Measuring actual amplifier power output

[shaggerz]

Thought I'd start another thread rather than hijacking the 'post your car audio setup' thread...

I have read a few things about measuring the actual usable output power of an amp.

Tools needed: Oscilliscope, multimeter, sine wave generator or other source of 50Hz test tone, non-inductive dummy load capable of handling at least the expected power output of the amp.

The idea is to connect the dummy load to the amp output (it should be 4 ohms or whatever resistance you want to test output power for) and play a 50hz tone through it. You want to measure the sine wave at the output with the oscilliscope and turn the volume up until just before the waveform shows visible clipping (flattened peaks). You can then measure the voltage and current at the output (this is why pick a 50hz tone... use the multimeter to measure AC voltage) and use ohms law to work out just how many 'watts' are being produced.

Biggest problem with this method is that oscilliscopes are DAMMED EXPENSIVE!. Other problem may be that it doesn't work. Can anybody (12VoltKing) confirm or deny this would work?

Or better still, a CHEAPER way to test?

[michaelw]

It might work except for a few things.

The AC response of a speaker is different to a resistor, personally if I were to do this I would use a speaker and a scope with a current clamp.

Your ear does not hear in electricity. This test ignores the efficiency of the speakers.

Why 50Hz?

It might be that the above is the standard for amp testing but other than bragging rights I do not see much point in the excercise.

If you want to test your system I think you would be better off buying a dB meter from Jaycar, Dick Smith et. al

Mike

[shaggerz]

I do not care about sound output. I want to know amp power.

1. You can NOT measure with a speaker because it is an INDUCTIVE LOAD and will give a false reading.

2. 50Hz because that is the frequency that normal AC oscillates at, therefore it is the frequency that the multimeter expects to see. Not 100% sure if this matters, but 50Hz is within the audible range and within all amps frequency response ranges and shouldn't matter.

3. When selecting an amp, which of these would you choose?

a) Audioline 2x400 watt
b) Alpine 2x150 watt

It's pretty easy to look at the numbers and assume the Audioline would cane ass, but unfortunately thats not the case at all.

Wouldn't it be nice to be able to measure the output power so you could say 'the audioline is actually capable of 2x75 watts and the alpine is actually 2x190 watts'

The way it stands, without measuring, nobody has any idea just how powerful any given amp really is, so we all speculate and guess and ultimately make a decision based on either false numbers (those people who buy 'Boss' and 'Audioline') or the brands reputation in the hope that they were being modest about the capabilities of their product.

[Tasmaniak]

[QUOTE=shaggerz]
2. 50Hz because that is the frequency that normal AC oscillates at, therefore it is the frequency that the multimeter expects to see. Not 100% sure if this matters, but 50Hz is within the audible range and within all amps frequency response ranges and shouldn't matter.

QUOTE]

A DMM doesn't expect to "see"any particular frequency. It doesn't count how times the polarities swap sides. Unless, it's an automotive meter that has a tacho on it and it can count the pulses from the coil. Although, this isn't true AC.

Generally, as Michael described, you are better off using a clamp and pen and paper and do some simple maths. Sure, I've used a cro before to measure power output at a particular frequency, but that is when I'm searching for resonant (Qhz) Frequency of a speaker.

[michaelw]

To address your points in order.

1. I know a speaker is an inductive load, why bother quoting an amplifier power on an artificial load. The reactive component of the speaker load is but a part, there will exist (esp at low frequencies) a siginificant fraction of the load being resistive.

2. If you are going to use a frequency I would use a few, music is at more than one frequency. Your comment about dmms 'expecting' 50Hz is wrong, besides we were talking about cros here.

3. If I choose an amp based on one number I would be one of those people who are born every minute. Look at all the specs of the amp, distortion and frequency response would be more important. That is if you want to listen to music rather than make noise.

Mike

[Tasmaniak]

I find a really good spec to pay attention to on an amplifier is "slew". you can look upon the slew rate of an amplifier a bit like the acceleration rate on a car. e.g. No use having a car that can do 180mph if it takes all day to get there. same as having an amp that can 1000wrms if it takes a full second to produce it(all amps are faster then that but I used it as an example)

[shaggerz]

michaelw you make an excellent point there.

Would it stand to reason that one could construct a graph of power vs frequency using this method?

I would love to see amps come out with a graph like this on the box. That would be fantastic

When buying my own amp I looked at power output and THD as the 2 main factors. My decision ended up being between Cadence and JL Audio, but because JL had much better THD for their rated power (albeit somewhat lower) I went with them over cadence. Stats below.

JL Audio e1200 Class A/B Monoblock Full-Range Amplifier Specifications:

Rated Power 12.5 V: 120 W RMS x 1 @ 4 ohm
200 W RMS x 1 @ 2 ohm
Rated Power 14.5 V: 175 W RMS x 1 @ 4 ohm
275 W RMS x 1 @ 2 ohm
THD at Rated Power: 12.5V: <0.08% THD, 20 Hz - 20 kHz
14.4V: <1% THD, 20 Hz - 20 kHz
S/N Ratio: >104 dB referred to rated power (A-weighted, 20 Hz – 20 kHz noise bandwidth)
Frequency Response: 10 Hz – 25 kHz (+0, -1dB)
Damping Factor: >200 @ 4 ohm / 50 Hz
>100 @ 2 ohm / 50 Hz
Slew Rate: ± 22V / µs
Input Range: 200mV – 8V RMS
Dimensions (L x W x H): 9.80 in. x 9.25 in. x 2.36 in.
250 mm x 235 mm x 60 mm

My favourite words in there are 'THD at rated power'. I would take this to mean that if one was willing to accept higher levels of distortion (I think 0.08% is quite low) then you could get more power out of it easily.

'clip point' would be an interesting figure as well.

Is it possible to design a 'clip indicator' that you can attach to an existing amp? or is this too complex?

[Tasmaniak]

Shaggerz, bear in mind though that the human ear doesn't detect distortion until it reaches 10%

[shaggerz]

Or in my girlfriends case, about 50% lol...

"for gods sake turn that down. Can't you hear the speakers crying?"

"no? sounds fine to me."

bah!!!

[michaelw]

You can if you want to build a clip meter. I would just put in polly switches to protect you speaker from it in the first place. Go look at jaycar and read up on them.

You hook them up in series with your speakers, basicaly what they do is when dc is across them they (clipping), the resistance goes up therefore lowering the power to you speakers. When they cool off they go back to normal. I would say they are good insurance for your drivers.

Mike

[shaggerz]

Have read about those and also see a kit on jaycar site for 'loudspeaker protector' which seems to serve a slightly more high tech purpose and could presumably be adapted to flick on a clip light when speakers in trouble.

with polyswitches, can you connect them in parallel to raise their trip current?

struggling to find data sheet.

[michaelw]

Don't parallel protection to boost current. It rarely works as planned due to slight differences in components.

Link is on the jaycar primer page.

http://www.jaycar.com.au/images_uploaded/polyswit.pdf

Mike

[michaelw]

I have to make a slight correction. Polyswitches will not protect you from clipping (my bad) they are for over-current. The two are related but not exactly the same. polyswitches are still good for speakers.

Sorry about that.

Mike

[shaggerz]

What do you think of the loudspeaker protector kit? I think it would be a good idea to whack one of these on the output of my sub amp. Killing the connection the moment clipping is detected should be a very good way to prevent premature sub death

I wonder if it could be modified to simply light an LED as long as clipping is detected.

[michaelw]

I like the idea. For the price I don't think it is worth going to the effort in redesigning it. With regards to lighting an LED I assume all you need to do is take the feed off the relay. Personally I would not do this as by the time you have noticed the led your speaker could be toast.

Most clip detectors are part of the amplifier as all that clipping that is, is the output swinging to rail voltages, thus you need to know what the rail voltage is to detected it, hence you need to get into the amp. The other way (and how I suspect the kit does it) is to have a low pass filter of some type that only allows DC through. When dc is detected something must be up so disconnect the speakers.

Mike

[jds303]

you can use a CRO to measure the maximum unclipped output from and amplifier, and i believe this is the most accurate way to determine this measurement. you get an actual tried and tested output power measurement with minimal distortion, rather than relying on manufacturers specifications, which are often incorrect (to ensure their product does approximately meet quoted specification).
you could use a clamp meter and ohms law and do some calculations to get these values at a given load. sure, they'll be correct, but can't as accurately detect the beginning of clipping.
50Hz is the frequency at which australian mains power (240V) operates, whereas in the US, the mains operates at 60Hz.
a 50Hz test tone is just a widely used frequency.

true CRO's are expensive, so borrow one from work. who's gonna know?
that's my 2c.