2-way with 8531G Revelator?

[dmitriy167]

Hi,

I've abandoned my previous idea for high efficiency, since I realized that my money might be better spent on something of higher quality. The Scan Speak 7" Revelator (sliced cone) driver might be a good idea, especially if I can get a used pair from somewhere since it's very expensive. I haven't seen that driver in any DIY 2-ways (I have seen the Sonus Faber speakers), so I was wondering if anybody here used it in that arrangement.

Thanks,
Dmitriy

[Mingo]

Have a look at Troels Gravesen's homepage and find SP95 (9500 tweeter) and SP98 (further down the page, I think) (9800 tweeter).

[kwng11]

mine is 8530K00, bass is very good, mid is quite clear, i think 8531G will be very similar.

mine is only 85.5db in sensitivity, but my Jolida 1701 (50W) can push them very well, so lower sensitivity of this woofer is not any issue to me.

The most i concern is how to get the best crossover to dig out all the goodness of the driver.

let's sharing ideas together~

[imjatse]

If you didn't see it Zaph's newest design uses the smaller version of the Revelator:

Zaph's Site

And the local diyaudio thread on Zaph's design:

here

[rabbitz]

Recently used the 18W/8531G00 with a S2905/970000 tweeter and a great sounding speaker... my best to date and are my mains. Very critical of the components used in the crossover as well as the partnering equipment.

The inverted drivers were to suit a 1st order crossover but didn't work so is now basically a 2nd order. I don't think the inverted arrangement is having a detrimental effect with 2nd order as it's sounds brilliant.

[alan-1-b]

Hi Rabbitz ,
that ScanSpeak woofer & tweeter combination is basically what I am planning to try !
What nominal cross-over frequency did you attempt the 1st Order slopes ?
I was thinking of attempting it at 2kHz , but am concerned about whether I'll acheive sufficient roll-off to adequately reduce the audibilty of the 18W's break-up resonance which seems to be a little above 3kHz.
Any comments would be appreciated , thankyou ,

[rabbitz]

I kicked off at 2375Hz and the peak was very noticeable. I then tried several others and even had severe underlaps on the crossover where the woofer and tweeter crossed over at different frequencies up to about 1000Hz. There were over 20 crossovers tried and tested.

None were successful and eventually nailed it with the AR series crossover which gives a steeper roll off on the woofer. What I was hearing was confirmed by Jeff Bagby's series spreadsheet which is invaluable. I'm sure you could use contour networks on the woofer for 1st order but all the others I've seen use a 2nd order as a minimum on the woofer and a good example is the one by Tony G.

Cheers

[alan-1-b]

Thanks rabbitz , and I've just now seen your reply to me in the other Scan thread -{coated versus uncoated 15W's} on basically the same question of mine , and I've replied there-under .

Further , on what you describe above , I realize there is more to this particular woofer than I had thought - 20 trial cross-overs !
I will look at the AR series crossover ; Jeff Bagby ; Tony G ; that you mention .
Really I do not like the audible effects of systems where one driver needs to be connected in reverse polarity to the other , so perhaps I'll have to go to a 4th order crossover if I decide to attempt with the 18W8531 , but even more complexity and I had wanted to keep this one simple !

I wish Scan still had the 8530 - coated cone option - available , that one of the contributors to the other thread said he has , as that would at least have less audible break-up around the 3k2 area . Perhaps I'll have to apply damping to the 18W cones myself ...

I see you have the 18S - sheilded version - for sale on your site , but other than its low frequency behavior being different , I suppose its midrange behavior will be close to that of the 18W - Scan's own data sheet for 18S showed very little difference there to theirs for 18W .

Thankyou again for your comments !

[tktran303]

In a 3 way you likely not need to worry about it, but at the typical crossover frequency for a 2 way (1.5-3Khz) you'll need a notch filter to shape the response to get rid of that 3.2Khz bump, which is clearly visible on both the FR and CSD.

For ideas, take a look at Dan Wesnor's Ella project, and John Krutke's ZD5 project for the kind of low linear and non-linear midrange distortion these these drivers are capable of.

I am impressed with the perserverence. 20 filters? Applying damping?

I think you guys deserve to treat yourself with some measurement equipment, CAD crossover software and some textbooks...

[alan-1-b]

Thanks for comments tktran303 .
Textbooks I have access to , and if I had the money to afford all the test equipment , etc ... , that I'd like , then I'd be tempted to spend it on good quality manufactured speakers - perhaps from B&W 800 series - but then I might not be completely happy with those and decide to modify them !
What I really need is sufficient time available to do sufficient measurements and constructing , but time cannot be bought !! - as I have no doubt many of the contributors to this forum know too well ...