IRPS instead URPS

[ChocoHolic]

Hi Subwoofer Friends!

....some fundamental thoughts:

Up to now I got good results with URPS (Under Resonance Principle Subwoofer).
Typically such designs use very small closed cabinets, which result in high resonance frequencies. The frequency response is equalized with a LT and if required damping can be adjusted by a negative output impedance of the amplifier.
The disadvantage of URPS is a poor efficiency, as the speakers output drops with second order below the resonance frequency.
Additional I do not like resonance frequencies in the range between 40Hz....70Hz, as they do not match the taste of my ears.... (to much interaction with room resonances)... So for URPS I prefer resonance frequencies above that, which really drives the required power
to crazy values....

For a new and much more powerfull subwoofer I am exploring a new idea. IRPS. In Resonance Principle Subwoofer.
In general I plan to use the subwoofer below 45 Hz.
If I squeeze a sadhara driver into a 20l box, then the resulting resonance frequency would be around 50Hz... some standard URPS design (and absolutely not my favourite resonance frequency!)...
I played around with adding mass to the cone in simulation. Normally a higher mass reduces the efficiency, but this is only correct if we run the system above the resonance frequency.
Also the simulation showed that a Sadhara in 20l box has 3 db more efficiency at 20Hz if a mass of about 200g .. 300g (!!!!) is added to the cone and the resulting system resonance frequency is trimmed to about 30Hz...35Hz. (For correct damping it would of course require to use a negative output impedance of the amp.)

Well, I did a test with a cheaper speaker. I picked a small Visaton W100S and measured the required input, which is required to get +/-2mm cone excursion. First I measured the original driver (fres=67Hz). Then I measured it with 25g added to the cone (fres=35Hz). The results also show that with added mass the required input power was less for frequencies below 43Hz.
Please refer to the attached zipped Excel Sheet.

Comments/hints/experiences welcome!

Bye
Markus

[BillFitzmaurice]

Adding mass will not add wideband sensitivity, but it will lower Fs. The lowered Fs will allow the driver to operate more effectively to lower frequencies. However, what you gain in the low end you'll give up in the high end. That's generally OK for a sub, you wouldn't want to do so for a wide range woofer.

The bottom line is that what you're doing merely duplicates what would happen if you simply bought a driver with the same specs with the exception of a higher Mms and of course the resulting lowered Fs, along with a lower broadband sensitivity and lowered high frequency F3 that would also result. The most common usage of adding cone mass is with small drivers, 5 inch or less, where one may desire a lower Fs than is available from a stock driver.

[brummkreisel]

ChocoHolic,


I don't understand why you don't like resonance frequencies (of the CB) around 40 to 70 Hz when you equalize the system anyways (using a LT and not just a shelving filter)? Where and why should there be more room-interaction? How do you prevent such an interaction by having the CB resonance (i.e. not the resonances of the room) above, say, 100 Hz? After all, you equalize you system to a linear frequency respone, hence you do very well get room resonance excitation. Am I missing something here?

Cheers,
BK

[ChocoHolic]

Hi BillFitzmaurice!
I am glad to hear your words.
Your comments match to what I learned from various books and several DIY projects.

I already used increased mass long time ago, but at that time I wanted to use the speaker also at higher frequencies and I had no possibility to use a negative output impedance in order to control the damping.
It worked acceptable, but it burned "poor efficiency and low damping" into my brain.... and I simply forgot to think in that direction for a small bandwith subwoofer.....

But in my current project it seems to be the best approach.
I was also investigating an approach with ACE electronic, but with that the second resonance of my drivers turned critical. The speakers Le with the cone mass causes a 150Hz resonance and if the RDC is completely compensated (as required in ACE electronic) this resonance gets undamped.....

When I was selecting the speaker, my main goal was to get a driver with very high volume displacement. So the linear Xmax of +/-27mm of that 12" drivers are exactly what I was looking for.
And in fact when I go for IRPS, then a full bridge class D amp with rails of 90V will be sufficient. This will simplify the amp design and reduce the risk of burning the voice coils.... step by step the concept is getting somehow reasonable....

Bye
Markus

[ChocoHolic]

Hi Brummkreisel!

Equalizing with LT does not change the natural resonance of the system. LT simply increases the power at lower frequencies, by a matched filter.
Changing the resonance would be possible with a complex output impedance of the amplifier as the ACE electronic does.....

40Hz....70Hz is critical because the typical room resonance between floor and top of the room is around 70Hz, related to typical heigth of 2,40m. Most rooms have the second and third main resonance (room length and width) somwhere between 40Hz and 60Hz.....

Of course a larger room may also have resonance around 30Hz..., but my ears can accept a 30Hz resonance easier than 70Hz.
Heavy resonacing subwoofers between 50Hz to 70Hz are the typical boom designs, which I can accept for a party, but not all the day.

In general you could reduce issues with room resonances also with matched (matched to the room) notch filters. But I always try to survive without that and find a suitable position for the subwoofer.
A subwoofer which does not boom up the room resonances with it's own resonances, gave the best results during my former projects.

Bye
Markus

[richie00boy]

We seem to be at odds as to what a Linkwitz Transform does. It seems to me that what I know as an LT is the answer to all your prayers. It changes the resonance and also the damping by complex pole manipulation, which is why I was completely baffled when you said you used a negative output impedance amplifier to change the damping when using LT.

I have been using subs from crossover frequency, down through resonance and below, for many years now. Using LT and getting great sound.

URPS imples dual-integrators or 2nd-order filter for the EQ to me, LT is a completely different animal.

[ChocoHolic]

Hi Richie!
For me a LT is something like this:
http://sound.westhost.com/project71.htm
The LT circuit does nothing but the reverse frequency response of the
box. And by this it compensates the roll off (to some extend...).
Or am I to blind to see the special trick?

Let's have a look to the damping of the speaker itself.
Damping is describes how fast the swinging of the speaker decreases
after an input signal stops...
I do not see an improvement of this by using LT.
What is the strukture of your LT?

Bye
Markus

[richie00boy]

Ahh Chocoholic, you missed a LOT!

The LT is not simple amplitude compensation, it is a pole-shifting circuit that manipulates a given F and Q to be a new desired F and Q. Hence the amplitude is compensated, but so is the phase and the damping.

You are supposed to chose target Q to be different to what your box Qtc is! That's half the point of using the LT.

[ChocoHolic]

....hm, for me it is clear that LT can/must of course not only compensate amplitude, but also phase and both Q of the closed box. Also the new Q for the new overall frequency response can be adjusted to any desired value.... yes...

But my concern is the speaker damping.
Let's assume no input signal.
Move the cone to the excursion of 5mm (by external mechanic) and then let it swing back. How fast does stop swinging?
The decrease of swinging will be the same with and without LT, because in both cases the amps output will always give 0V to the speakers terminals.
But a negative output impedance would result in a defined signal to the speaker terminals, which is opposed to the speakers natural swinging and the swinging will stop much faster.

In fact you can easily try this by knocking on the cone of the woofer
in your normal box.
Try to knock if the speaker terminals are open (or simply amplifier off), then you will hear a boom which decreases slow.
If you knock at the cone, when your amplifier is connected (turned on, but no music signal), then the boom will decrease much faster.
If you would use an amp with negative output impedance you could adjust values, that you will not hear any boom anymore.
Well, in reality to much damping does also not sound right.
In fact my trials with my old URPS gave best listening results with high damping, but not max. damping.....

[richie00boy]

Ahh, that old chestnut

The LT provides the overall wanted level of damping by changing the damping characteristics of the signal applied to the speaker.

Thus if you want to tighten the response of your speaker that takes a long time to settle down, (in simple terms) the LT only provides a little bit of signal that is over-damped, so that when it combines with the speaker the net result is the tightly damped speaker you wanted.

It's all about overall Q. If one filter has a Q of 1.2 (under-damped) and another filter has a Q of 0.42, then they combine to give an overall response Q of 0.5.

As an aside, when you added mass to your speaker that time, as well as lowering Fs you also raised Qts, which was exactly what you didn't want...