Critical Q redux

[Dave Jones]

Soooo. If you take a gander at the "woofer" me thread, you'll see that I'm thinking about doing a subwoofer that's basically the same as the "Critical Q" subwoofer on the web. I went back and took another look at that site, http://members.ozemail.com.au/~joeras/sub_index.htm

Joe Rasmussen (the Critical Q dude) measured his Peerless 83500 XLS, and he came up with a higher Qts than the published spec (.23 rather than .20), and a smaller Vas (124L rather than 138L. He figured an effective volume of 42L. How'd he get that number? I get 33L.

He also describes a very different method of modifying the plate amp than the manufacturer documents.

Can anyone tell me what kind of EQ is really needed? (Assume a "normal" amount of room lift.) It looks to me like a boost of 6dB in the low 20's would be indicated.

[Dave Jones]

ttt

[kelticwizard]

Dunno, Dave. I plugged Rasmussen's own numbers into WinISD and came up with the same thing you did-the box is Qtc=0.50 at 33 liters.

On the other hand, if you put the Peerless into a 42 liter box, (using Joe's measured Thiele-small parameters), your Qtc=0.46.

However, at Fc, that difference is only 0.7 dB, so maybe not a major issue.

On the other hand, Joe Rasmussen somehow says that the Fb of the box will be at 39 Hz. Using Joe's own measured parameters, I came up with an Fb of 35.7 Hz, instead of 39.

Rasmussen's calculations seem to be off, but not by much, and his measurements of the Thiele-Small parameters seem to indicate a woofer that conforms pretty close to the published specs.

I honestly think you can go either way on this and still have what you are looking for. I would think that it would be wise to measure the Thiele-Small parameters of whatever XLS you buy and go for a Qtc of between 0.5 and 0.46. I think your numbers are going to be different from Joe's although they might be reasonably close.

PS: measuring the WinISD simulation, at 42 liters you are -6 dB at 39 Hz and -6.7 dB at 35.7 Hz.

[kelticwizard]

Here is the WinISD simulation using Joe Rasmussen's own measured numbers for a box of 33 Hz, (blue) and 42 Hz, (red).

Red = Qtc is 0.46
Blue= Qtc is 0.50

Not a lot of difference.

[SY]

Dave, I'd try it without the boost first. Room placement can do a lot of that work for you. When I eq my subs (downward firing, placed at a wall) to get a supposedly flat response, they're too much. And the in-room response measurements show that.

[kelticwizard]

Dave, here is a chart of step responses for various values of Qtc. As you can see, there are only minor differences between Qtc=0.5 and Qtc=0.71. If you ever saw the step responses of Vented Boxes, which some people, (including myself), think sound just fine, then you would see there is no real reason to get nervous about this. Regardless of what you do, you are going to be much, much closer to Qtx=0.5 than Qtc=0.71.

[kelticwizard]

Now here are the step responses for various vented alignments. Generally, the a=1 and a=4 are the most useful-the other two are fairly extreme setups. I don't know about the Keibs and SC4 alignments, because Small does not go into them into detail. but you get the idea-vented boxes give bigger output, but not as good step response. And step response would be a good indicator of transient response, I would think.

Both closed box and vented charts are by Small, with the colored printing added by myself.

When you compare these charts, now you know why Rasmussen says that building a closed box with a Qtc=0.5 is like adding a servo to a box with a different alignment, especially a vented alignment-even one of the better ones.

It should also be pointed out that the lower the frequency, the less distortion is audible. Considering the advantages in output, cone excursion requirements, etc., that is why the vented box is popular. So the argument goes on. But here are the charts to show how well each follows a single pulse.

[kelticwizard]

Here is Subwoofer Simulator, written by our own member, F4ier, and how it simulates a Peerless 830500 in a 42 liter box. I am using the Peerless numbers here-didn't feel like changing them to Rasmussen's numbers. But Rasmussens's and Peerless' numbers are close, so it makes little difference.

These curves are for an amp driven at 28.3 volts-which translates into 200 watts for a 4 ohm woofer. The 830500 is essentially a 4 ohm woofer-okay, maybe a 5 ohm woofer. Not much difference.

The upper green line is the response using typical room gain-the same curve that Rasmussen uses, only added to the simulation chart. This was made possible by a file by Serow, to be added to Subwoofer Simulator.

The lower green line is the response without room gain. As you can see, room gain adds quite a lot.

The red line is the impedance, with the peak right where Rasmussen says it is, more or less.

Room gain of course varies room to room, and between various parts of room. These are averages.

But this shows why SY is saying position the subwoofer before you EQ-room gain is like an EQ in and of itself. And it will vary in different parts of the room.

PS: About that dropoff over 100 Hz-Subwoofer Simulator takes into account the speaker's response if the cone were perfect. The high frequency rolloff due to inductance is possibly over emphasized-that is why the program is called Subwoofer Simulator.

[Dave Jones]

Originally posted by kelticwizard
Here is Subwoofer Simulator, written by our own member, F4ier, and how it simulates a Peerless 830500 in a 42 liter box.

To my untutored eye, that still doesn't look too flat. Consider also that I want to use an off-the shelf box if possible. I found one expensive one that's about 30 liters. Otherwise, there's just 1 cu ft (28L) boxes. Subtract about 6 liters (?) for the woofer itself and a plate amp if I go that way, and thats 22 liters. Even if I can get 25% back by stuffing it, I'm coming up short.

There is an apparent sparcity of woofers with Qts below .4 or so, making it difficult to build a subwoofer with a Qtc of around .5. Other possibilities are some 8" or 10" woofers by Peerless and Scan Speak. Using those might require multiple drivers, so we are talking about some bucks.

[GM]

Parts Express has 2.0 an 3.0ft^3 subwoofer cabs in different finishes IIRC. Don't expect >20% apparent Vb gain through stuffing, though theoretically it's 40%. WRT flat response, look at the driver's high inductance (Le). It's plain to see its effect on the FR in the factory plot.

GM