diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Full Range (http://www.diyaudio.com/forums/full-range/)
-   -   T/S parameters measurement questions (http://www.diyaudio.com/forums/full-range/141469-t-s-parameters-measurement-questions.html)

IG81 30th March 2009 12:36 PM

T/S parameters measurement questions
 
Hello,

I just bought two 5" fullrange speakers from Q-components (McBride). They model is McBride Linear A 5.5.

They have the following parameters from the manufacturer:

Fs: 70Hz
Re: 6.7 ohm
Qms: 3.91
Qes: 0.86
Qts: 0.71
Vas: 8.82l

I first tought I would put those in sealed boxes of ~65l, producing a Qtc around 0.75.

Then I measured myself the parameters and arrived to the following results:

Fs: 71Hz and 70Hz
Re: 6.8 ohm and 6.9 ohm
Qms: 5.13 and 4.63
Qes: 0.49 and 0.47
Qts: 0.45 and 0.43
Vas: did not measure yet

I think I could still use them in a sealed box, but would need to have it be of a smaller volume to bring Qtc back up to ~0.707. Or what else would you suggest?

Would playing music through them for a couple of days "break them in" enough to measure a change in T/S parameters?

And last thing, are my following formulas correct? I've seen them at a few places, but just want to validate with the experts here.

Rc = Zfs / Re
f1 and f2 = sqrt(Rc) * Re
Qms = [Fs * sqrt(Rc)] / (f2 - f1)
Qes = Qms / (Rc - 1)
Qts = 1 / [1/Qms + 1/Qes]

Thanks a lot!

GM 30th March 2009 05:04 PM

Greets!

Wow! That's very unusual! Normally, they publish much lower Qts specs than actual, so unless you intentionally wanted higher Qts specs, you lucked up.

Yes, any driver with a < 0.7 Qts can be put in a 0.7 Qtc sealed alignment, but until Vas is measured, you're wasting your time choosing an alignment, especially sealed to a given Qtc.

Depends on how hard you drive them, but the way it works is that specs change in complementary ways during break in, ie. Fs, Qts falls and Vas rises, so while the specs afterward may be measurably different, any basic box alignment will be the same as based on its out-of-the-box specs. What changes during break in is the shape of its response.

Proven T/S measurements: http://www.diysubwoofers.org/measure.htm

GM

IG81 30th March 2009 05:22 PM

Quote:

Originally posted by GM
Greets!

Wow! That's very unusual! Normally, they publish much lower Qts specs than actual, so unless you intentionally wanted higher Qts specs, you lucked up.

Yes, any driver with a < 0.7 Qts can be put in a 0.7 Qtc sealed alignment, but until Vas is measured, you're wasting your time choosing an alignment, especially sealed to a given Qtc.

Depends on how hard you drive them, but the way it works is that specs change in complementary ways during break in, ie. Fs, Qts falls and Vas rises, so while the specs afterward may be measurably different, any basic box alignment will be the same as based on its out-of-the-box specs. What changes during break in is the shape of its response.

Proven T/S measurements: http://www.diysubwoofers.org/measure.htm

GM

Thanks for the info!

Actually I was going for this 0.71 Qts in hopes of trying a huge box that would get me down to a Qtc ~0.74, but now I'll have to make a smaller box. But I used to have a -3dB frequency at 70Hz, and now with similar Qtc in a smaller box, it's at ~117Hz. Bummer...

I really wanted to go Sealed box on this project, in hopes to get something smooth like my KEF 303's. I'll check out what it looks like in a BIB, I'd prefer that to a vented box.

Also, I'll try measurements the way it's described in your link.

Cheers!

GM 30th March 2009 05:44 PM

You're welcome!

Well, if high SPL isn't required you can always add series resistance to increase Qts to fill up the larger, lower Fb alignment you planned.

GM

IG81 31st March 2009 12:21 PM

Quote:

Originally posted by GM

Proven T/S measurements: http://www.diysubwoofers.org/measure.htm


Regarding the method used to measure Vas in the above link, where it is stated to use a box approximately the cube of the nominal speaker size, is it meant to say that for a 5" driver, to use a 5" x 5" x 5" inner dimensions box?

I guess this is better than the other ways that require a box of the expected Vas, I'll be using much less wood here!

Thanks!

GM 31st March 2009 03:47 PM

Right, the object is to raise the driver's measured free air Fs at least 50%, so while the test box Vb isn't critical, its swept Vb is, i.e. box plus cone Vb, so if this size doesn't do it, then add a known amount of scrap wood to reduce its Vb until this criteria is met. FWIW, for expediency's sake, I ultimately skipped the box and just measured it clamped face down to my workbench and while I don't recall ever comparing the results to a larger box measurement, the speakers I calc'd from it performed as I expected which is all that mattered to me.

The downside though is that this small a Vb has to be very accurately measured and only use just enough signal to barely move it to keep from distorting the surround. Obviously, with any speaker box measurement, a perfect seal is required for best accuracy, so use putty or similar around mounting hardware as even the minuscule air loss around threads can affect the results.

GM

IG81 31st March 2009 05:04 PM

Quote:

Originally posted by GM
Right, the object is to raise the driver's measured free air Fs at least 50%, so while the test box Vb isn't critical, its swept Vb is, i.e. box plus cone Vb, so if this size doesn't do it, then add a known amount of scrap wood to reduce its Vb until this criteria is met. FWIW, for expediency's sake, I ultimately skipped the box and just measured it clamped face down to my workbench and while I don't recall ever comparing the results to a larger box measurement, the speakers I calc'd from it performed as I expected which is all that mattered to me.

The downside though is that this small a Vb has to be very accurately measured and only use just enough signal to barely move it to keep from distorting the surround. Obviously, with any speaker box measurement, a perfect seal is required for best accuracy, so use putty or similar around mounting hardware as even the minuscule air loss around threads can affect the results.

GM


That all sound good. Thanks again!

IG81 31st March 2009 08:28 PM

In regards to adding series resistance to raise Qes, it is not something I knew could be done.

I simulated a dynamic loudspeaker in Spice and drove it with a 1V voltage source with a 0.01ohm output impedance and incrementally added resistance in series with the driver lumped model to see the behaviour. It indeed raises the electrical Q of the resonance. Of course this is gained at the expense of efficiency.

I did not get my values right yet for all the elements in the model, but I still get how it works. Thus, I was not able to calculate proper Q's yet and by how much does it varies with series resistance.

Therefore I ask, how much resistance is typically added to raise Qes in such a situation?

GM 1st April 2009 07:19 AM

True, not many 'free lunches' in audio design.

For a series resistor:

Qes' = Qes*(1+Rs/Rvc), where Rs is the series resistance and Rvc is the driver's DC resistance, so:

Rs = (Qes'/Qes-1)*Rvc. Remember to include the series resistances of your amplifier/cable/inductors in Rs, so the actual resistor you buy may be somewhat smaller than calculated value. Or do as I do and just use a pot to dial in the sound, then measure it for the fixed resistor value.

Qts' = Qes'*Qms/(Qes'+Qms)

Vas is unchanged

n0 = 9.614*10^-10*Fs^3*Vas(liters)/Qes'

SPL = 112.02+10*Log(n0)

As many times as Ron E, me and others have posted this info I'm surprised I couldn't find it in a search. :(

As much as required to get the desired response, which of course assumes there\'s enough amp power and speaker power handling to offset the \'hit\' on efficiency, though like most things there\'s a point of diminishing returns based on the alignment, so no set value limit. That said, up to 20 ohms of bass boost was available on some tube amps way back when, so after I switched to SS power I\'d use a cheap 25 ohm pot to dial it in by ear in-room, then measure it to make a fixed resistor filter. Can\'t recall ever needing more than this with up to a matching nominal impedance being the usual limit.

GM

IG81 1st April 2009 01:11 PM

Hi GM,

I used the measurement method in your link and got about the same resulsts as my previous method. Except for Qms, which measured ~15-20% lower, but that did not end up changing the Qts much of course.

I still have to build a box to measure Vas.

Thanks for the formulas regarding series resistance. I guess another benefit of adding the former is reducing the reactive portion of the total load the amplifier sees.

My drivers should be capable of ~105dB top and will not be used for monster SPL\'s, so I guess I could afford to shave a few decibels off the top.


All times are GMT. The time now is 04:26 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2