To measure the Vas of a driver one must input a driver diameter (cone + ½ of the surround), from which, I assume, that DATS calculates the effective cone area used in the Vas calculation. I am in the process of measuring the TS parameters of a KEF SP1587 “Uni-Q” (coaxial) driver. My question is: would it accurate to use a calculated ‘effective diameter’ based upon normally measured diameter minus coaxial tweeter diameter? (That is, convert the result of the two diameter measurements to a fictional effective ‘diameter’).
A second question did enter my mind: would it be more accurate to use to actual cone ‘surface area’ rather than ‘diameter’? (Although a bit more complex to measure, the surface area can be significantly different than an area based solely on the diameter, especially with ‘deep’ driver cones).
A second question did enter my mind: would it be more accurate to use to actual cone ‘surface area’ rather than ‘diameter’? (Although a bit more complex to measure, the surface area can be significantly different than an area based solely on the diameter, especially with ‘deep’ driver cones).
Unfortunately you can't just subtract diameters to determine the effective Sd because Sd is related to the square of the diameter. Use D(effective) = square root of (Dwoofer (squared) - Dtweeter(squared)). Use that diameter value for DATS.
IIRC the SP1587 is the driver in the Q100 speaker. If so, a couple of other people have already measured it:LINK1 and LINK2. Search for Q100 in LINK2.
IIRC the SP1587 is the driver in the Q100 speaker. If so, a couple of other people have already measured it:LINK1 and LINK2. Search for Q100 in LINK2.
That can’t be right. Both of those put the driver sensitivity under the observed system sensitivity. See KEF Q900 Speaker System HT Labs Measures | Sound & Vision
You can go either way, but as others have noted, the T/s parameters are a good guide, but not gospel, for speaker alignment, especially with various amounts of stuffing. 
Don't spend too much time trying to attain 0.0001% resolution, and instead spend it measuring the results empirically.
Best,
E
Don't spend too much time trying to attain 0.0001% resolution, and instead spend it measuring the results empirically.
Best,
E
You can go either way, but as others have noted, the T/s parameters are a good guide, but not gospel, for speaker alignment, especially with various amounts of stuffing.
I repeat, a technique that determines a raw woofer has a lower sensitivity than the commercial system employing that woofer is not valid as to that woofer.
You should not trust the efficiency number calculated from an impedance measurement. In addition, this number is typically calculated at 1Watt instead of at a fixed input voltage so it somewhat meaningless for a driver that does not have a resistive 8ohm impedance. The fact that T/S model is only a 1st order approximation adds further inaccuracy.
To determine the sensitivity look no futher than a frequency response measurement at 2.83V/1m. I trust Zaph's measurement to be accurate here: about 89dB around Zmin (~200Hz). This is inline with KEFs own spec of 86dB for the Q100 once you factor in a few dB of baffle step compensation.
As with all systems, you should measure the actual woofer and tweeter yourself, at the same time in order to to gauge the difference in sensitivity and design the crossover. There is too much error in T/S efficiencies, measurement system calibration and driver production variation to do it any other way.
Last edited:
(I guess TMM responds quicker than I do!)
What you're seeing is the old 1W/1m vs 2.83V/1m difference in measuring SPL. If you look at Zaph's T/S parameters you'll see 83.92 dB efficiency and a DC resistance of 3.0 ohms. Here's what he says about his T/S measurements:
"Also note that the SPL number on the T/S parameters is an efficiency number based on one watt. Additionally, this number is calculated rather than actual. The SPL shown on the response graphs is a sensitivity number based on 2.83 volts and is actual rather than calculated."
On the other hand, the Sound & Vision article states: "All passive loudspeakers were measured with grilles at a distance of 1 meter with a 2.83-volt input and scaled for display purposes." And it also states that the Q100 sensitivity to is "88 dB from 500 Hz to 2 kHz."
By my calculations the difference between 1W and 2.83V for a speaker with a DC resistance of 3 ohms is a tad over 4 dB, changing Zaph's 1W value of 83.92 dB to about 88 dB. This matches closely to the measured SPL in his frequency response graph which is done at 2.83V.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.