Before going 1st order, I tried this...

[4343mod]

This is the exact circuit in the 4343 XO, minus provision for a rotary switch or L pads.

This crossover is built on a 1/4" thick piece of Aluminum, 24" x 16".
I built only one, though I have another piece of Al to build another one. After a few weeks of listening to this, I decided to abandon it and build the 1st order versions.

The price of the parts for this one XO was about the same as the parts to build 2 of the 1st order XO's that I use, since this XO has about twice the parts as one of my 1st order ones.

The sound is superior to the OEM XO, with greater transparency and detail. I still prefer the sound of the 6dB ones, they bring me in even closer. The only draw back between the two is the 1st order XO's really show the drivers inherent flaws.

The 2 inductors you see on the LF is due to a change I made after building this. I disconnected the 5mH air core and went with the 9mH ERSE laminated core.

What do you guys think? Is it over the top?

[4313B]

"The only draw back between the two is the 1st order XO's really show the drivers inherent flaws."

What flaws? Are you sure the filters are complementing the drivers properly?

"What do you guys think? Is it over the top?"

No. Good job!

[4343mod]

Originally posted by Giskard
"The only draw back between the two is the 1st order XO's really show the drivers inherent flaws."

What flaws? Are you sure the filters are complementing the drivers properly?

"What do you guys think? Is it over the top?"

No. Good job!

Thanks, Giskard. Well, I think JBL CD's are inherently "rough" sounding, at least compared to say, a Morel dome midrange like the MDM-55; and to make our CD's smoother requires more agressive filtering (like above) that diminishes transient response and detail. Of course, the MDM-55 is a different animal, and is a wimp compared to the 2420. Response of the '55 is like my 2420 with a sock stuffed in the horn.

If you or anyone can recomend a different coil and cap combo for a 1st order bandpass than the 9uF/2mH that I'm using, I'd certainly give it a try. Heck, I'll try anything!

[4343mod]

Bump
The 2121 uses a 2mH choke, the 2420 choke is .25 to .10mH, depending on the mood.

Clearing up a mistake I made 21 months ago. Still going stong and time indeed does fly when you having fun!

[Lancer]

I'd like to see a schematic so I could plug it all into a spice program to see what kind of voltage drives it yields.

[spkrman57]

I am a big fan of 1st order crossovers, my only limitations using them is the power handling ability is greatly reduced. That is unless you want to replace diaphrams often!
Ron

[stevem]

This crossover is built on a 1/4" thick piece of Aluminum, 24" x 16".

I might be wrong about this, but aren't you supposed to keep conductive metal away from inductors? Or does this just apply to ferrous materials? Would this crossover sound any different on a piece of masonite (or pegboard)?

[4343mod]

Here is the 1st order with current values in use. All resistors are 12W Mills. I also sometimes run the LF with a 12dB/octave XO, coming up...

[4343mod]

w/ 12dB/octave LF

[Lancer]

Pleasure. It's a great thread and will belongs here. Now, let's delete our chatter and clean things up...

What?

Here is the 1st order with current values in use. All resistors are 12W Mills. I also sometimes run the LF with a 12dB/octave XO, coming up...

Here's what it plays out like. Stock 4343 in gold.

[boputnam]

What?
Here's what it plays out like. Stock 4343 in gold.

I'm not sure I should pay up for that response...

[4343mod]

Thanks for plugging that in Lancer, I appreciate you taking the time in doing so! I'm burning Man With The Horn on a Tascam 5000 right now, and it sounds wide and deep using the 12dB LF.

Can you reverse engineer and give values that look nice? I'll buy the parts, solder them up on new wood sheets, listen back and forth between the two and invite everybody over for a listening party after I have an opinion. You are all invited!

[edit] Ok I see the peak in the MF and the red is the 2405.
I dunno, what are you going to do with that 2121? Its like 88 or 89dB, in between a 98dB woofer and a 100+ dB CD, so you got to bring it up, IE no padding. You get the peak, but in the original 4343, tell me honestly- Has anyone heard the 10" even work in a stock cab? Mine was inaudible at close range before the new filters, and now when you listen to the baffle, you actually hear that driver sharing. the more I look at that plot, the more I nod and think "wow, yeah" The slot would be on fire if it was down much more, 2uF is pushing it. look at the MF, HF, and UHF- They still share parallel lines up to the apex. I could use a smaller shunt resistor on the UHF to heat it up, but the presence is actually pretty good here in RL. I would still like to try an optimised model.

Any tips for better value selections? Please?
I think I'm done with my edits. *pew*

[Ian Mackenzie]

Firstly the original filters were design to provide bandpass (and stop bands) for the 4 drivers to work tightly together. Secondly the filters were adjusted to incorporate contours to flaten the combined response of the respective drivers and ensure the combined response was flat as possible.

The cruves are voltage drives not frequency response curves, as such they show the attenuated voltage to each driver.

While the 1st order slope may offer some benefits, note these drivers were not designed for 1st order crossovers.

Ian

[Lancer]

Can you reverse engineer and give values that look nice?

I no longer have any of the components you are using.

I dunno, what are you going to do with that 2121? Its like 88 or 89dB, in between a 98dB woofer and a 100+ dB CD, so you got to bring it up, IE no padding.

The 2121 should be a mid 90's driver, a tad more efficient than the 93-94 dB 2231. If you use 6 dB/octave filters and cross it over correctly you will probably loose a few dB of sensitivity.

I would still like to try an optimised model.

Yeah, it could be fun. Maybe you should think about picking up something like SoundEasy so you can dial in the slopes/crossover points you want and try them. Measure the results of what you are hearing.

I suspect Ian is right though. It could be tough to get these particular drivers to sound "right" with 6 dB/octave slopes. I think you need some test gear to help you.

I'd personally upgrade the 4343's to 4344's. I consider the 4344, 4345, and 4355 the end of the line for the 43xx series. My opinion is they had the most refined filter designs for those component types. No telling what someone with the time, inclination, and proper combination of software/hardware test gear could come up with for them today though.

[Ian Mackenzie]

Just to add a dollop of tomato sauce here I have found that biamping the 4344-45 requires a re think of the system voicing in terms of driver integration.

Subject to me running some detailed acoustic measurements the behaviour of the midrange bandpass filter (2122H in the 4344-45) alters in bi amp mode as its no longer a pure passive bandpass filter with customised 12db slopes but an 18 db electronic /12 passive bandpass.

To cut a long story short I had to pump up the 2122 a db or so to stop the horn from sounding like it was screaming its head off. Why? I really don't know but have a theory.

My reasoning, again subject to measurements of voltage drive is that bandpass filters typically have a specified Q, and as such the value of that Q can effect the shape of the filter response and even add some real gain at resonance. My hypothesis is that removing the passive low pass elemens has altered this balance and overall level slightly and some re alignment of the mid driver level is required to ensure overall system flatness.

Heck, I am sure most members just think they can plug and play and yeah biamping is always best and then we hear of mutterings of dissatisfaction, horns sounding like sea shells, poor imaging and worse listener fatique.

Frankly, pull passive mode is far more managable and more stable as a known set of variables.

Its a bit like asking yourself why do tiles fall of the Shuttle sometimes and not others or have they always been falling of but we weren't aware of it?

Well unfortunately with speaker system crossover design its all a careful balance and everything tends to react, so if you upset some aspect of the design it will mean adjustment is required elsewhere.

Hence if you start playing with 1 st order filters its a whole new world of hurt to deal with before sanity and order will be restored.

Ian