Specific 2-way crossover question (long)

[aabill]

Hi

Background:

I'm building a pair of speakers to use with a good audio set-up. Enclosures will be relatively small, to be positioned either side of the audio stack which has television on top. Stereo only - I'm currently not interested in surround-sound. No sub either.

After a *lot* of poring over specs, and reading what people have to say, I've decided (won't bore you with the detail of the journey) on a 2-way using Monacor SPH-165KEP. I chose the LF driver first, then looked at tweeters and the Hiquphon OW1 really just chose itself (LDSG helped) - extended ruler flat response, minimal off-axis variation, sensitivity almost an exact match to the bass unit, etc, etc. Listening levels will usually be quite low so the relatively low sensitivity of these two units won't matter. I'd be interested in confirmation (or otherwise) that this is a good, sensible pairing.

The question ( at last - )

Because these units have a fairly smooth extended response in the crossover area I'm having difficulty deciding exactly what crossover frequency(ies) to use so I'd appreciate your thoughts on what to use as a starting point for what I expect to be a trial and error exercise. I think there's no reason to use more than 1st order slopes - would you disagree? why? and are there any other particular issues I might have to watch out for?

Naturally I'd be most interested to hear from anyone who's used this LF drivers with the OW1 (or OW11/111)

As a measure of me, FWIW and not to become a focus of this thread, I'm one of those who tend to rely on measurement first and ears second, and go round the measure-change-listen cycle as often as it takes to feel satisfied with the result. On the other hand, I am NOT one of those who claim to hear any difference, or who believes there *is* any audible or measurable difference, between the sounds of two different manufacturers' high quality speaker cables -

TVMIA

Bill

 

[bogicp]

Hi

Background:

I'm building a pair of speakers to use with a good audio set-up. Enclosures will be relatively small, to be positioned either side of the audio stack which has television on top. Stereo only - I'm currently not interested in surround-sound. No sub either.

Ok, your target point is clear and reachable... two small monitors.

After a *lot* of poring over specs, and reading what people have to say, I've decided (won't bore you with the detail of the journey) on a 2-way using Monacor SPH-165KEP.

If you already buy it, ok, no problem, but if not...
try to understand problems about stored enegy in drivers, because manufacturers don't give this measurements in public.
Siegfried Linkwitz do these tests and you can view their results at
Linkwitz Midrange Tests
If you need shaped tone bursts you may use BoZo Audio Test signal , and any sound editor like cool edit or whatever....

I chose the LF driver first, then looked at tweeters and the Hiquphon OW1 really just chose itself (LDSG helped) - extended ruler flat response, minimal off-axis variation, sensitivity almost an exact match to the bass unit, etc, etc. Listening levels will usually be quite low so the relatively low sensitivity of these two units won't matter. I'd be interested in confirmation (or otherwise) that this is a good, sensible pairing.

Sorry, I'm never use this drivers.

The question ( at last - )

Because these units have a fairly smooth extended response in the crossover area I'm having difficulty deciding exactly what crossover frequency(ies) to use so I'd appreciate your thoughts on what to use as a starting point for what I expect to be a trial and error exercise.

Good starting point is a mid point of overlaping region.
If you have high pass and low pass (-3dB) frequencies for tweeter and woofer (respectively), you may use fstart=sqrt(fmid*ftweeter)
(sqrt=square root).

I think there's no reason to use more than 1st order slopes - would you disagree? why? and are there any other particular issues I might have to watch out for?

1. Yes, I disagree.
2. Because of possibility to your listening levels cannot be always small.
3. Over-excursion of tweeter and greater distortion. There isn't much needs for power to overload any driver below resonant frequency, and distort sound at all frequencies near to crossover frequency.
3a. You cannot imagine crossover filter like switch (on/off), this is probably possible only with digital brick-wall filters (200dB/dec or more), and in real usage you must expect that drivers overlap each other even with steep filters, but with first order filters, you expect that tweeter can do their job at 400-500Hz,... this is hard to obtain, even with small power.

Naturally I'd be most interested to hear from anyone who's used this LF drivers with the OW1 (or OW11/111)

As a measure of me, FWIW and not to become a focus of this thread, I'm one of those who tend to rely on measurement first and ears second, and go round the measure-change-listen cycle as often as it takes to feel satisfied with the result. On the other hand, I am NOT one of those who claim to hear any difference, or who believes there *is* any audible or measurable difference, between the sounds of two different manufacturers' high quality speaker cables -

TVMIA

Bill

If you like measurements and if you like to hear what you measure try to use (linked above) BoZo Audio Test signal, It's fundamentaly based on wavelets, and there isn't much more measurement programs, that work in this principle. They still use wave based transformations or impulse principle, not wavelets (wavelets=small waves)

Best regards,

-boggy

 

[Emiel]

The SPH-165KEP seems to be usable for producing frequencies up to 7KHz. Also it shows a natural roll-off of about 6dB/oct above aproximately 7000Hz. Maybe it's an idea to cross this high to leave the xo frequency out of the voice range.
I have no experience with either of these drivers, but it seems to me that you could make use of a 6dB/oct filter at about 7000Hz for the Monacor, which combined with the natural roll-off would make 12dB/oct.
The tweeter could then be crossed with a 12dB/oct filter at 7000Hz, which would leave you with -42dB at aproximately 900 Hz for the tweeter, which should not be a problem.

Good luck with the design!!

Kind regards, Emiel

 

[aabill]

Emiel

Thanks for taking the trouble to reply, but I don't think what you suggest would be satisfactory. I haven't come across any off-axis plots for this bass driver (if anyone has I'd love to see them), but I'd expect off-axis performance with a mid-bass at those frequencies to be unacceptable, especially given the reputedly tremendous off-axis performance of the Hiquphons

But what an interesting idea nevertheless!!

Kind regards
Bill

 

[aabill]

Boggy

Thanks for your careful responses, which will feed into my thinking, particularly re the need for a steep enough crossover attenuation to keep out of the tweeter Fs. Unfortunately Mr Linkwitz hasn't tested the bass-mid I intend using, but I'm encouraged by the fact that nonody seems to have anything other than unreserved praise for it (other than mention of a small curable problem with an air leak around the centre plug). I don't think I can go too far wrong with it.

Best regards
Bill

 

[carlosfm]

...but I'd expect off-axis performance with a mid-bass at those frequencies to be unacceptable...

That's right, most midbass drivers have bad off-axis response above ~4Khz.

Thanks for your careful responses, which will feed into my thinking, particularly re the need for a steep enough crossover attenuation to keep out of the tweeter Fs.

Go for second-order Bessel, and cross at around 3.5K.
If you can't use first-order, Bessel second-order is the next best thing, at least for my ears.