Starlet Turbo/Aria Ultra Light

[henri22]

Sorry for my Englisch, but ...

Does anyone still have the filter shematics of the Starlet Turbo? I can't find the origional one's anywhere and the project files are not working on the new version of boxsim...

This is what i have:

W200S8 2x
W100SC8 2x
SC10N 2x
WS20E8 2x

My options are:

A slim nice Aria Ultra Light (floorstanding, sloped) or
A Starlet Turbo build as an Trapezium with 2 small BR's in the back
Maybe anyone with another nice idea?

I'm planning on using good filter components ... So no budget-cuts there, wich would be the best?

I have an optional subwoofer to company the Aria UltraLight(Subwoofer 20.40 with the GF200G).

Greets, Henri

[pluto]

Hello Henri,

do you know Istec1003´s database of boxsim projects?

This might be what you´re looking for:
http://home.arcor.de/istec1003/boxsi...arlet-Upgrade/

Good filter components: Difficult to answer if asked in such a general way. Visaton components are okay. Maybe you ask again after you´ve made a choice for your crossover network, considering those parts you are not sure about.

Kind regards, Peter

[henri22]

Well,

If I look at the crossover for the Starlet Turbo there are 15 components, and there is still a dip at about 7000-8000Hz.

Here:


And If I build the SC10N+W100SC_8 Ohm version I get a nice and small(inexpensive) crossover network,

Take a look:
Project data

So:

Is the dip really going to affect the sound of the speaker, and is it not that good, as I search through the forum. Or is it still a good speaker?

And, is the Project from the SC10N_W100SC correct?

Thanks,

[SvenT]

I think the starlet would be a good speaker
the small dip between 7 and 8 kHz can certainly not be heard
don't worry about such small dips

the project you posted is much too loud in the middle and high frequencies!!!
the box is too big for the W100S and the Bassreflex tuning-frequency is too high,
so the box will have a very limited maximum volume

[pluto]

Hello Henri,

the simulation of the Starlet according to Starlet_W200S_W100S-8_SC10.BPJ from Istec1003´s database looks convincingly good. I agree, don´t mind the dip.

I have two recommendations, however:

- The shielded version (SC) of the W100 is slightly louder than the S-type. 3,3 Ohm instead of 1,5 Ohm in the midrange filter should fix that.
- Increase the closed volumen for the W100SC from 1 to 3-4 l. I expect less compression and better damping of the sound radiated from the back of the driver.

Comparing the respective simulations, your version of an Aria Ultra Light is similar to the official Aria Light (just the SPL, not the filter, of course), aside from the range between 500-1000 Hz. Therefore, it will lack bass and fundament, if the speaker is not placed close to a wall - which is the application the Aria Light seems to be meant for.
The tuning frequency is okay. 45 Hz is roughly where the highest levels can be expected in music. If you decrease the volume of the cabinet, this will boost the upper bass a bit and slightly increase the maximum SPL, but more than modest SPL will not be possible, either. If you want to use a Subwoofer, go for a sealed cabinet and a high pass filter.

I´d prefer the Starlet Turbo.

Kind regards, Peter

[henri22]

I'll simulate on the Starlet some more for the W100SC,

Can I use the standard setup crossover and volumes as in the Aria Ultra Light? Or do I have to change it because of the loudness of the w100SC, and do I have to increase volume as well?

[henri22]

Here I've got two options:

One has a high Impedance, the other a low imp at 4000 of about 2 Ohm, wich can I better use?

Var1:

variation1Low impendance

Or

Var2:
Variation2 High impedance

Greets Henri

[UweG]

An Impedance of 2 Ohms means, that there are 4 Watt@2.83 Volt. More than three of your four Watt will heat up resistors. Use the high impedance crossover. In my opinion, the level of the tweeter has to be lowered a bit, but that's a bit of matter of taste.

[henri22]

I think I've got it right this time!

Project data:
Lower impedance at the Tweeter

Gr. Henri

[pluto]

Hello Henri,

the first filter seems to be based on Lonzo´s Merci. The second version looks better. Your latest version I haven´t checked yet.

However, you can make version II cheaper and use standard Visaton components if you choose
15 Ohms (serial) - 3.3 microF (serial) - 0.33 mH (0.6 Ohm, parallel) - 4.7 microF (serial) - * - (SC10N).
If the sound is too bright (Uwe´s opinion), a resistor (1 to 3.3 Ohm, serial) at the "*"-position will help.

I suppose, 2.2 mH and 10 microF will be sufficient for the W100SC-8. Probably, a notch-filter is not necessary; you may check that out by varying the value of the resistor involved.
SPL in the lower frequency range will be substantially higher than in the simulation. Just compare the simulation of the Aria Light with Visaton´s measurement of that speaker or compare simulation and measurement of Lonzo´s Merci.

Although the simulation is not a perfect representation of the real loudspeaker, I would use the real dimensions of the speaker. Making it a floorstander results in about +1 dB in the lower range.

Do you consider using the Ultralight as a bassreflex floorstander while combining it with a subwoofer? (I would still prefer the Starlet Turbo.)

@UweG: The proposed filters are likely to result in a relatively well-balanced sound, if you take in to account the simulation-measurement discrepancy observed for the W100S-family. Such a discrepancy exists also for the W130S (and maybe also for other drivers, whose SPL were measured under the same conditions); see Thomas Reimann´s Regalo for example. Certainly, it is more satisfying to do an almost perfect simulation based on formulas alone, but I would not hesitate to include empirical corrections, if there are differences in the order of +-3 dB.

Kind regards, Peter

[henri22]

Thanks for the filter, I'll try it out on boxsim soon...

With the Starlet Turbo:

The problem I get with replacing the W100s with the w100sc(and with 3,3 Ohm in 4 liter instead of 1,5 Ohm) there is a dip between the w100sc and the sc10n at about 2000-2500hz.

Can that be compensated? Or is it not really influencing the speakers performance?

Gr. Henri

[pluto]

Hello Henri,

I see no advantage offered by the 3rd filter version for the Ultralight. Furthermore, I would not cross over the SC10N any lower than 2 kHz. It is not wise to work in the vicinity of the fundamental resonance of a tweeter, especially if it´s damped by ferrofluide.

Starlet Turbo with W100SC: With respect to the SPL curve, the closest you can get to the version with W100S, is to use
33+4,7=37,7 microF (serial) - 3,3 mH (2,5 Ohm, parallel) - 3,3 Ohm (serial) - 22 Ohm (parallel) - 0,68 mH (0,8 Ohm, serial) - 15 microF (parallel) - 0,15 mH (0,4 Ohm serial) - (W100SC-8).

There will be still 1 dB less SPL around 2 kHz.

But:
- A dip will never disturb as much as a peak with the same excursion. A slight dip (and -2 dB is a slight one) in the range around 2 kHz will make the sound a little softer and the stage a little deeper. Some would prefer that.
- Positioning of the speaker (parallel to the walls or directed towards the listener) has a larger influence.
- The different SPL measured for the W100S and W100SC may be partially due to variations which will occur from sample to sample, even if you choose just one version. So the resonance peak or the dip may look different or may be practically nonexistant in your case. The SC-version is louder due to the stronger magnet; that´s a fact, however.

Kind regards, Peter