Seas U18RNX/P + Scanspeak D2905/97000

Status
Not open for further replies.
Jump to Latest
C

cucicu

Member
Joined 2011
#1
HI everybody,


I have realized a 2 way speaker based on seas Idunn, same box, same woofer but different tweeter.

Unfortunatelly I can't make any measurement, but I imported the frd and zma files for the woofer.

For the d2905 I have the measured files.

I tried to realize the system, maintaining the same crossover of idunn for the woofer, but it seems too forward on the mid side.

So I tried to make some simulations, and i found that that the woofer requires a bigger BSC (3,3mH) to compensate the baffle step and get alligned withe the bass zone.

this is the sim, I would like to know what you people are thinking about, thre 3.3mH BSC seems too large.

regards


idunn10.jpg

xover10.jpg
 

Attachments

#2
The SS tweeter and the Seas DXT are totally different animals, so that i would rather try to rework the filter on the tweeter side than the woofer side.

And yes 3.3mh is way too big, suggesting you need to atenuate the mids. Seems to be more straightforward to remove part of the attenuation on the tweeter...
 
Last edited:
#3
I tried with 2.2mh/10uf on woofer like seas idunn but the tweeter is very difficult.

2nd order 6.8uf/0.30mh damped with 5 ohm series sounds too midy; reducing the cap to 5.6uf there is a dip and also is not good with 3 order. Dont know what to do
 
#5
Hi Cucicu,

I'm a little worried about the final tweeter tilt. I think 3rd order may work well for you, but the issue I see is the tilt of the tweeter. I don't usually want it going upwards like that. I would make 100% sure that is your desired tweeter tilt before worrying about the woofer.

Best,

E
 
#6
I seriously doubt that the frd and zma files you are using a any raliable, and what you hear has probably nothing in common with your simulation.

It's the problem with simulating vs measuring: simulating is not a short path to avoid measurments.

If you don't measure, better build a kit.😉
 
#8
The phase difference changes with frequency; it's 45 degrees at the crossover point and increases to 90 degrees as the frequency increases. I attached the phase plot of an example speaker, one with a 1st order and the second with a 2nd order, both at 100 Hz.
 

Attachments

  • example.JPG
    example.JPG
    94.1 KB · Views: 425
#9
Oh okay taking the continuously varying impedance of the drivers you can calculate the phase difference. Crossovers though can only offer orthogonal phase shifts. But blended with the drivers' impedances you get the 45.
 
Last edited:
#10
Yes, the tweeter has a rising response. Maybe it needs a zobel network?

I agree also about 3rd order on tweeter. Maybe would be better a 2nd order.

I was looking at the zr71 zaph design wich uses the er18rnx woofer (quite similar to the u18) and he uses a 3.3mh/12uf.
 
#12
The tweeter is not diffucult to flat. Troels always use for this tweeter an LR in series to compensate the rising inductance of the VC but this is not the issue. It is not tremendous the bump over 10khz. Also a zobel works good but these are refinement
..
 
#20
GringoAudio said:
I can't find Frd files for drivers I have either. No need to be snotty. I got a rudeness warning from the admin here for using an exclamation mark here once so be careful.😉
Click to expand...

There are no usefull frd nor zma anyware but in YOUR computer after you have performed YOUR own measurements.

All the rest is messing with wrong simulations based on useless data..:nownow:
 
Status
Not open for further replies.
Top Bottom