Designing Usher 8945P 2-way: Seas TBFC or Pls HDS tweeter can be C/O'ed at 1.5 kHz?

[Jay_WJ]

Are there any tweeters that can be crossed over at 1.5kHz with -18dB/oct acoustic slope?

I'm designing a 2 way bookshelf speaker using the Usher 8945P driver, and really want to use the Peerless HDS (or Seas TBFC or TDFC if better) at this c/o frequency because that's the highest c/o point that can provide proper response shaping for the 8945P's noticeable dip at 1.2 kHz.

My initial design requires a 1.5 to 1.6 kHz c/o point with about -18dB/oct acoustic slope (NOT electrical 3rd order) for the tweeter. Can the HDS tweeter or the Seas be used in this case without a problem?

I don't usually listen very loud, but even so, what if I end up blowing the tweeter?

[augerpro]

I just measured a ScanSpeak 9800 and was amazed:






And distortion at 95dB:






Distortion isn't great down low adn I don't know if the excursion would be limiting but otherwise this looks pretty usable down low. At least to me

Have you considered using the HDS in one of those MCM waveguides? That should drop th usable XO point down some.

[Dave Bullet]

Have you measured the 8945P in the intended baffle?

Have you simulated or measured a baffle for the 8945P to give a rise at 1KHz?

Often baffles that are 220 wide for a 7" driver will cause a rise in the 1KHz area to offset any driver dips.

Try playing with the FRD consorts' BDS to see if you can model an enclosure to give a rise. This might allow you to more safely crossover higher.

Depending on how loud you are playing this (and whether the woofer would be excursion limited at high volume) - I would think 4th order acoustic slopes would be safer with most 1" domes at 1.5KHz.

Cheers,
DAvid.

[Dave Bullet]

Here's another crude way to tell if there is enough cut for the tweeter...

Have a look at the tweeter slopes for a bunch of 2 way designs. How much energy is going into the tweeter around it's Fs? Say for example the max is -40dB - that would indicate whether the slope on the Peerless is rapid enough - also allowing for the Fs of the tweeter you are comparing with.

Another poster here blew an 810921 (recent post) - might have more info.

I'm just trying to think of other lateral ways to evaluate....

Cheers,
DAvid.

[Jay_WJ]

Originally posted by Dave Bullet
Often baffles that are 220 wide for a 7" driver will cause a rise in the 1KHz area to offset any driver dips.

The 8945P's response dip at 1.2 kHz is unusually deep and cannot be offset by diffraction effect. Existing designs for the 8945P (and even the 8945A, which has a less severe dip) use a 1.9 kHz to 2.2 kHz c/o point. This does not help to eliminate it. A notch filter right before the dip frequency, combined with electrical 3rd order filter, can remove it. But it requires a low c/o point for the tweeter.

[ttan98]

Jay,

I have in hand a pair of 8945A, haven't made any measurements as yet(still burning in) even though I have the tools.

There are a no of sites which deal with the design of this driver. The few points that arise are:

1. A rise(a bump) in the 850-900Hz region, this rise is noticeable during listening and needs to removed via a notch filter.

2. A dip around 1.2khz(first time mention by you), I read in a no of places are not so noticeable and sometimes preferred. This dip is not corrected in the designs I came across.

look at this site for the design:

http://murphyblaster.com/content.php?f=usher2way.html

cheers.

[Jay_WJ]

The dip is related to the bump. The bump is mild and wide on the measurement on an infinite baffle, but it becomes very noticeable by baffle diffraction.

Usually, a dip around 1000Hz may rather be offset by a diffraction effect. But in the case of the 8945, the dip is at 1.2kHz and exaggerated by a diffraction bump around 900Hz to 1000Hz---it is more so in the P version. And this is why Dennis Murphy and A Feyz P used a notch filter to reduce the effect for their 8945W and 8945P 2-way's.

My approach will flatten out this bump/dip. But it requires a very low c/o point.

[sreten]

Originally posted by Jay_WJ


My initial design requires a 1.5 to 1.6 kHz c/o point with about
-18dB/oct acoustic slope (NOT electrical 3rd order) for the tweeter.
Can the HDS tweeter or the Seas be used in this case without a problem?

Hi,

I'd say yes, given that they are also attenuated relatively a lot.
The midband efficiency with BSC of that bass driver should be
low, so the excursion requirements on the tweeter are lower.

The guy blowing tweeters referred to in this thread simply
has no sensible idea of how to correctly test tweeters.

Personally I'd use a Peerless HDS driver (with Seas tweeter), great
value combo, instead of trying to work round the Ushers problems.

/sreten.