But don't use TOO high slopes for the high pass. High slope filters at that low frequencies will give tremendous group delay....

Yes...

You have to remember that every 6db of crossover you add moves the phase back 90 degrees...  If you have say 24 high pass on your subs and an 18db low pass on your bass, they will be 90 degrees out of phase at the crossover frequency...  You can correct for this on an LMS though...

You can correct the 'phase pushing' electronically, but you can't undo the delay the crossover introduced.

Asymmetrical filters are 'no good'. Filters should always be acoustically symmetrical. If not you can't get flat response around the crossover frequency. Symmetrical means that if you have a 24dB LR highpass for your mids you should have an according 24dB LR lowpass for your subs. If your top does a 12dB/oct. roll off at 120Hz and you want to cross over at 120Hz it would be a good idea to use a 120Hz 12dB Butterworth highpass for that top and a 120Hz 24dB LR lowpass for the subs (just as a starting point).

You can of course set a 12dB hp at 30Hz and a say 24dB lowpass at 80Hz for your subs. The lowpass and the highpass for the same band do not need to be of the same slope and chracteristic. Crossing over below 80Hz to another sub, like X1 from 30 to 60Hz and HD15 from 60Hz up I also would recommend to stick to lower slopes like 18dB or even 12dB. 12dB LR could be a good thing then.  Every filter will introduce more delay..... the higher the slope, the more....

Ahhhhh! I just typed in this post and it lost it all and told me I needed to log in! I'll try again...

My planned system includes 2 X1's and 2 X15's per side. I am currently thinking about the crossovers. I am planning to build my own crossovers rather than buy a controller.

If I have an 80Hz lowpass and a 30Hz highpass filter for the X1's, a 2kHz lowpass and a 80Hz highpass filter for the SN 15 MB, all at 12db/oct and a 2kHz highpass filter for the SD 740 N at 24db/oct, will all the drivers be in phase and have the same delay?

matttcattt wrote: If I have an 80Hz lowpass and a 30Hz highpass filter for the X1's, a 2kHz lowpass and a 80Hz highpass filter for the SN 15 MB, all at 12db/oct and a 2kHz highpass filter for the SD 740 N at 24db/oct, will all the drivers be in phase and have the same delay?

No is the simple answer....

You will get a delay around the ports on the X1.  You will need to delay the SN15MB to get it in time at the crossover point.

You will then need to delay the SD 740N by the same amount +/- the distance behind/in front the compression driver is than the mid cone.  Obviously this will depend on which horn you've used...

Because you've used a 12db low pass on the mids and a 24db highpass on the tops, they will be 180 degrees out of phase at the crossover point.

Best bet is either use some test tones at your crossover points and set the delays by ear or download smaart from siasoft and let your computer do all the hard work for you...

How much will I need to delay the SN15MB's?

I havent actually built anything yet, but I am planning on using the horn.

How do I work out how much to delay the SD 740N's by?

I havent actuall used anything yet. Wont the two 12db/oct filters in series have the same effect as the one 24/oct filter?

@ Timber_MG: That's why I said 'have to be ACOUSTICALLY symmetrical'. That does not necessarily mean 'electronically symmetrical and at the same frequency'....

@ Matttcattt: Even if you leave out the group delay each speaker produces, the answer would be no. That's because you have a highpass and a lowpass on each band (except the HF output which doesn't need a lowpass), but the delay caused by the highpass for the subs won't affect the MF output. So if you have 15ms group delay for the subs highpass and 8ms for the subs lowpass this adds up to 23ms, but a symmetrical highpass for the MF would only show the 8ms......

The only REALLY WORKING way would be to measure the delays. Calculating those values is too complicated by far.... But at least in the bass range a good guess can do better than running without setting delays at all....

I use a anayliser that plots the GD. You can also do it with Smaart.

Put in a tone and then click on the calculate distance button. This then gives you how far and hence the time it is from you.

"GD"?

Wow, that Smaart is expensive software. Is it worth it?

You can download the trial and use it for 30 days and decide yourself if it's worth it...

And if it takes you longer than 30 days to measure your delays then you've got problems...

Dom wrote: And if it takes you longer than 30 days to measure your delays then you've got problems...

Or a really bad BP6 or higher.... :)

It would be a good idea to consider 'filtering options' BEFORE you get your hands dirty.... Just think about the speakers you're going to use. If it's a 15" and a small 1" HF you know for example that you need to crossover the 1" very high (2kHz and above) and that the 15" won't play good that high.... so you know this wouldn't be a good box.

Or you want to build a 15" 3way cab. So you get an 8" for the midrange, now you have to check what's the lowest frequency the 8" can play well.

I don't mean that you should have the crossover ready before the speaker. I just think that you should know what you want first....

But the X15 is a 2 way design... A 15" will go up to 2k wont it?

I realise that I need to plan.

I don't think that a 15" speaker should go higher than 1200Hz. Even that is in many cases too much if you want clear mids. It's plain physics. A cone with voice coil has a certain weight. This weight will act like any accelerated mass. The bigger the cone and the VC, the higher the mass. And the higher the frequency, the higher the acceleration. So high frequencies AND big cone add up to 'not that good at all'.

It is NOT about 'mids are weak in SPL' or 'the freq plot isn't flat', it's more a question of details in sound.

I understand the problem, and I agree, a three way design, or smaller bass driver would be better, but would Rog design a speaker that wouldnt work well?

I think the HF driver has a highpass at 2kHz, but the 15" somewhere around 1100 or 1200Hz or such. This is because the HF driver has a big hump around 2kHz and below and the acoustic crossover frequency is much lower than 2kHz.

Oh.

If the crossover frequencies are set for the designs im using, i dont have to do much planning really.

Of course, if you build already existing designs, they will perform like intended.

In my case most existing deisgns won't fit my needs.....

This will indeed get us bandpass characteristics. Of course higher freqs will 'escape' because the 'hole' is directly in front of the speaker... but I guess the low mids (100-250Hz) won't be good....

It's not only beaming I was talking about. It's more the fact that the mids, even if the freq plot is smooth, won't be that detailed. I have some really smooth sounding 15/2 type cabs, but it isn't comparable to 10" and smaller mids.....