EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Steve Zarg on October 08, 2014, 11:25:33 pm
-
Hello,
I was working with an audio speaker today that wasn't working. I used my Fluke 179 on it to get an idea of the DC resistance and the reading fluctuated wildly.
I then used my DER DE-5000 LCR meter and found that it would show stable "R" readouts at each selected frequency, while the results varied with a trend of increasing "R" values at lower frequencies.
It made me wonder why the Fluke readout was fluctuating so much. Is it possible it uses a test signal that is some sort of sweep through frequencies? I guess I imagined the Fluke used DC to test resistance.
Does anyone have any ideas?
Thank You.
-
Out of curiosity, I grabbed a speaker and attached a Fluke 87 to it.
It did the same as yours and widly fluctuated on the reading.
I thought it might be since the coil and magnet are reactive to air movement around it and simply put, affecting the reading.
I suspect the induction of an a/c signal causes the meter to go :wtf:
Placing the speaker cone face down on a flat surface stabilized the reading.
-
Thanks for the reply.
-
If I remember correctly, multimeters often use pulsed signals to measure resistance. This is to save battery life and prevent heating of the resistor. The frequency and voltage of the pulses typically vary with the range song the meter is in. Unfortunately, pulsed signals and inductors don't interact well together. You will at best, get readings that confuse the meter or at worst actually turn the system into an oscillator. What is likely happening is the former. This is not uncommon. The meter tries a range, sees a load outside this range and switches ranges. This range has a different frequency and voltage, which means the meter sees a completely different load (due to the relationship between inductors and frequency). The LCR is not affected because it uses a constant frequency and measures attenuation of said frequency. Actually, it does more than that but for the sake of this discussion that description is enough.
Sent from my LGLS990 using Tapatalk
-
To measure resistance, most DMMs pass a small constant DC current (mA, uA) through the DUT, measure the voltage across it and deduct the resistance value by solving according to Ohm's law.
They don't have any problem measuring a speaker coil resistance fairly accurately (8 Ohm).
-
Speakers are not resistors. They are reactive devices that also produce energy output from mechanical input. When you try to measure a speaker with a multimeter, the speaker is also producing an output from the sound of the ambient environment corrupting the measurement. The LCR meter uses a different method of measuring the resistance and you will get a better reading. But because of the reactive nature of the speaker, it will vary with frequency.
-
Unless you try to measure your speaker at a rock concert, in a lab, ambient noise is too small to move the coil through the magnet and corrupt your multimeter readings by any significant amount..
An LCR meter can't give accurate impedance readings because the test current is AC, inducing errors as it passes through the coil in a magnetic field.
As an example, measuring the inductance of a 0.5W 8Ohm speaker at 100Hz gives an inductance of 440uH and a reasonable ESR of 8 Ohms. At 100kHz, the inductance has gone down to 55uH and the ESR up to 24.3Ohms.
The coil DC value, measured with a DMM (4W) gives a resistance of 7.8Ohms.
-
To verify the speaker nominal impedance accurately, the LCR meter should use the same frequency as the one used by the manufacturer to characterize his product, which usually gives an impedance 10-20% higher than the DC result.
This is the impedance of an 8 Ohm Bose speaker for a wide audio frequency range:
(https://www.eevblog.com/forum/testgear/question-about-fluke-179-and-resistance-measurement/?action=dlattach;attach=112502;image)
-
I did the same test with my Fluke 179 on a small 8? speaker and the reading was perfectly stable.
-
Perhaps Steve should try to measure a good speaker with his Fluke 179 to ensure proper DMM operation.
-
Unless you try to measure your speaker at a rock concert, in a lab, ambient noise is too small to move the coil through the magnet and corrupt your multimeter readings by any significant amount..
Try it you might be surprised. One of my multimeters puts about 1.5mV across the coil of a 4 ohm speaker while measuring resistance and showing 3 digits of resolution.
Flipping to ac volts and handling the speaker by its frame it was quite easy to get 0.1mV rms (which would be around 200 counts pk-pk on the resistance reading).
-
FWIW, your meter puts a current through the coil, not a voltage. About 350uA in your case.
Speaker sensitivity and low DMM test current could influence readings with coil movement, but not enough to compromise a rough test of resistance.
My 8Ohm 0.5W speaker resistance reading barely moved with loud noises in front of it or tapping on the membrane, with a test current of 1mA.
Your AC volt test bears no relevance here.
-
FWIW, your meter puts a current through the coil, not a voltage. About 350uA in your case.
Kind of hard to put a current through a coil without also putting a voltage across it FWIW.
My 8Ohm 0.5W speaker resistance reading barely moved with loud noises in front of it, with a test current of 1mA.
8 ohms and 1mA would make it about 6 times less sensitive to voltages induced in the speaker coil than mine.
Your AC volt test bears no relevance here.
Of course it is relevant. AC voltage induced in the coil sums with the dc voltage from the meter test current and the voltage is what the meter is measuring to display resistance. Higher frequencies (which is probably all you would get from a 0.5W speaker) would get averaged out by meter. I was measuring a 50W speaker with response down to 50Hz.
-
Of course, but the constant current is the clincher here. The varying voltage is an indication of the resistance under test.
Like I said, a lower current value, makes resistance measurements more sensitive to coil movements.
You don't say what meter you are using, but measuring low AC voltage is tricky, due to errors in the true RMS converter. Usually only valid from 10% of range, with crest factor 3 or less.
The bottom line is that it is not difficult to measure the resistance of a speaker coil with a multimeter, to find out if the coil is broken or to verify the nominal impedance of a speaker.
-
The bottom line is that it is not difficult to measure the resistance of a speaker coil with a multimeter, to find out if the coil is broken or to verify the nominal impedance of a speaker.
The OP said "fluctuated wildly".
In a quiet room with the speaker on the bench the meter read 3.88 ohms. If I picked the speaker up by its frame I had difficulty holding it still enough to get less than 10 counts of noise on that reading.
-
Your measurement of 3.88 Ohms DC resistance (Re) is fine for a voice coil of a 4 Ohm impedance speaker.
The OP tested a broken speaker.
Some manufacturers will give the value of Re for their speaker voice coils:
-
Just to see how much the noise would influence the DC-Ohms reading, I hooked a 15 cm diameter 8 Ohm speaker up to the Agilent 34461A in 2W OHM measuring mode and captured the displays in 3 stages:
1. Speaker covered with a big book
2. Speaker open with ambient lab noise
3. Speaker slightly moving on the table
Here are the results:
-
Nice. Do you know the value of the test DC current on the Agilent 34461A?
-
The 34461A has 1mA DC current in this range, measured with a 34401A.
-
A DC current through the coil doesn't affect the readings. Any movement of the coil trough the magnetic field would, to some extent. So the coil should be measured with the speaker lying flat, free of disturbances.