Quick & Dirty 4430-Inspired Two-Ways Part I

[Zilch]

We have to figure out a passive network for the 2431/Waveguide combo. I prefer passives.

Yup. I'm starting with the AM6212/00 factory crossovers you cited earlier. These are a decidedly different species; we'll be taking meaurements to determine what is required as soon as the other (rectangular) version arrives.

Just as well, perhaps. N3134 is a big stumbling block to an L200 upgrade. As always, I'm looking for something simple. With a little more investigation, we may find that N200B is applicable here, for those who have them. For L200/B upgrades, I'm aiming to reuse the LE85's that come with, but for a "sweetened" 4430, the 2431H seems quite promising at this point. Even with the wrong crossover, they sound good!

The waveguide alone may be an improvement to 4430. It's certainly a more reasonably priced option for DIY, and with the throat adapter, may accommodate 2426/7. An even cheaper choice would be 2418H-1, as used in AC2215/00, though that's gettin' more like 4425, probably. We'll see....

[Zilch]

0.11 uf Metallized Polypropylene (MKP) capacitors are kludged across C10 and C110 on solder side of 5235 crossover mother board to lower 30 Hz Q=2 bump filter to B380 26 Hz tuning per BX63(A) spec. Mod also lowers 20 Hz filter to 18 Hz if dipswitch S1 is enabled as well as S2, the new 26 Hz setting (heh, heh....) Factory had a little kludge of their own there to replace a missing trace on this particular board.

Did it work? Well, it's not apparent on the RTA display capture shown below, but Pink Floyd says "YES!"

Note significant 2431H HF extension to 20 kHz and likely beyond on the RTA display. These ARE improved drivers, at least when used on PT waveguides.

L200's, one port in each blocked with plumbing test plugs for 26 Hz box tuning, await arrival of Giskard's 2235's. At approximately 5.25 cu.ft., they dwarf wimpy-weenie B380's. For standard 4430 tuning, merely remove these proprietary precision port tuning devices.

Does this begin to make sense yet?

[Robh3606]

What's the resolution on the RTA?? How many dB per LED?? What's the throat size on the 2431's??

Rob

[Zilch]

RTA resolution is switch selectable for 1 dB or 3 dB per row. That's a 3 dB capture. 1 dB is, ummmm, "messy."

Still using N3134 crossover with them.

2431H is 1.5 inch throat, bolt-on....

[Robh3606]

So you would go direct on the new horns?? Were they expensive??

Rob

[Zilch]

List for 2431HPL is $399, and the horns are $84, both from Pro Parts.

Yeah, they're direct bolt-on. See pics on previous page.

Owing to neodymium magnet and less bulk in the horn, the entire assembly is very light, maybe 1/3 the weight of 2426/2344A. It could possibly self-support....

[Robh3606]

Thanks those horns are very reasonable, the drivers are not that bad either. Are they aluminum, aquaplas dusted aluminum?

Rob

[Zilch]

I believe they are aluminum, which by many accounts sounds smoother than titanium. There may be other design enhancements, but I'm not about to open this pair up to "discover" them just yet.

There's no published spec sheet that I've been able to find, either, just their specs in a few current systems. We'll measure them to determine their actual performance, once the other waveguides come in. All I have to go by so far is what I hear and what I see on the RTA in comparison to known drivers we've studied here.

They need different EQ than 2426H, for example, and they exhibit extended HF in comparison. They sound very good on the PT waveguides, and they're not bad on the P-Audio horns, either, as I suggested to Guido earlier.

They've been manufactured for about a year now, I think, maybe longer. My two units are "vintage;" final test dates 2/20/04 and 5/26/04.

[Zilch]

Zilch!

We have to figure out a passive network for the 2431/Waveguide combo. I prefer passive networks.

All of the action is in the HF section of the 4430 passive crossover, it seems, which stays in play when biamping. It'd be good if the same approach were continued so they could be used either way. The HF portion of 5235 is a simple single pole.

The LF section of N3134 is easily replicated and maintained, either passive or active, except L102, perhaps. I don't know why the LF of AM6212/00 is so "busy." Looks like a couple of notches to tame 2206H, which we don't need. Nobody's highlighted any special role of 5235 EQ card LF filter that I've seen. I'd like to keep 5235's separate 26 Hz bump filter when biamping, tho, obviously.

I am confident others here will help us, once we determine what is actually required.

[4313B]

This seems to be a good time to post some old notes -

The 2431 is an improved 2430 and offers better high frequency response. It has an improved surround. The 435AL is the consumer version of the 2431. They all use aluminum diaphragms with improved kapton surrounds.

The 2425/2426 is fairly breakup free up to around 15 kHz. It starts to roll off around 2 kHz and is easy to passively EQ. The aluminum diaphragm offers better damping while the titanium diaphragm offers greater reliability.

The 435BE is the consumer version of the 2435 and has a larger back can to lower the useable response from around 1 kHz to around 650 Hz. It is quite happy down to 800 Hz. Slope roll-off is curved such that passive EQ is difficult. Roll-off starts around 7 kHz. Digital parametric EQ is the way to go. Watch out for latency in digital crossovers.

The 045 drivers are compression drivers with very small apatures, 0.35", that allow for excellent HF dispersion. They are flat to 50 kHz whereas the ring radiators drop like a rock around 15 to 20 kHz. The ring radiators are extremely efficient at the expense of ultra flat wide bandwidth response.

There's no published spec sheet that I've been able to find, either, just their specs in a few current systems. We'll measure them to determine their actual performance, once the other waveguides come in. All I have to go by so far is what I hear and what I see on the RTA in comparison to known drivers we've studied here.

Yes, the burden of measuring these various drivers rests with you.

The LF section of N3134 is easily replicated and maintained, either passive or active, except L102, perhaps.

Indeed it is. L102 is a 0.6 ohm DCR inductor - easy enough to aquire.

[Mr. Widget]

The 435AL is the consumer version of the 2431.

The 435BE is the consumer version of the 2435 and has a larger back can to lower the useable response from around 1 kHz to around 650 Hz.

The 045 drivers are compression drivers with very small apatures, 0.35", that allow for excellent HF dispersion.

Hey Giskard,

I have a couple of questions/clarifications. If you haven't already discussed these with the powers that be maybe you could inquire.

Doesn't the 435AL also have the larger back cap like the 435Be but the 2431 does not?

Aren't there also differences between the 435Be and 2435 in the amount of aquaplass dusted on the diaphragm and/or ferro fluid added?

What are the performance differences between the 045Be and 045AL? The cost difference is significant.

Widget

[Zilch]

Thank you, Giskard!

The Be verisons are HELLA expensive, alas, but what you say about 2431H is all good, except, perhaps, for the "difficult to do passive EQ" part.

I'm surprised 2431H hasn't gotten more play from my experience with it so far.

[Hmmmm. Wonders what 435AL is used in? Maybe there's some clues there....]

[Guido]

All of the action is in the HF section of the 4430 passive crossover, it seems, which stays in play when biamping.

Yep!

I did a spice analysis with the AM6212 xover but do not have it here. I can post it tomorrow. There was one or two notch(es) to tame the response peaks. If we succeed with implementing these into the 4430/4435 networks that's the ticket!

I still didn't start my 4435's.

[Guido]

The 2431 is an improved 2430 and offers better high frequency response. It has an improved surround.

Ha! I was wondering why the 2430 and 2431 had the same price till mid 2004 and then suddenly the 2431 price where rised by 100 USD. Now I understand

[Zilch]

I did a spice analysis with the AM6212 xover but do not have it here.

I previously cited your earlier exchange with Giskard regarding AM6212 crossover topology here:

http://audioheritage.org/vbulletin/s...ead.php?t=2175

It seemed incomplete and I couldn't quite follow it. I'd say it's worth a new look and update, now, yes!