Alpair 7 and SDX10 WAW / FAST.

[Wiggle8]

I've got a new FAST arrangement I am working on here. The full range driver is a Mark Audio Alpair 7 in a sealed 5 liter enclosure. The SDX10 is in a 48 liter rear ported enclosure tuned to approx 22 Hz. The boxes still need touching up but I have been gently breaking in and testing the Alpairs. Very impressed with the beautiful midrange on these.

I will be reconfiguring my PC to act as an active crossover during the testing. I am envisioning a crossover of approximately 225 Hz, 2nd order LP on the SDX and 1st order HP on the Alpair to work with the natural roll off of the driver in the sealed enclosure.

Updates to come as I configure and test the crossover.

 

[SlainNothing]

very nice!


i suggest trying 333hz xo; although using sdx10 as midbass will need significantly more power

 

[Jerms]

My concern would be that the subs won't play nice higher up. Whether that's a relevant concern I don't know. There is no advised upper Hz limit on the CSS page I grabbed details from.

Best of luck with the PC crossover. Will you run bafflestep filters for both HF and LF too? As you have different size enclosures for them.

 

[Wiggle8]

My concern was also that I wanted to avoid crossing over too high to avoid the vocal range and avoid going too high for the subs to handle gracefully. I had seen Jeff Bagby use the SDX10 crossed over at 200Hz with good results so I feel comfortable closer to that range but will also experiment higher.

For the baffle step, I am still not clear if the Alpair 7 is baffle step compensated for a particular baffle width. I have run baffle step simulations for both the SDX and the Alpair. I think I am crossed over low enough that the BSC will not be required on the SDX10.

 

[Wiggle8]

Got the last edge tape on and a good amount of the staining done today. Tomorrow I hope to get the other bass unit stained and then soon move onto the crossover.

 

[xrk971]

Perfect setup to try Harsch XO. You can move too back and forth for time delay. 4th order woofer lowpass, 1st order electric highpass will give second order acoustic highpass on top.

http://www.diyaudio.com/forums/multi-way/277691-s-harsch-xo.html

 

[Wiggle8]

What is the effect of combining the drastically different crossover slopes? It seems that the response curve would be not ideal.

 

[xrk971]

The 4th order and 2nd order with positive magnitude phase and delay give an approximate quasi transient perfect behavior and nearly flat phase. You get the nice steep filter to cut out woofer peaks in HF and still get similar response temporally as a all first order XO.

 

[Wiggle8]

Interesting. And there is also a time delay on the full range driver correct?

 

[xrk971]

Yes, assuming a flat baffle and zero acoustic offset between drivers, the full range or "tweeter" is delayed by half the time period of one cycle at the xo frequency. You know you did it right when the bass response of the combined signal dips a little below the bass only. There is a 1dB wiggle but that is nothing at 300Hz and what you get is time aligned absolute phase.

 

[Wiggle8]

Got it fired up ☺️. Haven't tried the harsch yet but will tomorrow. Just have it set up through the PC on eqapo, FIR filters convolution at 225Hz, just first order on HP and 2nd order on LP. I am in a bad room and listening too close but can already sense a lot of promise.

 

[xrk971]

If you have FIR convolution you can have a transient perfect response without Harsch. Flip Alpair upside down so CTC distance between drivers is closer for better realism.

 

[Wiggle8]

I will stick with FIR for now then, but may end up using the harsch if I switch to the mini DSP.

The SDX10 seems just fine playing up in the upper bass and lower mids. I am impressed by the added richness that the SDX brings to the sound.

 

[Wiggle8]

After a bit of tuning I am liking about a 4 dB boost at 250 and 400Hz, accomplished using parametric EQ in rephase. Pulls a slight bit of presence but smooths out the bottom end and adds some richness.

 

[Mayuri]

I was planning a similar build as yours but decided do go for the flat enclosure instead. If you are interested, I recommend you try to add an vibration dampening layer between the enclosures. Maybe one of those rubber slide stoppers you put under carpets, just cut to shape?

 

[SlainNothing]

What are the specs of the amplifiers?

I see small class d at the top? and custom class d for the sdx10?
Generally, by ear, class d needs much more dsp than class a/b (poor damping factor on class d)

 

[Wiggle8]

Dayton Dta120 for the Alpairs and L15D Pro for the SDX10.

 

[SlainNothing]

class d (class t) on xbl^2 has always added to the hallowness and thin sound (but very crisp and dynamic). i can see why the +4db adds much warmth/rich

imo class d on any of xbl speakers didnt "sound right"

koda10
tempest 15
extremis 6.5
chr70
aplair7
AA 6.5 mids
exodus 10
exodus anarachy 6.5

 

[Wiggle8]

Interesting, I did drop that down to a 2dB boost at those frequencies, seemed like the best balance. I wonder how much that effect would matter here where the XBL driver is only handling the lowest part of the frequency range. I do have a Dayton APA150 I could test in there too.