major questions about SETs and sealed enclosures

[keto]

Hey all,

For a driver of Fs=34.7hz, Qts=.33, Vas=312L, and solving for Qtc=.707, gives a sealed volume of 87L with an f3 of 74hz. Great!

Now, I would like to include the SET damping factor, to get a 2nd-order alignment that is Qtc=1 and then "tune" the coupling cap and parafeed cap (the speaker has a very flat impedance vs frequency curve) to that f0 frequency, or the frequency between the hump and the slope at which the response is at 0dB. And voila, end up with something approaching a Linkwitz-Riley 4th-order high-pass.

Is there a way to calculate the f0 frequency for the aforementioned 87L enclosure and driver and SET amp?

Or, is 87L no longer the optimal enclosure volume to get such a 2nd-order, Q=1 roll-off?

Any help would be greatly appreciated. I would like to have an approximate Vb before building the box (my present 60-120L is too big of a bracket), and a better idea of where to bring in the subs (I'm guessing somewhere between 80 and 105hz, again, too big of a bracket), also with the LR4 filter.

many thanks,

--keto

[ssssly]

I don't know the math for the impedance matching, I'm sure someone smarter than I will comment on it though.

However I do know that if you are going to calculate things down to that level that you can not rely on the manufacturers publshed specs for the speaker. You are going to need to test the speakers to find their actual parameters.

Also if you are building a simple sealed enclosure I would suggest making it about 15% too large and then reduce volume with bags of sand, or the like, instead of trying to build it to exact specs the first time.

[keto]

Thanks, that's good advice. I just realized that I already have a 60L enclosure, which is the calculated Vb for the driver with a 25% increase in Qes, Q(speaker+box+amp)=1 and f0 of 86hz. I guess I'll start there.

[Paul Joppa]

Just increase the Thiele-Small parameter Qe by the factor (Rdc + Req)/Rdc. Rdc is the dc voice coil resistance, Req is the output impedance of the amp (usually about 1/3 of the nominal load, e.g. 2.67 ohms on an 8 ohm tap). I do it in boxplot, works great.

However, your speaker impedance is probably not flat at the resonance unless Qm is much much lower than Qe - which would be VERY unusual! The parafeed cap will act like a series cap and change the behavior of the alignment. Unfortunately Benson's hard-to-find and even harder-to-read book is the only quantitative reference I know. I use PSpice to simulate it rather then using an analytical design - still a pain in the backside but not quite as severe!

[keto]

@ Rdc = 6.0/7.5 ohms (disturbing that they are so different), I increase Qe by the approximate 40% and get (Qm=5.5, Qe=.35x1.4=.49) Qt=.45. 80L (2.8 cu ft) and f0 of 86hz. Great! The driver has a "flat" impedance curve, that errr.... doubles between 90 and 45hz. http://www.aespeakers.com/phpbb2/viewtopic.php?f=3&t=1677 Maybe I'll add a cap on the input rather than depending on the parafeed cap for the 4th pole. Okay, I've got the theoretical size and frequency. Thanks!!!