PDA

View Full Version : DIY time aligned XO, 12db/oct @ 1khz



readswift
03-23-2008, 08:17 PM
can I chime in with some time Time Align™ stuff ? :idea:
http://www.geocities.com/kreskovs/CrossoverdocN.html

I'd use it with 4435 , what do you think?
Amp idea ? I can choose between 80W Class-A and 400W class-D / side. ( for now ,on the long run i want both :applaud:)

yggdrasil
03-24-2008, 04:37 AM
After reading quickly through the documents my first concern is that the overlap will be dangerous to the HF driver. This could again be one of my short circuits since the summed response will be flat. But again the rolloff is at a lower frequency, so I think you will end up feeding the HF driver more low frequency signal.

Given that this is true you will also feed the 2234's with more high frequency signals. There will be no danger of damage to them, but my guess is that they do not sound at their best at even higher frequencies.

Lastly I see the question of acoustical centers. I don't remember if both the 2234's go all the way in the stock 4435. If that's so, you will have a serious challenge finding an acoustical center, so you would probably have to make a workaround.

Good luck.

readswift
03-24-2008, 05:14 AM
the woofers aren't parallel . Only one works up to 1khz and second is low passed at 100hz with a first(?) order LP AFAIK. So this part is fairly simple.
The overlap as shown in figure 15, well , judging by the looks, it isnt *that* bad, of course I have zero real life knowledge -yet . BTW I have 2426j (16ohm by spec ) HF instead of the 2426h ( 8ohm).

yggdrasil
03-24-2008, 05:41 AM
The 2426J will have an impact on the stock compensation network for the HF driver, which I assume you will be keeping.

Regarding the low frequencies reaching the HF driver you will have to put both the stock and your new design through a spice simulator to see which voltages comes through to the driver at your highest expected sound pressure, and then add some headroom.

If the results are acceptable, then go for it. If not you will blow dia's.

readswift
03-24-2008, 10:37 AM
the single best picture about the original crossover. I dunno if I want to keep the analog HF compensation, doing it in digital realm might add up better :
http://audioheritage.org/vbulletin/attachment.php?attachmentid=5318&stc=1&d=1107977866

Ian Mackenzie
03-24-2008, 01:23 PM
The 4435 passive crossover is already min phase/group delay type where both the woofer and horn are electrically in phase and the horn is time aligned.