Visaton al170 with vifa xt25BG (help please)

[Shirokuma]

Total Beginner here (still).

I used boxsim and winisd and created a pair of floorstanders 2-way with following specs:
woofer: visaton al170
tweeter: vifa xt25bg
dimensions: 100cm x 30cm x 20cm, 18mm mdf, braced, internal volume of 41.727 L
port size: 16.4cm x 1 cm x 12.4 cm

I added my boxsim files, a picture from the crossover and from the speakers (not in its listening position). Amps used
marantz nr 1603 (2x50w)
Lab Gruppen 1000 (2x 350w)
Source: Pro-ject Debut Carbon Esprit

When finally firing up these babies I could have cried:
They sounded muffled: a wooly sound, as if somebody put some blankets over them. I also had the impression it started ok but they got 'woolier' especially compared to what i had before (own design, same dimensions visaton al170 with dsm25 ffl).
I also had the impression they sounded better with the more powerfull amp.
They also are not very sensitive, I have to crank up the amp (first one) quite hard.

Can anybody help me please? Some ideas?
I know I'm not an expert, but I 'm here to learn.

Thanks already, I hope one day I'll be experienced enough to help others on this forum.

 

[LightBit]

xt25bg seems to lack frequency response and phase data in your simulation. It is therefore simulated which might give you different result. See attached screenshot.

Other possibility is that simulation overestimates baffle step. Reducing tweeter attenuation will probably help.

 

[Lojzek]

Shirokuma, for starters flip the polarity on either tweeter or woofer, both enclosures, then report back what happened.

edit: Boxsim was using frd and zma of XT25BG which can be seen at driver tab in boxsim, just before frequency response electr.

I have simulated in XSim and the polarity does look to be wrong when in phase and the raw driver XT25 spl seems wrong also being 95 dB, while it is in fact around 92,5 db/2,83V/1m.

 

[LightBit]

Black line is too perfect (simulated).

 

[Lojzek]

It is but the issue extends beyond that. This is a quick sim with XT 25 and original filter, polarity in phase, which may sound as described by Shirokuma.

 

[Shirokuma]

I already tried switching polarity but the effect remained the same. Perhaps I made a mistake entering the specs from the xt25bg?

 

[Lojzek]

You have entered the TS specs from the manufacurer spec sheet but the software recalculates that to about 95 dB/2,83V sensitivity which is unrealistically high. See the original pdf messurement that specifies graph to be measured at 2,83V/1m having about 92,5 dB. So you should at least decrease padding for a couple of dB to improve the HF spl. Increase the 3R resistor to a larger value like 6R8 and then it will be better.

edit: Try this TW filter mod. 4R7; 4u7; 0.33mH; 8u2

 

[Shirokuma]

Excellent I'll try that out. I really like the al170 and heard nothing but good about the xt25. I thought i'ld give the combination a try. The combination is also pleasing to the eyes, to me at least.

 

[Shirokuma]

You mean like this then?

Thank you. I'll try this out asap.

 

[Lojzek]

No, first 4R7 resistor, then the HP 3rd order electrical.

 

[Shirokuma]

Ok, my bad. I'll try this. Thank you.

 

[system7]

Lojzek is a terrifically safe pair of hands when it comes to simulation.

I like the idea of a third order 4.7uf/0.33mH/8.2uF filter on the tweeter with 4.7R in front to adjust level.

I imported this similar VIB 170AL bpj project into the projekte folder in Visaton Boxsim simulator.

I then used a 4 ohm tweeter, the mylar Visaton DT94-4, to guess how your Vifa XT25 will work. There are two versions of the XT-25. One is the 30 at about 90dB. And the double-magnet 60 at about 92dB.

I tried to reuse your existing components here. Lojzek is on the money that the tweeter is too quiet. I also thought that your 6kHz notch is too deep on the Visaton AL170 metal bass, and preferred a 22R resistor if only for phase alignment.

I think there is a case for adding the 4.7R to the 3 ohm tweeter to get the classic bell-shaped impedance curve which amps like. Because otherwise, a 4 ohm tweeter leads you to using much smaller coils and bigger capacitors like the 0.15mH you find in B&W designs.

Hope it helps anyway. It's fairly easy to raise tweeter level by reducing the input resistor, and quite crucial to get it about right.

 

[Lojzek]

Steve, I am probably worse than you have described me, but I thank you anyway.

 

[Shirokuma]

I have a feeling that, the more one gets in to this hobby, the more one realises how hard it is. At first i thought, oh is that it? High pass filter, low pass filter, speaker: done. Now i'm like Homer simpson: ''what have i done? What am i doing? Wat will i do?'' thank god for this forum and the kind and clever minds on it.

 

[Mario Pankov]

The XT25 won`t make the best partner to the AL170, its a darker sounding tweeter better combined with smooth paper or poly cones. Visaton have a waveguide that you can use and solve most issues, the flat ring radiator will load it very well. But, this is not solving your problem. If you remove the tweeter and connect it through some cap, does it sound ok? To me it sounds as if the tweeter is not connected properly or you have too much attenuation on it (resistor network).

 

[Shirokuma]

AH, the plot thickens..... I used it first with the dsm25ffl which sounded ok, but my cabinetwork sucked sod i thought i'ld give them an upgrade

 

[tuxedocivic]

If you need actual measured data for the XT25BG60 on that size of baffle I have it available for download. I also have a testing video showing off axis and other info to assist in design. Let me know.

 

[Lojzek]

This is the cause of miscalculation

I have spent some time figuring out what happened in the process while Shirokuma got Boxsim to have calculated wrong the tweeter sensitivity and I am positive that Scan Speak has a typo in specs with Fs=589 Hz while it should have been Fs=489 Hz.


Re=3 R; Fs=489 Hz; Vas=0.02 litres; Qes=0.43

no=9.613*(10^(-10))*Fs^(3)*Vas/Qes=0.005228


SPL(1W) = 112.2 + 10*log(no)=89.43 dB


SPL(2,83V)= SPL(1W)+10*log(8/Re)=89.43+4.259=93.69 dB

Try with Fs=589 Hz and you get the Boxsim's SPL(2.83V)= 96 dB

 

[boswald]

"Question Everything". A good method, was a popular bumper sticker fourty and fifty years ago.
Sometimes, when all else fails, read the instructions; if that fails, question the instructions.

 

[Stixx]

I didn:t read every line of the thread but have a similar build in BoxSim...
and I replaced the tweeter with measurements of a similar driver (XT300.. might be older).

Hint: Have a look at his attenuation pad before the tweeter... 3 Ohm in // is awfully low andf pulls the tweeter to 81 dB in my sim. Which sounds as the OP described.