someone with a 3 dB peak in their hearing at 2.2 kHz! :p And yes, it can probably be optimzed for the 2446.Interestingly Tom Holman designs just such a dip into his Audyssey software. :D

What impressed the hell out of me was that impedance plot.


Widget

[4313B]

What impressed the hell out of me was that impedance plot.Yep.

[kegger]

With the other compensation circuit I'll be curious if we feel the need for a little extra extension
at the "extreme" top, regardless of ones hearing ability out there, that extra there can really add
to frequencies one can hear, little bit of air and or shimmer could be wanted in those upper octaves.

I'll get to hear it soon enough and find out if it seems lacking..:)

[Zilch]

I sold all my waveguides to guys that couldn't buy them from JBL Pro directly (overseas). And I don't have any 2452H's anymore either.I have all that stuff, of course, and there's a 476Be local here.

All "we" need is some TruExtent (which is also local) diaphragms.

[I will inquire.... ;) ]

[Ian Mackenzie]

Interestingly Tom Holman designs just such a dip into his Audyssey software. :D

What impressed the hell out of me was that impedance plot.


Widget


Agreed,

Its far better damped than a the simple series LCR network commonly used like that in the 4344 and some PA Cd crossovers.

[4313B]

Let's SEE it! :)

[And how a 2452H with TruExtent diaphragm performs on the same waveguide.... :thmbsup: ]

I have all that stuff, of course, and there's a 476Be local here.

All "we" need is some TruExtent (which is also local) diaphragms.

[I will inquire.... ;) ]I finally got around to digging this old post up: #27 (http://audioheritage.org/vbulletin/showpost.php?p=217849&postcount=27)

Why are you still interested in these waveguides after I bent over backwards to get you a pair of H4338's? The H4338 is the best horn JBL has produced to date. I've not seen your project employing them. I wish I had the skills necessary to build a pair of wood H4338's.

[timc]

I wish I had the skills necessary to build a pair of wood H4338's.


Maybe you could bribe JohnW to make you a pair ;)

[4313B]

Maybe you could bribe JohnW to make you a pair ;)There's no way that I know of to get a model of one.

[timc]

Ah...I thought you just lacked the woodworking skills.

But don't you already have a pair of original 4338's? Can't you make a casting mold with them as a base?

[4313B]

Ah...I thought you just lacked the woodworking skills.

But don't you already have a pair of original 4338's? Can't you make a casting mold with them as a base?I wouldn't risk it. I'm currently trying to come up with a decent grille design for my H4338's so they are protected from kids.


The only two tasks that I have left in this thread (that I am aware of) are measuring Dan's horn and measuring the 2446 with the Truextent diaphragm on the 2311 or H93 horn with the 4355 network.

[originaltubino]

Belated thanks for the great new info posted here, such as the FR of the TAD 4001 with EQ. Since my living room system includes the 1601a and TD-4001 on TAD clone horns, this greatly interests me!


The H4338 is the best horn JBL has produced to date.

I would very much like to get my hands on a pair of those horns. Maybe the combined talents of the Toledo group could eventually produce them in wood. But to be honest, I would have to hear sonic advantages over the radial wood horns before diving into that project...

[4313B]

But to be honest, I would have to hear sonic advantages over the radial wood horns before diving into that project...Yep. They aren't for everyone.

The H4338 and 1400 Array/SAM1HF horns are very similar in performance. The wonderful thing about the SAM1HF is that one can actually buy it here in the U.S. I don't know what the hell JBL was thinking making it available here, maybe someone goofed. :rotfl:

[eso]

... I wish I had the skills necessary to build a pair of wood H4338's.

What do those look like? Any pictures handy?


I might be able to help you out with that. Exotic Veneers are extra.



eso

[4313B]

Here's a photo I found of them mounted in some enclosures.

[eso]

I could do that. I would either need some good drawings/dimensions/radii or a physical example to reverse engineer.



eso

[Dingusboy]

I could do that. I would either need some good drawings/dimensions/radii or a physical example to reverse engineer.



eso

Perhaps something like this Laser scanner (http://www.david-laserscanner.com/) could get the job done with little investment and even less worry.

[originaltubino]

Perhaps something like this Laser scanner (http://www.david-laserscanner.com/) could get the job done with little investment and even less worry.

Or one of those doohickeys like a superfine comb where the steel-pin teeth push in to make a copy of any surface curve. That photo makes it look do-able, but is the throat transition something especially complex?

[eso]

Or one of those doohickeys like a superfine comb where the steel-pin teeth push in to make a copy of any surface curve. That photo makes it look do-able, but is the throat transition something especially complex?

No doubt there is some complexity within the throat. Notice there is a slot that is narrower than the throat of the driver. I would guess then that the sides are expanding faster than the top and bottom coming together while transitioning from round to ~rectangular. That is one of the difficult areas, the other being getting a nice veneer pattern on the compound curves of the mouth.

The rest is fairly easy.


eso

[Zilch]

I finally got around to digging this old post up: #27 (http://audioheritage.org/vbulletin/showpost.php?p=217849&postcount=27)Yes, I had that in my file here, but wasn't about to repost it.


Why are you still interested in these waveguides after I bent over backwards to get you a pair of H4338's? The H4338 is the best horn JBL has produced to date. I've not seen your project employing them. I wish I had the skills necessary to build a pair of wood H4338's.It's not about me, and it never was; I only ever showed what others could do, that they might, if they wanted to.

We are very much alike in this respect, and it has worked out well. I hold you personally responsible for this.... :p

[Chas]

We are very much alike in this respect, and it has worked out well. I hold you personally responsible for this.... :p


:yes::yes::yes::yes:

[JoMoCo]

Yeah, lets see that diaphragm performance measured at Zilchlabs...maybe with the CLIO wavelets????.....would be nice to see it both on a 90x40 wave guide and a 2380 horn w/the 2445?....:hmm:

[4313B]

Yeah, lets see that diaphragm performance measured at Zilchlabs...maybe with the CLIO wavelets????.....would be nice to see it both on a 90x40 wave guide and a 2380 horn w/the 2445?....:hmm:Measure the whole fuckin' lot at "Zilchlabs"...

Moderators, please whack this thread.

[Mr. Widget]

Measure the whole fuckin' lot at "Zilchlabs"...

Moderators, please whack this thread.I thought that request was quite peculiar myself... I don't think it was meant to trivialize your work, I think it was posted out of ignorance. The assumption being, it would be interesting to see what these diaphragms do in a more familiar driver and affordable horn.

I think most of us who work with a variety of horns and drivers can make great use of the information posted on this thread... unfortunately not everyone gets it.



Widget

[JoMoCo]

I thought that request was quite peculiar myself... I don't think it was meant to trivialize your work, I think it was posted out of ignorance. The assumption being, it would be interesting to see what these diaphragms do in a more familiar driver and affordable horn.

I think most of us who work with a variety of horns and drivers can make great use of the information posted on this thread... unfortunately not everyone gets it.


Widget

What is another measurement on CLIO a problem? If it is I fail to see it. If for some reason it was taken to "trivialize" the research and hard work already posted, that is not true and WAY off base...

[Mr. Widget]

What is another measurement on CLIO a problem? If it is I fail to see it.
As for CLIO, the information posted into LEAP was gathered using CLIO, and very likely more accurately so. The amount of clutter at Zilchlabs makes me a bit leery of treating any of the posted measurements from Zilchlabs as "laboratory" measurements.

If for some reason it was taken to "trivialize" the research and hard work already posted, that is not true and WAY off base...I didn't think you were meaning to be as rude as I originally took your post and as 4313B obviously took it.

The reason it seems rude is it feels like you are questioning the validity of the original work.


Widget

[timc]

Question: Why should CLIO give any different results?

I have measured with TrueRTA, CLIO, HolmImpulse, WinMLS, NTI system, and the Copland DRC.

They all measure the same, so why?


I applaud 4313B's work on this, and I'm tempted to try a combination of 2446/Truextent after seeing the measurements.

[originaltubino]

I think most of us who work with a variety of horns and drivers can make great use of the information posted on this thread...

Hearty agreement. That's not changed by any subsequent posts.

[timc]

As for CLIO, the information posted into LEAP was gathered using CLIO


Hehe...we posted at the same time.

[JoMoCo]

Please notice the key words in my post are wavelet analysis, which to my understanding is a newer graphical tool not all are aware of or use & requires CLIO measurement 8.5 or above. and two other horns not included in the research already posted.

Pretty simple really. Personal negative interpretations past that? not stated or intended or mine:dont-know

[cosmos]

Personally, I feel troubled that this discussion and the work invested by 4313B has somehow seemingly come under fire. All of the information presented is valuable and it took considerable effort to achieve. Once again, I thank 4313B for these efforts.

There is great value to achieving test results for the purpose of developing filters to best release the capabilities of any great driver/diaphragm/horn combination. Unfortunately, I know precious little about filter design. Among what little I do know, is the belief that any change to any one of the three can cause significant changes to the subsequent response. This is the reason I asked 4313B to test my newly developed horns with my drivers and my diaphragms. Once this is obtained, there is great hope to be able to put the information to use in developing a filter to release it's potential. Frankly, I am stoked. The horns would be in 4313B's posession at this moment, if I hadn't decided to clean up the throats before sending them to him.

Further, I own the best version of TrueRTA and bought it in hopes of testing systems. However, to date, I haven't been able to trust the readings I get from it. I do, however, trust 4313B can achieve the results I cannot. Once again, thank you 4313B. :applaud:

That said, it's easy to understand why everyone would like the combination in their system tested. Is there anyone here reading this that wouldn't? (unless they are already ahead of that)

[Mr. Widget]

Please notice the key words in my post are wavelet analysis, which to my understanding is a newer graphical tool not all are aware of or use & requires CLIO measurement 8.5 or above. and two other horns not included in the research already posted.

Pretty simple really. Personal negative interpretations past that? not stated or intended or mine:dont-knowI acknowledged that you were interested in the other horns... as for the wavelet analysis, in my mind those colorful plots were added to sell new CLIO licenses and are not terribly useful... the reason why I haven't bothered to upgrade my CLIO software and why I didn't acknowledge it in your original post.

If you like those plots and are curious about how those specific horns might perform, why not PM Zilch and ask him to buy a TruExtent diaphragm and post away. He doesn't seem to mind buying all manner of bits and pieces to test.

I am a little surprised that you can't see how your post could have been misunderstood, but I think we all know where everyone is coming from at this point. Thanks for graciously clarifying your post.


Widget

[JuniorJBL]

I would like to thank 4313B for the measurements here. It will serve reference for me later on.

I know that he puts great effort into making sure that they are as spot on as he can get them givin there isn't any anoic chamber to further make the measurements better than they are.

Not often do people have LEAP to do that, let alone post them here for all to use.;)

EDIT: I see that I have repeated others but still stand by my statement!:applaud:

[spkrman57]

I thank 4313B for what he has accomplished and shared with us!:applaud:

I for one will be able to use this data for my own personal system in the very near future.:)

It's a shame the thread is going this direction as there is a wealth of information contained within it.:(

Regards, Ron




The only two tasks that I have left in this thread (that I am aware of) are measuring Dan's horn and measuring the 2446 with the Truextent diaphragm on the 2311 or H93 horn with the 4355 network.

[Doc Mark]

Evening, All,

I, too, would like to thank 4313B for undertaking the task, and for sharing his results with the rest of us. Were it not for several LH members, like him, we would all be the poorer, most certainly!! Even though I am not undertaking any projects right now, I still have high hopes for quite a few such things in the future. Without guidance, education, and information, from our JBL Elders, like 4313B, many of us would be lost in the dark, without a clue. I think that his contributions, and those of the others who share such valuable testing results and information, make LH the special place that it is, and I very much appreciate that!! Thanks, very much, Sir, and please continue to shine the light of JBL truth on those project you undertake! Take care, and God Bless!

Every Good Wish,
Doc

[4313B]

It's a shame the thread is going this direction as there is a wealth of information contained within it.:(I told you we needed to go private with it and I should have followed through... we might have actually accomplished everything to good end.

In any case, cosmos doesn't need his horn measured now and I really couldn't care less about the Be diaphragms in the 4345's or 4355's so I'm returning the horns and drivers this morning.


Measurements talk....all else may be odiferous...Thanks for sharing your ignorance in my thread... :barf: I addressed the point of the thread in the very first post because I knew some idiot would show up posting something like this...

[1audiohack]

Jumping back a bit, if some one has a loose H4338, I could 3D model it with our Faro Arm. A completely non damaging light contact proccess.

We could probably rough them in our four axis mill, never cut wood in one of them, yet.

[Earl K]

I really hate to ask this ; but, how about locking up this thread now ?

- It started to go south at post 40 when the grand-poobah of the Econwavers figured this thread somehow should be about them .
- JOMOCO merely walked through the door that Zilch opened for him .

- I can't think of anything that would rankle Giskard more than a telegraphed attitude ( that he too should be happy to act as a sales tool for P.E. and eBay ) .

- My hats off to Giskard for hanging in as long as he did .

<> Earl K

[scott fitlin]

OK, Earl, and 4313, your request has been acknowledged, and it's done. The thread is closed.

Unless one of the administrators feels differently about this, it will remain closed. It does seem as if the original intention of measuring the be diaphragms from TrueExtent has been satisfied.