Rear Horn bandwidth

[e-side]

Currently I'm trying to design a BLH for a 6.5" bass-mid driver. t/s parameters:

Re = 9,4 Ohm
Fs = 54,7 Hz
Qms = 4,03
Qes = 0,89
Qts = 0,73
Vas = 12,8 L
Sd ~ 130 cm^2

I downloaded Keele's paper on horn design. He mentions 2 corner frequencies (Fhs and Fls), which indicate 'the range over which a driver will be suitable for use as a horn driver considering small-signal operation only'. In my case, [Fhs]~75 Hz and [Fls]~40 Hz. The lower driver compliance corner frequency [Flc] is ~ 20 Hz, because Vb is almost infinite for a BLH. The first HF-breakpoint [Fhm] occurs at ~ 150 Hz.

Q1) When choosing the lower and higher cutoff frequencies [Fl] and [Fh] for my BLH, what should I consider as the widest bandwidth possible for this driver? 20-150 Hz (Flc - Fhm)? Or 40-150 Hz (Fls - Fhm)?

Q2) Will a high cutoff frequency of 150 Hz be high enough to get good bass reponse out of this 'fullrange' BLH/driver combination?

thanks in advance

Erwin

[GM]

Greets!

High Qts driver BLHs are best designed as big vent BRs, so choose its BW based on blending it to either the baffle step or the driver's Xmax excursion limit, whichever is highest. As Till notes though, a good mass loaded TL is a superior performer overall: http://www.diyaudio.com/forums/showt...512#post899512

Don's math is for choosing a max BW design, which tells you what flare frequency to work from to determine the other specs, with Leach's math the most accurate.

GM

[e-side]

Originally posted by GM
Greets!

High Qts driver BLHs are best designed as big vent BRs, so choose its BW based on blending it to either the baffle step or the driver's Xmax excursion limit, whichever is highest.
GM

I'm not sure if I should go for corner or wall placement. The horns will be placed in a quite small room (4x3x2 meters, LxWxH), against a wall. They'll be near a corner (max 50 cm away from one), so maybe I could go for corner placement. I that case a 50 Hz cutoff is possible without compromises.

I seems it's not possible to get a flat frequency response up to the upper cutoff freq when it's set to the baffle step frequency: Fh ~ (115/W(B)), W(B) in meters...

I also have 2 Audax drivers:

Re = 10,7
Fs = 45
Qts = 0,25
Qes = 0,26
Qms = 6,7
Vas = 20,3 L
Sd ~ 130 cm^2

According to Keele the first HF breakpoint occurs at ~ 360 Hz. Is this a suitable high cutoff freq, or is it too high for bass output?

[GM]

Greets!

Room gain will begin in earnest around 60 Hz, with up to a 12 dB/octave rise below ~43 Hz, so depending on how well built/sealed it is will ideally determine the speaker's F3, sysQ. This implies that a sealed cab with an effective 2nd order roll off (0.707 Qtc in-room, so any boundary gain beyond 2pi requires lowering it) and ~60 Hz F3 in-room is the ideal. Your previous driver has a little too high a Qts and the Audax's is too low unless some form of EQ is used, so it's suitable.

How high the mass corner is just determines how much BW must be amplified to flatten its in-room response and above ~250 Hz (some say as high as ~400 Hz) our hearing acuity improves enough that the comb filtering between the horn and driver begins reducing clarity, so at ~136 Hz it shouldn't be a problem, but in a small room where reflections are rampant, who knows? Still, well stuffed furniture, selective room damping can solve any of these type problems.

Since the room is relatively small, rearranging it so that the speakers bracket a corner (making it a common virtual horn) and doing something to make the corner behind the listening position either diffusive or well damped is best.

If this layout isn't an option, then I recommend designing them as side firing into the corners, which allows for a bit of tuning adjustability. This of course requires mirror image speakers.

GM

[AndrewT]

Hi,
I am not sure the Fls formula is relevant.
Most designers ignore it.
They seem to adopt the flare rate and mouth area as the determining factors for bass low limit.

150Hz as a bass cut off frequency will kill the bass frequency. It will almost sound like a mid range.

But you will gain in intelligibility on voice.

[e-side]

Originally posted by GM
Greets!
How high the mass corner is just determines how much BW must be amplified to flatten its in-room response and above ~250 Hz (some say as high as ~400 Hz) our hearing acuity improves enough that the comb filtering between the horn and driver begins reducing clarity, so at ~136 Hz it shouldn't be a problem, but in a small room where reflections are rampant, who knows? Still, well stuffed furniture, selective room damping can solve any of these type problems.
GM

Just to verify my thoughts: In both cases (the high-Q driver and the Audax) I may get good results if the BLH's upper cutoff frequency is set to the mass corner breakpoint (Fh = Fhm). Is this correct?

Originally posted by AndrewT
Hi,
I am not sure the Fls formula is relevant.
Most designers ignore it.
They seem to adopt the flare rate and mouth area as the determining factors for bass low limit.

For BLHs the Fls parameter isn't important I think, but it's used in calculations for front horns (the rear chamber volume).

150Hz as a bass cut off frequency will kill the bass frequency. It will almost sound like a mid range.

I'm not sure if I understand correctly. Do you mean 150 Hz as an upper cutoff frequency?

Thank you both for your help

Erwin

[AndrewT]

Hi,
have I misunderstood this

Will a high cutoff frequency of 150 Hz be high enough to get good bass reponse out of this 'fullrange' BLH/driver combination?

I meant a high pass filter cutting off the bass too high up.

But, in hindsight, maybe he meant low pass filter to remove all the upper frequencies.

Surely, in a BLH, you need as much from the rear side to allow matching with the front wave which has a low limit determined quite effectively by cone size (Sd) and Xmax. This frequency will be highish for a typical BLH optimised driver type.

I do not have fiqures for a for a small Sd and low Xmax, but for the Tannoy Westminster (Sd=750sqcm Xmax~=4mm) the rear horn is full power upto 200Hz and continues above this to crossover to the front horn which is some 500mm to 600mm square. A 13cm cone without front horn will need support from the rear horn way above the 200Hz quoted from the Tannoy example.

[GM]

Originally posted by e-side


Just to verify my thoughts: In both cases (the high-Q driver and the Audax) I may get good results if the BLH's upper cutoff frequency is set to the mass corner breakpoint (Fh = Fhm). Is this correct?

Thank you both for your help

Erwin

Greets!

You're welcome!

Grrr! I hate it when I post confusing BS. I often wind up writing responses in pieces over a day or more and occasionally I either lose track and/or synopsize too much to save time.

Anyway, I indicated that in theory your mystery drivers have too high a Qt to get a ~flat in-room response without EQ and that the Audax had a high enough (maybe even too high) a HF mass corner (Fhm) to work, though the '~136 Hz' was a typo and should have been ~346 Hz, or close to the ~400 Hz upper limit.

GM

[GM]

Originally posted by AndrewT
Hi,
I am not sure the Fls formula is relevant.
Most designers ignore it.
They seem to adopt the flare rate and mouth area as the determining factors for bass low limit.

150Hz as a bass cut off frequency will kill the bass frequency. It will almost sound like a mid range.

But you will gain in intelligibility on voice.

Greets!

It's relevant to me per my first post, but then I don't design horns quite the same way as the others that have published how they do it, but they performed well so apparently there's more than one way to figure them.

'Fhm' is the upper mass corner frequency, so ideally you want the horn 'done' by this point if there's no baffle step issues. Regardless, a 150 Hz vented Fb doesn't ever seem a good plan to me unless it's ~aperiodic.

GM

[e-side]

Originally posted by AndrewT
Hi,
have I misunderstood this
I meant a high pass filter cutting off the bass too high up.

But, in hindsight, maybe he meant low pass filter to remove all the upper frequencies.

I'm sorry for being unclear. Indeed, I meant the low-pass filter idea.

Originally posted by GM


Greets!

You're welcome!

Grrr! I hate it when I post confusing BS. I often wind up writing responses in pieces over a day or more and occasionally I either lose track and/or synopsize too much to save time.

Anyway, I indicated that in theory your mystery drivers have too high a Qt to get a ~flat in-room response without EQ and that the Audax had a high enough (maybe even too high) a HF mass corner (Fhm) to work, though the '~136 Hz' was a typo and should have been ~346 Hz, or close to the ~400 Hz upper limit.

GM

Thank you for your explanation. It's my bad English that causes this confusion.

Thanks again for your help GM and AndrewT!