The 4345 2010

[Ian Mackenzie]

Funny title for a Thread ..Probably how long it will take to complete!

Read On

After what, about 4 years of enjoyment I have decided its time to give the O'l 4345 a make over.

Background:
Following numerous ideas of modifications, new baffle arrangements and updated crossovers I decide it made more sense to offer the key drivers to a caring 4345 member and then re organise the system to my own recipe. The idea of using a larger horn and driver with lower end bandwidth did not seem to want to work as well as the stock system

The 2245H stays, the 2122H goes as does the horn and slot.

Why do it? I think much of today's digital hi resolution music calls for a system with a lot more detail and better imaging. The 4345 goes loud and is dynamic but lacks the accuracy and resolution compared to todays contemorary systems.

Some people may still wish to use a Tad driver on the original exponential horn but I think the horn is also a major limiting factor.

Driver Choice:
There are still benefits of having a strong extended LF woofer to provide a 32 hertz low end bandwith. You really cannot do that properly with a hi sensitivity 2 way system .

The plan is to run a suitable 12 or 15 inch mid/bass driver down to around 80 hertz and run that up to 900-1200 hertz. Possibly a GPA 12-15 inch mid/ woofer or an AES 15M with low distortion motor. Both have excellent mid/bass characteristics and crossover flexibility to the horn.

While I like the bass mid transition of the stock system I think this will work better as will the 12-15 inch transition to the larger horn.

System benefits:

This system should sound more open, image properly, have better tonal character and have blazing resolution. Of course the bass we have all grown to love will still be there. In a sense the system is an augmented 2 way. The true mid / bass driver and the horn will cover 7 octaves and the full vocal range while the 2245 will support the bass & deep lows and a ribbon will provide ambience above 14000 hertz. By carefully setting up the mid/bass and the horn the system will project a very coherent sound.

Moving the crossover points around makes the 2245 more of a true sub and the 12-15 driver more of a true mid bass. The same applies to the horn.

Horns:
I would like to use one of these horn.

http://www.geocities.jp/arai401204/Horn/A480FL/A480FL.html

or

http://www.geocities.jp/arai401204/Horn/A1K/A1K.html

Driver for the horn will be a Tad 2001 or a Neo Eightensound driver.

Above the horn I plan to try a small Ribbon super tweeter.

Its still a 4 way with many of the hallmarks of the original stock system.

I think when it is done it will be perhaps be more beautiful than the stock system with some resemblence to a Westlake set-up

Below are some images of the horns and response curves.

They are quite smooth without any equalisation and have good controlled coverage.

The plan is to make these smaller horns as a step up to the larger horn for my other Tad Fat Boy project.

The system will be biamped with provision for triamping.

I dont think it will be that expensive to update and drivers will be purchase as I progress.

What have I done with the old parts?

Some have been ear marked for a caring 4345 owner.

Ian

[Joe Alesi]

The plan is to run a suitable 12 or 15 inch mid/bass driver down to around 80 hertz and run that up to 900-1200 hertz. Possibly a GPA 12-15 inch mid/ woofer or an AES 15M with low distortion motor. Both have excellent mid/bass characteristics and crossover flexibility to the horn.

Hello Ian,

Nothing says you have to keep it "all in the family", but midbass thwack is also very good from the 2226 and possibly, although I've not tried it, the 2206.

Best
JA

[Ian Mackenzie]

Hi Joe,

Yeah those woofer were a starting point.

The 2020 is also a maybe if I am prepared to be flexible. Same goes for the 1200FE

The newer breed of drivers are miles better than the old vintage models, they are of much lower distortion with modern engineering techniques.

The GPA Altec drivers in some circles are considered the Hier to the Tad monarchy. The AES 15M is a drop in replacement for the Tad in some minds.


Ian

[Ian Mackenzie]

The point of no return:eek: No backing out now

Well all the surplus drivers are spoken for.

I had a brain wave a moment ago.

Looking at the tad 1601a it is possible to run it as a sealed system.

This would simplify my project for driver search as i could also use the 1601a in a vented system fot a high sensitivity 2 way. It would also make voicing more consistent and perhaps assist in sorting out the crossover network.

The tad 1601a is good up to 900 hertz based on a other users. My criteria for the mid bass bandwith is therefore met in addition to the above benefits.

Using Bass Box my inital sims indicated in a sealed box of about 1.5 cu fts it will have an fc of 75 herts for a QT =0.707 , the F3 is also 75 hertz and output is around 128 db @75 hertz and 115 dn at 35 hertz

I may have mess with the internal volume of the exiting enclosure for the trial version to get the mid bass volume or just make a stack using my spare 4645 sub enclosure and rig a trail system that way to see how it goes.


(For those not keen on spending up big the 2206 is a maybe for the mid bass)

(A some what less adventurous horn option is one of the DDS fibreglass horns that can be ordered in 1, 1.5 and 2 inch throat sizes)

The green curve is the sealed box
The red and the curves are reflex boxes.

Ian

[Dr.db]

I know this thread is 10 years old, but have you actually tried the Yuichi A-480 horn with the TAD TD-2001?

[Ian Mackenzie]

The short answer is NO.

It’s was an idea only.

There were several issues that made implementation less than satisfactory.

1. Incorporating the larger mid enclosure and horn made system integration sub optimal.
The problem being it wound become a PA stack with the horn no where near optimal location for seated listening.

2. At the time acquired a Tad 4001 which l think is better overall than the Tad 2001 but l was never happy with its performance. The system required a uhf which wasn’t the path l wanted to take due to the location required for alignment of the drivers.

If you have ever listened to a larger system with a vertical stack of drivers in a 2-3 metre listening distance it’s not an audiophile experience. At best it’s a wall of sound. Some people are happy with that and each to their own.

In the ensuing years l have improved the resolving power of stock 4345 through a deeper understanding of driver and crossover operational characteristics and subtle refinements including a more precise approach to system setep and voicing with associated amplification and signal path.

There is nothing wrong with trying the above idea if you have the time but l know where it will end up.

[Dr.db]

Hi Ian,

thanks a lot for your detailed answer.


I was actually thinking about just swapping the midhorn and middriver and leave the residual parts of the 4345 as is.
Besides that, I was thinking about a different baffle-layout. A strictly vertical alignment, more like the westlake tower speakers.

From top to bottom:
TAD 2001 + yuichi A480
JBL 2405
JBL 2122
JBL 2245

With this alignment all speakers can be positioned closely together. The Yuichi A480 protrudes way less than something like a 2397, so the 2405 and 2122 can be positioned closely without causing nasty reflections.
Besides that, the yuichi A480 should offer a more controlled dispersion than the 2308 acoustical lens, or am I wrong?
The loading should be better as well, as the hornmouth is bigger as well.

The TAD 2001 is said to sound less exhausting than the JBL 2425/6 driver. But I´m not aware if anyone has ever adjusted the crossover for this driver.


Regards,
Olaf

[Ian Mackenzie]

Hi Olaf

You need to build it and form your own opinion

The crossover is the barrier for most diy loudspeaker plans.

You need to acquire test and measurements equipment to make such an effort worthwhile.

[Mannermusic]

Hi Olaf

You need to build it and form your own opinion

The crossover is the barrier for most diy loudspeaker plans.

You need to acquire test and measurements equipment to make such an effort worthwhile.


Yes to all the above. My experience, for what it is worth, is to build a "4333" enclosure for the bass drivers and then experiment with (endless) treble ideas above. It puts everything in a reasonable position and allows for relatively easy revision. An on-going experiment. Otherwise, you never learn anything. And, as you say, you've got to learn to do crossovers. Not easy because we don't have sufficient tools - read anechoic environment - to optimize theory . . . etc. (big on the etc.). But, you've gotta get your hands dirty. Sometimes (most of the time) I think the Internet environment has actually damaged folks ability to improvise, create. So it goes.

[Ian Mackenzie]

+1

I understand you wanted to know if anyone else tried this combination.

I don’t recall but other members have anecdotally tried placing horns on top of the enclosure. Another member has used the Heil driver. One member used an oblate spheroid horn turned from wood. In each case this did some experimentation with the crossover. I would point out that the crossover either makes or breaks the whole project and it’s a case of trial and error even after computer simulations to get it to sound right.

It might take 12 months or longer to do all the work.
Ian

[speakerdave]

TAD 8 ohm small format drivers have been successfully substituted behind the 2307/2308 in factory 4345's using the stock crossover. Because the stock drivers are 16 ohm the parties doing this have been pleasantly surprised at the lack of audible response anomalies. The speculation has been that the use of tapped inductors in the stock crossover is the explanation. Since the redrawn GT/4313b crossover does not use them, similar success with diy 4345's may be more difficult to obtain because crossover tweaking would be required. Ian's caution is worth noting.

I also experimented with 16 ohm dusted diaphragms in 2450's on 2311's. They fit into the factory cabinet because of the ways the 2450 body shape differs from the 2441. That too was a big improvement over the stock setup, though it is true that in the end I settled on the TAD beryllium. Both of my alterations may have resulted in very modest response tilt and sensitivity changes, I don't know. Ian has said instrument-based setup of this speaker is required for optimum performance. Lacking that, I settled for using a lucky dime to adjust the L-pads. :)

[Earl K]

TAD 8 ohm small format drivers have been successfully substituted behind the 2307/2308 in factory 4345's using the stock crossover. Because the stock drivers are 16 ohm the parties doing this have been pleasantly surprised at the lack of audible response anomalies. The speculation has been that the use of tapped inductors in the stock crossover is the explanation. Since the redrawn GT/4313b crossover does not use them, similar success with diy 4345's may be more difficult to obtain because crossover tweaking would be required. Ian's caution is worth noting.

<<<<SNIP>>>> :)

Its' mostly a happy coincidence that things worked out ( IMO ).

8 Ohm TAD small format drivers and ( so-called ) 16 ohm JBL small format drivers have impedance curves within a couple ohms of each other.

Once either driver is located behind a ( sufficiently attenuated ) variable Lpad ( 8 or 16 ohm ) the actual load impedance becomes dominated by that LPad.

That's not to say that any DIY rookies reading this thread shouldn't get the appropriate test gear and start learning how to properly design speakers.

:)

[speakerdave]

. . . . any DIY rookies reading this . . .

You mean like dangerous amateurs? :p

[Ian Mackenzie]

Well put all.

I found the dusted 2425J diaphragm also an interesting option within the envelope of a carefully adjusted HF and UHF filter networks using precision test equipment and numerous measurements. Then a bunch of simulations of the
modified network to precisely match the stock response. Only then was l happy with it.

But without careful consideration to these adjustments (of the system overall) any plug and play comparisons cannot be taken as cut and dried subjectively. Carefully set up the 4345 design with the stock drivers in as new working order is impressive.

The Tad2002 in the stock network is a lucky call.

[Dr.db]

Thanks at all for the great input!

I understand, that you need test equipment to do a precise diy rebuild of the crossover.
As Ian pointed out, I was hoping that someone more talented than me had allready done this in the past.
The 8ohm TAD-2001 seems to work with the stock crossover as speakerdave said. But I would have to rebuild the equivalent crossover without the tapped transformers, so it probably wouldn´t work.
But the dusted 2425j looks like a smart idea. Would work without crossover changes, except for the level of the l-pads...

In that case, do you guys think the Yuichi A480 is a major upgrade to the original 2307-2308 horn-lens?

[Ian Mackenzie]

Thanks at all for the great input!

But the dusted 2425j looks like a smart idea. Would work without crossover changes, except for the level of the l-pads...


Unfortunately your statement above is incorrect.

Please refer to my earlier post.

Neither the response nor the impedance of a dusted diaphragm is suited to the stock network or the equivalent.

As was stated earlier you need to get you hands dirty then you will learn and grow with this hobby. In time you will start to understand how to design a diy loudspeak (There is no other around it.)

In the diy loudspeaker space one persons opinion (which is quite often the case) could be another’s bullshit.

And the measurements don’t necessarily tell you if it works. You need to make things and listen like everyone else.

If you don’t feel that is the road you can build a JBL design as closely as you can and get a good result.

We’ve been over this a thousand time

[Dr.db]

My mistake, I thought the idea was to aquaplas the existing diaphragm. I wasn´t aware of any aquaplased diaphragms readily available for 1"-drivers...

[marco_gea]

In that case, do you guys think the Yuichi A480 is a major upgrade to the original 2307-2308 horn-lens?

YES.

But, only once all the caveats discussed above regarding crossover and overall system integration have been duly considered.

[Ian Mackenzie]

My mistake, I thought the idea was to aquaplas the existing diaphragm. I wasn´t aware of any aquaplased diaphragms readily available for 1"-drivers...

They are only available now for large formal Jbl compression drivers.

Jbl only did acquaplas 1 inch diaphragms for the 275nd used in several systems in the1990’s.

The only other 1 inch aquaplas diaphragm are diy attempts about 10 years ago.

But my point as described early is under scrutinised measurements the diy acquaplas diaphragm response is different from the stock diaphragm and requires a modified network.

Does this matter? If you are just interested in swapping diaphragm on face value at home then No.
The same applies to any matter of horns.

BUT, if you are an audiophile and serious diy amatuer then making changes and getting then to work properly requires a disciplined approach, the right tools and significant understanding of what really matters and how things work.

I am an staunch advocate of the latter because it’s a waste of time otherwise if you are serious about audio reproduction.

Please read the commentary by this widely respected diy loudspeaker designer for an illustrated guide to how to succeed. The article also provides insights into the unique qualities of well executed four way system.

My own experience and investigations over the past 20 years indicates the drivers were and still are under utilised in these systems and with associated components. Meaning everything matters.

My unqualified advice is before you change anything buy or build the best the best amps you can afford for the mid, horn and uhf.

http://www.troelsgravesen.dk/The-Loudspeaker.htm

[Robh3606]

But my point as described early is under scrutinised measurements the diy acquaplas diaphragm response is different from the stock diaphragm and requires a modified network.

That certainly was the case with the dusted 2435's. Even if you don't dust them you need to run impedance and frequency response plots on the drivers to make sure you have the diaphragms properly re-installed.

Rob:)

[Dr.db]

Thanks for the link, very nice to read.

Regarding the diy-game, I get you.