Home ›
Site Links
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Categories
Please review the site Rules, Terms of Service, and Privacy Policy at your convenience. Rules, TOS, Privacy
Get familiar with the reaction system: Introducing the Reaction System
LCR meters
Here is the one I use:
That is a good price these days. I have a wall wart I use with mine so that a dying battery doesn't influence my results. Particularly useful for measuring driver DCR.
I have a signature.
Comments
It uses four wire measurement, FYI.
Odd that you would use this over ARTA or REW. If I were comparing commercial products for speaker design use, $100 for a Dayton woofer tester seems to make more sense than a $100 handheld LCR meter. In true DIY fashion, a few bucks in resistors and a lazy afternoon soldering a jig together works just as well.
I used DIY jig for years, they are not accurate at 1hz to measure DCR of a driver. Neither is the Dayton or SL units particularly accurate. Within 10% I'm sure.
Anyways, precision is important
Anyone doing DIY electronics work should have a regular multimeter which is perfectly capable of DCR measurement. For very small resistances, short the meter leads and subtract the lead resistance from the measured value. ARTA provides the ability to manually enter the DCR using a multimeter or estimate by following the slope of the impedance curve at low frequency. Using a soundcard and ARTA/REW, equal precision can be obtained with a half decent soundcard.
What matters for accuracy at low frequency is channel to channel consistency between left and right channels. This can be easily verified in ARTA using 2FR loopback measurement. Most any half decent hardware these days will have very good consistency here, just stay away from cheap junk like Behringer UCA202. Use sine sweep and drop the start freq right down to 5Hz. Noise signal is not good enough for low frequencies. Shorted lead calibration must be completed to remove the lead resistance. The software will estimate DCR by following the slope of the impedance curve from 5Hz to 0Hz. I find this method matches my Fluke meter with enough accuracy to not bother with using a multimeter for DCR determination. Anything I've measured is within 0.1 ohms, which provides 3.3% of precision measuring a 4 ohm driver with 3 ohm DCR, more like 1.5% error for an 8 ohm driver. It may even be preferable to estimate T/S with DCR at 0.1 ohm on the high side, as cable resistance from amp to speaker is often not included in many people's simulations. If we're talking about small DCR of inductors, again I find the results from this method more than adequate.
Let's do a bit deeper dive into the importance of precision here. I'm not trying to discourage your choice to use a handheld LCR meter, but just want to put some context to the importance of precision in driver Re. It may be worthwhile to determine T/S using measured Re, and then recalculating adding in a small resistance to create some error. Here I simply used ARTA, and the well calibrated impedance sweep to 5Hz, but manually entered Re for T/S calculation a bit on the high side. 10% of error here seems a bit extreme, resulting in 6.5 ohms instead of 5.9.
Speaker parameters with the correct Re:
Fs = 46.36 Hz
Re = 5.90 ohm[dc]
Le = 204.89 uH
L2 = 1245.32 uH
R2 = 5.28 ohm
Qt = 0.40
Qes = 0.44
Qms = 4.63
Mms = 14.59 grams
Rms = 0.916492 kg/s
Cms = 0.807904 mm/N
Vas = 10.25 liters
Sd= 95.03 cm^2
Bl = 7.550930 Tm
ETA = 0.22 %
Lp(2.83V/1m) = 86.91 dB
Added Mass - Constant Bl Method:
Driver unbaffled
Added mass = 14.00 grams
Membrane Diameter= 11.00 cm
Speaker parameters with imprecise/incorrect Re:
Fs = 46.35 Hz
Re = 6.50 ohm[dc]
Le = 204.33 uH
L2 = 1051.16 uH
R2 = 4.74 ohm
Qt = 0.45
Qes = 0.49
Qms = 4.68
Mms = 14.46 grams
Rms = 0.900419 kg/s
Cms = 0.815334 mm/N
Vas = 10.34 liters
Sd= 95.03 cm^2
Bl = 7.448244 Tm
ETA = 0.20 %
Lp(2.83V/1m) = 86.03 dB
Added Mass - Constant Bl Method:
Driver unbaffled
Added mass = 14.00 grams
Membrane Diameter= 11.00 cm
The dashed line is the correct Re of 5.9, the solid line is the incorrect 10% error of 6.5 ohms. Sensitivity is lower due to higher Re, otherwise the simulation is rather similar.
Yep, that's almost a db over a wide bandwidth - that's a fairly critical difference when level matching.
Use the right tools.
Yes, we’ll 10% error here is quite extreme, like I said 1.5 - 3 % (within 0.1 ohm) is more realistic form my experience. If your measured impedance is off by 0.5 ohms you should really have a close look at your calibration process. That impedance error throughout the entire sweep will affect your crossover design beyond just T/S values for the box model, an LCR meter won’t save you there. Level matching should be done by acoustic measurement, not T/S.