2435 HPL- 2 way 750hz

[4313B]

Hmm, I do wish Giskard would leave his posts up - sometimes they are an oasis of reason!

Because this kind of activity drives me nuts. I find it pointless. I don't care one whit about 2431's or waveguides. I want to see more of what people like John can turn out.

Certainly the 435al sounds rough as a badgers arse if it's left to run out to my uncultured ears - the 435be seems to fair slightly better.

Your assessment of the 435Al is the general consensus. I forgot, are yours aquaplased? G.T. had to aquaplas them for use in the Array. I personally think the 435Be is better than slightly.

But for the best HF performance, certainly in my limited experience, I'd say JBL got it right when utilising the 045's.

People at JBL really like them. They certainly are a big step above the old ring radiators and domes.

I look forward to the 476be -a major step up in every way.

It really is a fantastic piece of hardware.

[Robh3606]

"Looks like I have most of the parts to build Rob's filter, so I'll do that and compare the results with all three."

Hello Zilch

That filter is a work inprocess as I learn the software and get a better understanding of how the networks actually work. Until I get measurements from a bass driver in my cabinets it is guaranteed to be revised. I posted it with the Project May graphic to show what the basic topollogy was for a passive version of Gregs filter. I also wanted to show that the basic topology was good for different horns but that they would indeed be different filters. I don't think it's ready to be built just yet.

Rob

[Zilch]

"I don't think it's ready to be built just yet.

No problem, Rob. I'll wait for you to finalize your design....

[Zilch]

I heard tell that there was also a plan afoot on the forum to buy a pair of my 4355 and 4344/4345 networks and then reverse engineer them real fast and post the schematics.

I trust you know that ain't me, Giskard.

There's nothing more boring than building more than one of any network. Even the second one for a pair is a drudge.

[That's why I mirror-image them, so I hafta pay attention and stay awake long enough to complete the task... ]

[merlin]

Because this kind of activity drives me nuts. I find it pointless.

I do see your point.

Your assessment of the 435Al is the general consensus. I forgot, are yours aquaplased?

I really haven't checked Giskard - mine came installed in a shiney pair of 4338's. Must find out now - but it sounds like there's are really nasty mode somewhere below 15khz. The 435be is a lot smoother as you say - I was being polite

I need to build up some spares, then, apparently....

Many thanks for the thoughts Zilch. My intention at the moment is to stick with the active route - with some simple and judicious use of EQ. It's easy and works quite well IME. I'm only here for the music!

[4313B]

I really haven't checked Giskard

I checked, they're not. No biggie.

My intention at the moment is to stick with the active route - with some simple and judicious use of EQ. It's easy and works quite well IME. I'm only here for the music!

Good plan. Sounds like you've been talking to the right people.

[DJ Vincenzo]

I presume you mean "too loud."

The answer is "No."

If you read and understand Rob's posts, you'll see that 10 - 15 dB of the compression driver's SPL capability is consumed in performing the requisite compensation to achieve the desired frequency response with constant directivity horns/waveguides....

which desired freq? 750 X point? it doenst seem far from the 1000 jbl designed it for.......but im exicted...my brand new AM6400/95's just got drop shipped from Cali and im gonna try switching out the highs with these 2435's i bought on ebay a while back this weekend.

[Zilch]

which desired freq? 750 X point?

Not the crossover frequency, the desired frequency response, i.e., flat, in this case....

[Zilch]

Hi, Rob.

Would you run optimization on this topology for 2435HPL on PT-H1010 using your measurements for me, please?

It's kinda labor-intensive tweakin' this the "hard" way.

I've replaced the series notch filter with a broadband low-Q parallel one.

Same one as I used here, actually:

http://audioheritage.org/vbulletin/s...719#post135719

Frequency response is appreciably closer to flat, now +/- 2 dB or better. Compare to green curve in top group here:

http://audioheritage.org/vbulletin/s...8&postcount=66

Compare filter impedance and phase here:

http://audioheritage.org/vbulletin/s...2&postcount=28

Orchid HF curve is windowed.

M552 highpass is set at first mark past 900 Hz, i.e., 920 Hz, if accurate....

[Robh3606]

Sure when I get home I will run it and post it for you. The blue traces are the generic filter. The first set is the original. The second is with L3 changed to .050mH

Rob

[Zilch]

COOL! Thank you, Rob.

I'll go try that. Looks like 0.06 may be a good compromise not to pull down 14 - 16 kHz any.

The Spice sim looks very much like AM6212/00, but without that circuit's low impedance problems (~2 Ohms above 10 kHz).

[Zilch]

Wonder what they will sound like???

I listened to V1 most of the day today, and they were fine. It's an easy add-on once you've got the basic filter built. They don't sound bad with just that, either.

Biggest change came from switching to active. LE14H-1 certainly likes being connected direct, and the HF gained substantial clarity on the T-Amp as opposed to 3110s.

You're much faster than me and my soldering iron with these tweaks! Please leave them up at least overnight so I can get them printed and saved. Haven't had dinner yet, so I gotta take a break.

Here's V2, top, showing the effects of varying L3. It punches up lower down, tho, like squeezing a water balloon.

Lowering R3 progressively to 2.5 Ohms leveled out the notch at 10 kHz, but it's kinda back to just the core filter, looks like, i.e., the parallel notch is, in effect, defeated. Listening to that now. [Edit: Decidedly "forward."]

Other thing I have to do is measure a bunch of 2435HPL to find a "typical" for this work....

[Zilch]

O.K., here's V1.0 and I'm throwin' down the gauntlet.

+/- 2.7144 dB out to 18 kHz @ 1/12 octave smoothing.

Anybody with 2435HPL wants to play along, PT-H1010 waveguide, JBL part 338785-002 is $94.

It's not playing all that bad on compact PT-F1010 part 338786-001, either, $98 each.

Full biased version of the filter with lithium battery power and premium terminals is under $50.

Note: For those reading just this post, this is a compensation filter for this specific driver/waveguide pair, and is not a crossover network. It must be mated with an active or passive crossover at an appropriate frequency for integration into a system with a compatible low frequency section. These measurements were taken using a JBL M552 24 dB/octave L/R active crossover set to 920 Hz....

[kartsmart]

Will this filter work on the PH-316 horn ? How low would you think the PH-316 would be good for ? Of course that is with the 2435 drivers, I am In the prosess of setting up a PC RTA so I can play along

[Zilch]

Good question.

We studied PH-316 over in Mr. Widget's Horn/Driver Comparison thread. It may make it down to 800 Hz, as I recall, but you can look it up there.

Guido is using that horn with 2435, I believe, and working on modifying the throat for true biradial performance.

I'll try it on this filter and see what happens for you; this IS your thread, after all.

A BIG thumbsup on getting an RTA running....

Edit: Certainly listenable, it starts sounding "horny" below 900 Hz, tho. I'd say 800 Hz is the lower limit.

Try increasing L3 to knock down the hump just above 2 kHz a bit, maybe.

I'd try and figure out where the glitch at 10.356 kHz is coming from, too, though I can't say I actually hear it. See PT-F curve in #43, above, where it also appears. It's not this driver, specifically:

http://audioheritage.org/vbulletin/s...389#post147389

I should look at the influence of driver/throat mounting concentricity.

ReEdit: Had to move the driver an obvious 1/8" off center before the response was altered appreciably. That ain't it....