Designing a Crossover for High End Speaker (Raal 70-20xr + Audiotechnology Flexunit and C-Quenze)

Designing a Crossover for High End Speaker (Raal 70-20xr + Audiotechnology Flexunit and C-Quenze)

Hi.
For the last couple of years, I have been making my ultimate speaker project. I, in my opinion, have gathered the best parts for that project, where i have been helped by Troels Gravesen and the makers of both Raal and AudioTechnology. I am pretty much done with my project.

However i hit one of the last hurdles on the way. The crossover. I have the measurement for my drivers, in my baffle, and wanted to design the crossover in xSim etc. My only problem is that im not experienced enough, to make the fullest of these wonderful parts. I've uploaded my beginnings of trying to make a crossover, but as you might be able to see, its a mess. I know i can make it better myself, but i wanted guidance first. My question to you guys is how i minimize the baffle refraction on the 18h52 and make the transitions smooth? - The cabinet is set, and is a homage to my uncle, it is already build. What im wondering is how you guys would do it, is there better ways to smooth out etc.

I hope some of you have some ideas or know where i can seek more guidance.

The speaker consist of the following parts:
Raal 70-20XR-NC
Audiotechnology C-Quenze 18H52 06 13 SDKAM-LR
Audiotechnology Flexunits 12 B 77 25 10 KAP

Best regards, Mads
 

Attachments

Hi Mads
where do you live in denmark i live in western Denmark
best Bjarne
Hi Bjarne
I live in Ribe.
- Mads

Mads,

are you positive your driver measurements are spot on, in every respect? (near field+far field summation). Zma files of the AT units seem like as if they were done free air.
Mhm, the subwoofers are probably free air, as i had to use the measurements given from AT. Troels Gravesen measured my midtones and Raals with my baffle, so that should be exact.
 
you have a PM
Answered, thank you
Out of interest, did you ask Troels how much for him to design your XO?
Yeah, that was the plan. I was his last client, but my money fell short to keep up our schedule, so thats why we didnt get further.
Hello Mads,

did you have to pay for Troels' work on these measurements?

edit: I wonder why isn't there anymore official FR data of Audio Technology units.
Yeah, i paid him some money, but we never finished it off. He told me i was his last client. Regarding AT units, they updated their site. For some reason, they didnt include the data again. I once talked with the founder, who told me it would be updated some time, as far as i understood.
 
Hi @AlfonsDenmark - there are many more capable here than myself, but I still wanted to give you a take on a simulation - perhaps it can give you some ideas. I don't have XSim so here is the schematic.


atraal.jpg


Notes:
It's LR4 acoustic with XO at 200hz and 2400hz, SPL at 92db!
The RLC on the mid can probably be omitted, but I included it so you can see what it does.
I tried LR2 on the bass but there is a hump between 50-100hz that can't be flattened without a humungous RLC which also then causes a dip around 100hz right before the filter starts working, so I gave up on that.
I found that LR4 @ 92db levels this all out such that the hump + filter together match quite closely to the textbook curve from 50hz.
I also tried to ensure the filters track the target curves to -50db down from reference level; maybe that is unnecessary. It is easier to do that with LR4, though.
Looked at distortion measurements for the mid on hificompass and saw 200hz is quite OK for this driver. Maybe 300hz would be good, too.
Hificompass also recommended taking it up to 2.4khz, so I did that, which also gives more comfort to the ribbon.
The mid has a little response dip around 1k from the original measurements. Don't think much can be done about that, or if it's a big deal.

As the 92db level is so high, I wondered about attenuation and as you can see there is none on the mid (which should be a good thing) and a tiny amount on the tweeter, but both mid and tweeter can handle this in theory. Perhaps the system db should be set for 91db or such, and give you more adjustment options to balance the sound in the room.

Sim shows nice phase alignment with your provided measurements and this network, and impedance is healthy.
__

You have such nice drivers and obviously this project means a lot to you, I'd recommend getting more help locally if possible, or this thread can continue with more help from others. Another thought is to get a Behringer active XO and play around to find the XO points that sound best to you. I just selected them here as a starting point. The ribbon is meant to be able to go lower, and everyone says they always prefer second order slopes. But how it all sounds without testing and tweaking, who can say.
 
Hi @AlfonsDenmark - there are many more capable here than myself, but I still wanted to give you a take on a simulation - perhaps it can give you some ideas. I don't have XSim so here is the schematic.


View attachment 1115012

Notes:
It's LR4 acoustic with XO at 200hz and 2400hz, SPL at 92db!
The RLC on the mid can probably be omitted, but I included it so you can see what it does.
I tried LR2 on the bass but there is a hump between 50-100hz that can't be flattened without a humungous RLC which also then causes a dip around 100hz right before the filter starts working, so I gave up on that.
I found that LR4 @ 92db levels this all out such that the hump + filter together match quite closely to the textbook curve from 50hz.
I also tried to ensure the filters track the target curves to -50db down from reference level; maybe that is unnecessary. It is easier to do that with LR4, though.
Looked at distortion measurements for the mid on hificompass and saw 200hz is quite OK for this driver. Maybe 300hz would be good, too.
Hificompass also recommended taking it up to 2.4khz, so I did that, which also gives more comfort to the ribbon.
The mid has a little response dip around 1k from the original measurements. Don't think much can be done about that, or if it's a big deal.

As the 92db level is so high, I wondered about attenuation and as you can see there is none on the mid (which should be a good thing) and a tiny amount on the tweeter, but both mid and tweeter can handle this in theory. Perhaps the system db should be set for 91db or such, and give you more adjustment options to balance the sound in the room.

Sim shows nice phase alignment with your provided measurements and this network, and impedance is healthy.
__

You have such nice drivers and obviously this project means a lot to you, I'd recommend getting more help locally if possible, or this thread can continue with more help from others. Another thought is to get a Behringer active XO and play around to find the XO points that sound best to you. I just selected them here as a starting point. The ribbon is meant to be able to go lower, and everyone says they always prefer second order slopes. But how it all sounds without testing and tweaking, who can say.
Thanks for your contribution. I will definitely try to tinker around with your suggestion.
 
Be careful with complicated networks, especially when there is unknowns regarding measurements. Drive units offsets (x,y,z)?
You won't need 2.2mH in tweeter circuit. Use older B&W service manuals as a starting point for a x/o network.
 
Unluckilly the datasheets in Vituix can not say the real sensivity of the drivers. For illustration this Audiotechnology is a 88 dB/2.83V in real life. But that is not hard to adapt in the XO with resistors...It makes just a little harder to choose the bass unit to match the mid unit at baffle step.
 
Be careful with complicated networks, especially when there is unknowns regarding measurements. Drive units offsets (x,y,z)?
You won't need 2.2mH in tweeter circuit. Use older B&W service manuals as a starting point for a x/o network.
Yeah. I want to, as far as possible, to keep the parts as few as possible. Especially for mid and tweeter. I might be buying the Hypex FA501 for each sub, to have active crossover.
Unluckilly the datasheets in Vituix can not say the real sensivity of the drivers. For illustration this Audiotechnology is a 88 dB/2.83V in real life. But that is not hard to adapt in the XO with resistors...It makes just a little harder to choose the bass unit to match the mid unit at baffle step.
Yeah, i might cross bass active, to adjust it to the room etc.