3-way vs. 4-way

[Dr.db]

Hey guys,

I was thinking about a new project for a friend of mine.
He has listened to my system and wants similar parts, but we want to improve it a little compared to my DIY-speakers.

The following parts are on hand:
2234H
2123H
2441
2405

For those of you asking why 2234h instead of 2235h; the 2235 just sounded to muddy above 300hz and the 2234h with it´s lighter cone would be prefered for the 3-way...
Now there are two possible setup´s we are thinking about.

Option 1:
- 2234H
- 2441 on yuichi-horn
- 2405
- L300-crossover (800hz, 8,5khz)

Option 2:
- 2234H
- 2123H
- 2441 on 2397-horn
- 2405
- 3155-crossover (300hz, 1,2khz, 10khz)


We are wondering which setup would be the best.
A 3-way design with a big yuichi horn that allows to crossover the 2441 at 800hz? The big horn is really tempting and would sound great with the 2441. But on the other hand, the 15" woofer below 800hz doesn´t sound as snappy and dynamic as the 10" cone in the 4-way design...

Looking forward to hear your ideas

[1audiohack]

There are so many consideration to factor in a project like this. Some of mine would be:

How loud will it be expected to play?

What will the listening distance likely be?

Is the room live and reflective or absorptive?

Is the music to be listened to mostly field recorded or studio produced?

Here’s my thinking:

The 2397 is dispersive in the horizontal plane and diffractive in the vertical. A 2397 not mounted in a baffle sounds nearly the same from behind as in front. In a live room that could be a problem. With field music like orchestra, the sound from everywhere can be a plus.

If you listen loud, four ways get louder and sound better than two ways.

There is no doubt in my mind that no 15” woofer will pop like a 2123.

If real loud is not a main consideration and if the listening distance is not great, a two way integrates better and easier.

The Yuichi will have better pattern control and as a byproduct of that better imaging and would be my prefered for listening biased to studio recorded music where imaging is more a factor than envelopment of field recordings.

The 2441 should get to about 16k on either horn. I like to let them play up as far as they will cleanly and get the super tweeter to cover only what the large fromat driver wont. The system will play louder while maintaining the big system sound of a large format system.

Any of your listed possible combinations would require subwoofers for me.

These opinions are from a guy who likes the sound of large format HiFiPa. Maybe that disqualifies me.

All the best,
Barry.

[Chevelle]

I had the 4-way with 2123 lo-mid which has really "snappy" sound. But what I like more with my 3-way is the sound from a 2" driver as low as possible. I cross the 2445 at 700Hz with my 2311 short horn, let it go up to 16k and have a 2405 on top.
This is best in my ears for short listening distance under 3.5meters in a live room and medium listening levels. The 4-ways only had an advantage at very high listening levels.

Hjalmar

[Dr.db]

Thanks a lot for these declarations!

I like the idea of keeping the acoustic centers of the drivers close to each other to minimize group delay and phase problems.
The 2234 has a steeper e.g. deeper cone than the 2123, I get you. But this difference might be ~5cm, so most of the offset to the horndriver still remains. So I would think their shouldn´t be a big difference between 2234 and 2123 concerning the offset-issue, or did I miss anything?

I believe the reflections of the 2234 or 2123 underneath the horn would make a much bigger difference.
The narrower the dispersion of the cone driver, the less reflections will bounce of the bottom of the horn. So the 15" woofer should be clearly in advantage here.
You said the dispersion of a 15" woofer around 1khz should be around 100°. What about a 10" woofer at 1khz for comparison?



We are not tied to the 2123´s, if it doesn´t make any sense we will sell them.

There is only the chance to mount one 15" driver because of the size, so no 4435 clone on the bass department...

[Ian Mackenzie]

Thanks a lot for these declarations!

I like the idea of keeping the acoustic centers of the drivers close to each other to minimize group delay and phase problems.
The 2234 has a steeper e.g. deeper cone than the 2123, I get you. But this difference might be ~5cm, so most of the offset to the horndriver still remains. So I would think their shouldn´t be a big difference between 2234 and 2123 concerning the offset-issue, or did I miss anything?

I believe the reflections of the 2234 or 2123 underneath the horn would make a much bigger difference.
The narrower the dispersion of the cone driver, the less reflections will bounce of the bottom of the horn. So the 15" woofer should be clearly in advantage here.
You said the dispersion of a 15" woofer around 1khz should be around 100°. What about a 10" woofer at 1khz for comparison?



We are not tied to the 2123´s, if it doesn´t make any sense we will sell them.

There is only the chance to mount one 15" driver because of the size, so no 4435 clone on the bass department...

There is a lot more to it can l have discuss here

At lower frequencies the phase shift or group delay had a bigger impact

When we look at 700 hertz the acoustic centre of the woofer and the combined group delay of the low pass filter on the crossover results in a far better situation than attempting the 2123 at 1000 hertz or above

In terms of the suitable woofer no one is complaining about the 2216 in the M2

Tad go a step further with a sharper crossover slope of 36 dB that also introduces more group delay

Remember the A290 is a hypex horn and they are longer than the typical exponential horn

Tad monitoring has a reputation for a reason

The alternative if you think you must use the 2123 is mount it on. Pi mid wood horn for 200-1000 hertz

If you intend to use the 2123 then consider a different horn specifically design for a 1200 crossover point

That will have far less depth by design in a modern wave guide

There are lots of options

Frankly if you get the 2234 in there it’s pointless

[Chevelle]

Forgot to say that the 3-way sounds good with passive crossover, I use the Nelson Pass modified version of the 3133. The 4-way does not sound best with passive crossover. This system really wakes up when run active with a good amp for the 2234/35.

Hjalmar

[bldozier]

which kind of music are you playing and are you biamp your system

[Ian Mackenzie]

If you can do the big wood horn the run the 2234 up to 650 hertz with 4th or 6th order filters

Using the 2123 with those other drivers is pointless

Besides there will be too much group delay from the 2123 shallow cone to the compression driver.

Of course pro pa guys will quote it out of the “book” but that is not any where refined enough for the financial outlay in a home listening situation.

It will never sound right

The devil is always in the details and that is the crossover phase tracking and getting the 2/3 way to track your preferred room listening curve. That’s not easy as none of this is plug and play

At best you will have a “wall “ of sound. I stopped doing that about 20 years ago.
Greg’s designs are always the first port of call before you go off the reservation into the diy wilderness

With a 4 way is easier for the diy person to manipulate the levels to track your room curve but you still need some tools to measure the levels.

I currently shave the levels on the 43xx 4 ways to within 0.25 dB with LinearX LMS

The balance and audiophile “thing” snaps right in with that order of precision

[1audiohack]

....Of course pro pa guys will quote it out of the “book” but that is not any where refined enough for the financial outlay in a home listening situation....

Hi Ian;

To what does this refer? Sorry you lost me.

Barry.

[Ian Mackenzie]

Hi Ian;

To what does this refer? Sorry you lost me.

Barry.

What l was referring to was going to a pa handbook, making a number of assumptions and then building a (large) pa rig is not the same as designing and testing a domestic hi fidelity loudspeaker or a studio monitor.

[Dr.db]

I found this picture with google...
The yuichi horn protrudes a fair bit, would this be ok?
Would you mount the 15" woofer a bit lower to increase the distance to the horn, or can woofer and horn be mounted closely against each other?

[dn92]

I found this picture with google...
The yuichi horn protrudes a fair bit, would this be ok?
Would you mount the 15" woofer a bit lower to increase the distance to the horn, or can woofer and horn be mounted closely against each other?

Mount the 15" as close to the horn as possible. On Rey Audio Loudspeakers there is 1cm between TL-1601b and RH-3. On Exclusive 2401 and 2402 there was more space between due to the presence of a grid.

I know (I've seen and heard them twice, the owner is a friend, equiped with JBL 2235H and 2450J) the loudspeakers shown in this picture.
The horn is a clone of TH-4001 made by ALG Audio in the South of France. ALG is also manufacturing Arai290 horns.

[berga12]

2234h --> 300hz (or 250hz)
2123 --> 300 (250hz)-800hz
2441 --> 800 - 10.000hz
2405 --> >10.000hz

easy game.

with 2441 you can go down to 500hz in a Yuichi 290 or Iwata, in this case you can get rid of the 2123, but then you loose in "guitar type" instrument reproduction and Low voice dimension.

[ivica]

2234h --> 300hz (or 250hz)
2123 --> 300 (250hz)-800hz
2441 --> 800 - 10.000hz
2405 --> >10.000hz

easy game.

with 2441 you can go down to 500hz in a Yuichi 290 or Iwata, in this case you can get rid of the 2123, but then you loose in "guitar type" instrument reproduction and Low voice dimension.

Hi

I would suggest to use 2441 (AL) over 1250Hz, as such 'high FR' would not be a problem for 2123, and AL D16R2441 diaphragm will be protected from "over displacement" of too low FR.
Anyhow A290 (with fins) would be my favorite solution. From the Dr. Yuichi measurements, it seems that 2405 can be omitted.

Generally speaking using many drivers would introduce 'comb filter effect' round crossover frequency depending of the listener position
and can not be compensated to all spacial positions.


regards
ivica

[Ian Mackenzie]

The 2234 would be satisfactory up to 1000 hertz