simulating a onken

[zottinger]

Hi
I'm planning to build a set of small desktop speakers inspired by the uFonken.
I want to use Visaton FRS8 speakers and currently I'm trying to simulate these with Boxsim. From what I understand, the Onken type is somewhere between a bass reflex enclosure and a closed box, because the slim and wide ports create a higher resistance than a normal BR port, is that right?
In Boxsim I got the option to simulate a stuffed port, is this a way to get close to simulating Onken type ports? As an example I did a quick sim of the frs8-4 in a 2L BR tuned to 100hz, the dotted line is without stuffing, the other one is with what boxsim a "lightly stuffed" port (6,2g polyester wool/L).
I'm not sure about size and tuning, but I don't want to go much bigger. Actually this simulation looks much nicer than the 6L BR speakers I just build, which surprised me how good they sound. Since I'm new to all of this any comments on my general plan are appreciated.

Daniel

[planet10]

1st off the miniOnkens i design only borrow the concept of the slots from the onken.

Looks like you are on the right track... this sim program convolves the sim with actual driver FR?

dave

[zottinger]

Originally Posted by planet10

1st off the miniOnkens i design only borrow the concept of the slots from the onken.

Can you elaborate a little*. From what I found with Google and the forum search it seems to me that most of the onken designs are for woofers and the configuration is a way to accommodate big port areas while the miniOnkens and the MarkO
use it to get better damping. Is that what you meant?

*I'm trying to learn as I go along one of the reasons I want to use the same drivers in a smaller cabinet again to see what different cabinets do.

Originally Posted by planet10

Looks like you are on the right track... this sim program convolves the sim with actual driver FR?

Yes it simulates based on measured FR, but you can choose to use Thiele Small parameters instead, but you also need to know Le, which is sometimes hard to find. Unfortunately it is German, only and somewhat proprietary to Visaton. It's freeware and there is a free driver database for Visaton-made drivers, you can also enter your own measured data, or import them from a text file.
There is a thread trying to provide some translation:
BoxSim

[planet10]

An Onken follows a specific set of rules, which i ignore when designing the miniOnkens. I look for a specific shape of curves and work from there.

The effect i am looking for is to add some "R" to the ports to give them more tolerance to varying T/S parameters as the volume & the dynamics change.

You are on the right track with the curve that damps the lower impedance peak.

dave

[zottinger]

Thanks for your help
daniel

[chris661]

Hi Daniel.

I'm a little curious about your choice of driver. For a vented cabinet, wouldn't the lower Qts FRS8M be better? It also has 6dB more efficiency, which is surely a good thing...

Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories

[zottinger]

Originally Posted by chris661

Hi Daniel.

I'm a little curious about your choice of driver. For a vented cabinet, wouldn't the lower Qts FRS8M be better? It also has 6dB more efficiency, which is surely a good thing...

Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories

Hi Chris, there are a couple of reasons, first of all I already got a pair of frs8 for 5 Euro on ebay, besides that: I know of several successful designs using the frs8 in a 2l BR tuned to about 100hz, so I'm playing it safe here. Furthermore the consensus on the Visaton forum seems to be that you can get away without BSC on a FRS8 quite often, while with a FRS8M you almost always need one. If you compare the FR of both drivers you see a little peak around 1000hz and a slight rise after that for the FRS8M, while the FRS8 is flat around 1000hz and dips a little after that. And 1000hz is also where the baffle step with my planned baffle width would be. There was one more issue the M Version has, but I can't find it right now, cause I really got to go, will try later to dig up what it was. Daniel

[chris661]

Likely the peak around 10-15kHz.

Fair enough for your choice of drivers. I used an FRS8M per side with some woofers, so BSC wasn't much of a concern. I've also messed around with the FRS8 8ohm, and they benefit from phase plug and a couple of layers of (diluted) PVA glue. Cone break-up isn't so harsh, and the phase plug gives better treble dispersion. If they're cheap, it might be worth a go.

[zottinger]

I could not find this other thread about why to prefer the non M, however I did a sim of the same 2l 100hz BR-box for the M Version and there you can see what I mean between 1000 and 2000hz (dotted line is frs8M) where it is important. My baffle width will be between 11,5 and 12,5cm which puts my baffle step somewhere between 900 and 1000hz.


The frs8m unfortunately has a rising spl in this area even in a 135x165cm baffle (that's the so called din-Schallwand mentioned in Boxsim, so Baffle Step and FR add up and call for a BSC. From what I've this effect is even more pronounced in real application, but then again the Visaton-forum seems to be obsessed with BSC circuits.

Two questions:
PVA glue is the white wood glue, right? If so, is it the express or the non-express variant?
Do you have a pic of your drivers with phase plug, did you measure the speaker before/ after?

Somewhat unrelated:
Wohooo, coming back home earlier I found enough material for at least pair of test boxes in the dumpster of a construction site. Thats cool because this way I can get a feeling for the form factor, which is important with this built, since I want to make them luggage-friendly. I guess the material (usually used to protect wooden floors, when doing major reconstruction in a flat, I don't even know the German word) I picked up is just to soft to tell much about how they sound.
Daniel

[zottinger]

I found the thread I was thinking about, but it is basically about the same problem - the rising SPL of the FRS8M at higher frequencies. According to the guy who wrote boxsim you have to lower the efficiency of the FRS8M to basically the same level as the non-M to get the same frequency response.
If you speak German it is here:
FRS8 oder FRS8 M ?? - Visaton Diskussionsforum