Woofer testing WT3/ DATS methods (specified SPL?)

[gearfreak]

DATS will measure the impedance sweep in free air, but for the additional parameters requires one of these methods:

1.) Test Box Method - requires a suitable test box
2.) Added Mass Method - requires you to add a known mass
3.) Specified SPL Method - requires the SPL be known
4.) Specified MMD Method - requires the MMD be known

I don't have a suitable test box yet, so am trying to figure a way to use SPL or MMD methods.

(3: SPL Method)

For SPL I presume we are looking for 'free air' SPL measurements - but the spec sheets all are per enclosure.

e.g. as from 515-G series LF Speaker.pdf

enclosure/sensitivity (@1w/1.2m)
828G / 103db
816A / 102db
817A / 104.5db


This source:
http://stereointegrity.com/wp-conten...yofadriver.pdf

Offers SPL calculation via Reff Eff. (eta naught):

So, how do we calculate the SPL of a driver from n0? Use the following equation: SPL = 112 + 10 * log(n0 )

n0 is spec'd (6.0%)

so we derive:

SPL = 112 + 10 * log(.06)
SPL = 112 + 10 * -1.2218487496163563674912332020204
SPL = 99.7815

Can anyone vet or invalidate this approach?



(4: MMD Method)


Thought I had a way to get to MMd via calculation here:
Driver TS Parameters: Mmd & Mms >> speakerwizard.co.uk
Fs = 1 / 2 pi * (sqrt [Cms * MMs])

Alas no. But HornResp might provide an avenue... will check that out.

[GM]

Best overall to use the added mass method. Don't know what amount DATS recommends, but the pioneers worked it out to a fare-thee-well and arrived at adding enough mass to decrease the measured driver's free air Fs by [at least] 1.56x, but not much more, though IIRC most folks just shoot for a > 1.25x, then use the formula [or one of the many worksheets available on-line] to find the driver's Vas if not automatically done. I assume the DATS does the other calculations such as efficiency, Mms, Mmd, etc., using Fs, Vas, Qes.

A cornucopia of acoustical math calculators: Audio calculations in English acoustics calculator convert audio formulas sound dB audio system microphone electro engineering electronics formula sound recording studio useful stuff free audio calculator recording studio acoustic audio engineering s

If determined to work from SPL, then the math normally used is based on a 1 kHz mid-band efficiency to allow accurate comparisons between lows-mids-highs, which manufacturers don't necessarily follow for a variety of reasons, but what you should measure, i.e. don't rely on published n0! When I was actively designing, I vetted the published specs as much as I could using the various formulas and most T/S specs were way off, especially n0 and one reason why I always spec [very] large cabs to allow plenty of tuning flexibility. Best overall to just calculate it and if DATS does it, it will be interesting to see if they match: mh-audio.nl - Home

Mmd, etc., calculator: mh-audio.nl - TSP Calculator

GM

[gearfreak]

Thanks as always GM.

I improvised a test box (trash can - 4.467 cuft) and have those results for one. Higher Vas 19-21cuftm and Q values than spec'd, but perhaps not unexpected as you mention:
[ Measuring 604 D TS parameters ]

...and the test was rudimentary WRT bracing etc.

Will attempt added mass, magnets en route.
I've not yet found suggested placement, but am going under the presumption that near enough to the VC as one can get is a goal.

I have 4 to measure, so it will be interesting for us to see what agrees with what, or not. :-)

[GM]

You're welcome!

Correctomundo, with smaller drivers, folks use large sockets, coins, large fender washers, old lead wheel weights, or large diameter solder or lead hand formed rings. Basically anything that doesn't 'dance around'. For larger drivers, [driver] hole cutouts/scrap 'discs' or similar. Mine is an old plastic pickle jar with a rather deep, somewhat hemispherical, rounded off bottom that fortuitously clears even the 515's DC and just add rocks to get close, then add sand till it hits the target frequency, but you could make similar by sawing off the bottom and glue in a recessed wood disk and add a thick enough felt ring to it to protect the driver.

GM

[gearfreak]

Thanks GM.

I've done ~487g (a little more than a pound) attached about halfway up the cone and we're seeing no change in the Fs calculated when I run the test.

I figure either I am too far away from the VC or I need to add way more weight.
I will pursue the former before the latter.

[GM]

Added to a 604D?! What's the Fs?

I don't have my spreadsheets loaded to do the math, but I'd be surprised if its moving mass [Mms] is > 40-45 g, so it's not going to take but a fraction of a pound to lower its Fs 1.56x, implying it's frozen up. When I said rocks, I meant the small riverbed rocks my drive/walkway extensions are filled with.

I've never measured a 604, but seems like this much weight would have long since bottomed it out.

Yes, normally added around/on the dust cap.

BTW, what do you mean by 'Fs calculated'? It must be measured.

GM

[gearfreak]

yes, the DATS workflow is not clear to me, or else it would seem that the software has a bug. Regardless of what I try with the 'added mass method' test, Fs displayed does not change.
Good thing I stopped before I added 2 pounds to the thing! This is a shame, I'd like to have compared their Vas values from the different tests.

Nonetheless, using it to obtain Fs and +/-9db points (F1 and F2 if I have interpreted them correctly) lets mh-audio.nl - TSP Calculator arrive at some values.

The question I have is whether to use:
- the DATS and TSPCalculator method of calculating effective piston diameter: measure diameter as to +1/2 of the surround
or
- go with the published altec spec.

They differ by 1" so seem to have a substantial impact.

Further, we see that despite using the altec value (or very very close - 29.13 vs 29.12), the Sd that TSPCalculator arrives at is vastly different (666.63 cm2) than the altec spec (848.4).
So 'whether to include the surround' and how that is used by the calculator ultimately is not quite clear - or I've misunderstood it.

Ignore the tiny attachments if they show up below. Sharing the data was the point of this followup, but I can't make sense of the HB requirement of 19.5 kb at the moment.

[bowtie427ss]

I'm no expert with this to be sure. But, one thing that was stressed to me about the added mass method was the method of attachment to the cone with the idea that if it's not a rigid bond, some damping will occur between the mass and cone adversely affecting your results.

One of the more clever methods i've seen was the use of pairs of small magnets placed on opposite sides of the cone, and spaced evenly around the inner circumference.

[gearfreak]

Thanks Bowtie.

I am pretty certain that the DATS 'Calculate Vas' routine (or workflow, or my workflow) is buggy.
When I return to 'measure free air parameters' and use added mass, I get the expected change in Fs.

In this case attachment is pretty good.
I went with multiple small magnets so as not to leave an impression/bend/mark the cone. (appx 2mmx6mm) x8.
Ultimately a small 3"W x 5"H 'stainless' steel mixing bowl worked well to clear the dustcap and still be practically close to the vc. Anything smaller diameter makes the magnets to hard to attach from the backside. The bowl and magnet came in @ ~90g (~1/5 lb) and this dropped the Fs from 37.x down to 24.x hz. *

(for reference, a 16oz e.g. deer park water bottle is about a pound - so 1/5 full and one has a ballpark of the appx. necessary weight)

* ...as the 515 is front vented, I think this impairs the measurement (I'd have to cut a hole in the bowl).
Not a problem per se - but a small paint can at least 5" dia would probably be easier. I am on the hunt this afternoon. A coffee can came close but no cigar. I have a 1gal paint can, but it's cruddy.

ION: found GM's comment re SD for a 604. Looks like he prefers the more accurate smaller dim:

You're welcome!
...
WRT your sims, they're unfortunately completely bogus since BL is off by almost a factor of two. Even if BL is published, one should always calculate it when using published specs and why MJK has a BL calculator WS. Unfortunately it would still be off due to the Sd you used being too large for any 604 with 96"^2/619.354 cm^2 being more technically correct, though at a glance it doesn't change the MC sim, so assume it's a moot point.
[/FONT]