Some guidance w/b greatly appreciated.
I have a 3-way speaker with drivers (without crossover) and need to test frequency response of the 3 drivers in cabinet. I assume that the objective is to measure each using the same output voltage since the differing impedances, efficiencies and crossover goals will cause different loudness.
I bought an oscilloscope to do this and can see sine waves no problem and when driving my cheapo Lepai class D amplifier to 1/3 volume gain dB I can measure 2.83 volts on a simple DMM at 60Hz. But using the oscilloscope and my Denon receiver (AVR4306) I get a measurement in the millivolts Irregardless of the gain setting .. I'm going crazy.
I tried taking measurements without loads, dummy loads of 4 & 8Ohm and with drivers loaded. Same results. The only thing I didn't try was hooking the scope ground lead to actual ground as opposed to across the output terminals of the amps. But that doesn't explain why it worked on the Lepai with a simple DMM.
I did manage to eventually blow the Lepai with an accidental short. So now I'm reluctant to measure the Denon any further.
Could someone offer me with a simple sketch of the oscilloscope, amplifier & driver circuit so I can feel confident about that anyway?
Thanks in advance
Mike Wittman
I have a 3-way speaker with drivers (without crossover) and need to test frequency response of the 3 drivers in cabinet. I assume that the objective is to measure each using the same output voltage since the differing impedances, efficiencies and crossover goals will cause different loudness.
I bought an oscilloscope to do this and can see sine waves no problem and when driving my cheapo Lepai class D amplifier to 1/3 volume gain dB I can measure 2.83 volts on a simple DMM at 60Hz. But using the oscilloscope and my Denon receiver (AVR4306) I get a measurement in the millivolts Irregardless of the gain setting .. I'm going crazy.
I tried taking measurements without loads, dummy loads of 4 & 8Ohm and with drivers loaded. Same results. The only thing I didn't try was hooking the scope ground lead to actual ground as opposed to across the output terminals of the amps. But that doesn't explain why it worked on the Lepai with a simple DMM.
I did manage to eventually blow the Lepai with an accidental short. So now I'm reluctant to measure the Denon any further.
Could someone offer me with a simple sketch of the oscilloscope, amplifier & driver circuit so I can feel confident about that anyway?
Thanks in advance
Mike Wittman
IME this is best done with a voltmeter that has a fairly low voltage range. You don't have to use 2.83 volts out, but you do want to now what it is for reference. 60 Hz is a good choice, as meters are meant for that, but I've sound must digital volt meters are accurate up to 1 kHz and beyond.
A dummy load is a good idea, depending on the amp. The Lepai might have died if one side was connected to ground, as it is a floating output. At least the early ones were.
A dummy load is a good idea, depending on the amp. The Lepai might have died if one side was connected to ground, as it is a floating output. At least the early ones were.
Connect an 8 ohm resistor to the output. Use the scope probe to measure the voltage on each output terminal without connecting the scope probe ground lead.
I'd be very careful using a 'scope on an amplifier output - one side of the 'scope input will probably be grounded, and a class-D amp, for instance, will probably have a bridged output (most class A/AB amplifiers won't be bridged, but some may be). You risk damaging the amplifier by shorting an output stage to ground.
Bridged connection means that there are, in essence, two amplifiers, with outputs in mutual anti-phase.
Bridged connection means that there are, in essence, two amplifiers, with outputs in mutual anti-phase.
Does you scope have two inputs? If so use both channels and set it for A-B mode for a differential reading. Then set it to 1 volt per division and adjust the gain until a 1khz test signal is as close to 8 graduations as possible. 2.83V RMS = 8V p-p Your can use a single channel using the scopes chassis ground but your will need float the ground plug. I have a SpeakerCraft amplifier that does not earth ground the black speaker output and I don't need to float or A-B with it, but the cheap class D amp I bought for carrying outdoors shorted just like yours did, fortunately it wasn't fatal. I also noticed the SpeakerCraft delivers a nice clean sine wave and the cheapo class D delivers a fuzzy mess. They both sound OK.
For accuracy, a DVM is far better than the scope. As it is floating, it is safe with respect to ground.
So 2.83VAC is 1Watt rms into 8 Ohms. On a scope that would be an even 4V peak. You had problems connecting a scope to a class-D amp output? The first thing that comes to mind is that maybe you shorted the negative output to earth ground of the scope probe. Many, class-D amps are bridged so the negative speaker terminal is hot, half the output voltage. You can use two probes in differential mode, or you can just leave the scope ground unconnected and look for half the voltage (1.415V, 2Vpeak)) on one of the speaker terminals.
Mike Wittman,
You could float the oscilloscope by running it off a UPS meant for computers, printers and other office equipment. However, note that the UPS needs to remain unplugged from the wall outlet and therefore cannot charge its battery while you make measurements.
Alternatively, you could use an "instrumentation amplifier" whose output the oscilloscope can safely measure, even if the AVR happens to be bridged. A difference amplifier would also work, but not equally well, due to component tolerances and mismatches.
You could float the oscilloscope by running it off a UPS meant for computers, printers and other office equipment. However, note that the UPS needs to remain unplugged from the wall outlet and therefore cannot charge its battery while you make measurements.
Alternatively, you could use an "instrumentation amplifier" whose output the oscilloscope can safely measure, even if the AVR happens to be bridged. A difference amplifier would also work, but not equally well, due to component tolerances and mismatches.
for general purpose response testing, there is no need to set the output to a specific level.
Thank-you and yes, I believe this is what I did. I'll remember to turn amplifiers off before connecting in the future
Thank-you. Analog circuitry is not my specialty so I struggle with some of the concepts.
Thank-you. If found this interesting. I've read a bit on the different amplifier types but this is a very practical example.
Interesting. As noted I'm new to oscilloscopes so using it in A-B mode did not occur to me. Is there a way to test if the amplifier black terminal is earth grounded?
You're killing me 😉 I bought the oscilloscope becasue I assumed my DVM was inaccurate and distorting my readings.
Thank-you. I was definitely placing the probe ground on the other side of the dummy load resistor. Using the scope differential input then without connecting probe grounds seems the way to go the?
Now that's clever! After reading the responses I'm convinced the probe ground was messing with my reading and even dangerous to the equipment. Some YouTube's I watched explained this but I hadn't fully grasped the concept of a 'floating ground until your explanation...
Some things you can do safely...
1/ Connect the scope ground to the amp chassis (or input ground) and see if you have signal on just the positive speaker terminal and nothing on the negative terminal. Most conventional amps would be like that.
If you see the same signal on both speaker terminals then it is a bridged amp and the load sees twice the voltage... as one terminal goes up in voltage the other goes down by an equal amount.
2/ With the amp OFF and if you can measure a short (zero ohms) between speaker negative and chassis then its conventional. The scope ground lead can safely be connected to the speaker negative.
3/ If in doubt then just add a 47 ohm 0.25 watt in series with the ground lead of the scope. That way if there is any DC voltage present on the terminals (lke in some bridged amps) then the resistor lets you know (gets hot) and no damage is done.
Its also good practice if in doubt to just measure and see if there is any voltage between the point you want to connect the ground lead to and the scope ground. If there is voltage then don't make the connection.
1/ Connect the scope ground to the amp chassis (or input ground) and see if you have signal on just the positive speaker terminal and nothing on the negative terminal. Most conventional amps would be like that.
If you see the same signal on both speaker terminals then it is a bridged amp and the load sees twice the voltage... as one terminal goes up in voltage the other goes down by an equal amount.
2/ With the amp OFF and if you can measure a short (zero ohms) between speaker negative and chassis then its conventional. The scope ground lead can safely be connected to the speaker negative.
3/ If in doubt then just add a 47 ohm 0.25 watt in series with the ground lead of the scope. That way if there is any DC voltage present on the terminals (lke in some bridged amps) then the resistor lets you know (gets hot) and no damage is done.
Its also good practice if in doubt to just measure and see if there is any voltage between the point you want to connect the ground lead to and the scope ground. If there is voltage then don't make the connection.
Thank-you. What I was trying to do was volume match the 3 drivers in the cabinet at the crossover points. Since they're different efficiencies I thought given the same input the tweeter and midrange would be louder than the woofer and I was trying to avoid this. Could you suggest an alternative method?
Thank-you for all your suggestions. As a word of caution; this is the kind of screwup you get when a software engineer (me) meets the analog world. LOL