power output, can it be measured?

subdude67 · #1 07-27-2011, 03:04 AM

is there a device that you can connect to an amp's output so that the actual output in watts that it produces can be measured?

johnk_10 · #2 07-27-2011, 03:18 AM

sure. you can measure it with a signal generator, a dummy load (resistor capable of handling your amp at full output), and volt/ohm meter set to AC volts, then multiply the volt meter's reading times itself, and divide it by the resistance of your speaker (ohms). in other words if your amp is putting out 20 volts into a 4 ohm load, it would be 20x20 (=400) divided by 4(ohms)=100watts.

i usually simultaneously view the output on a scope to see when the signal (sine wave) is starting to distort.

i also use my compu-bias which makes it really easy: Compu-Bias™ Computerized Tube / Valve Bias Meter

staindbass · #3 07-27-2011, 05:29 AM

i use old radio shack / realistic power meters from ebay originally made for stereos. the method above is probably more accurate. i assume you could do the above measurement with the meter attached to the speakers.

billfitzmaurice · #4 07-27-2011, 05:53 AM

Quote:

Originally Posted by subdude67

is there a device that you can connect to an amp's output so that the actual output in watts that it produces can be measured?

You can't measure watts, you can measure volts and then extrapolate the power via Ohms Law. You can't measure it with your bass as the signal source, as the voltage reading is not a constant. If you have a peak hold meter you can measure the peak output with the bass as a source.

craig.p · #5 07-27-2011, 09:05 AM

I'd not trust any power measurement I made without at least a scope so that, as John K has already said, I could see where the sine wave begins to ""go wobbly." In this context I wouldn't trust a voltmeter alone for jack.

wcriley · #6 07-27-2011, 09:33 AM

True if your only interest is "How much power can my amp produce?"
Bill's method of using a peak hold meter would answer the question of "How much power is my amp actually putting out to achieve my desired SPL?"
I think a lot of bass players would be surprised by the answer to the second question.

JTE · #7 07-27-2011, 09:47 AM

Yes, as noted above it CAN be done. When Alembic wrote the gear reviews in Guitar Player Magazine (in the late '70s and early '80s) that's one of the many things they did. They'd measure the actual output (at a consistent THD for every review) and compare it to the advertised rating. They basically took none of the manufacturers' ratings as true, but checked all of them and noted differences between the specs and what they measured.

I miss those reviews...

John

beans-on-toast · #8 07-27-2011, 09:48 AM

There are inline power meters like this one, available for instrument amps.

These sort of meters aren't the most accurate. If you want to monitor your output as a means of protecting your speakers, if you understand the limitations of the device, maybe it's good enough.

nashvillebill · #9 07-27-2011, 09:51 AM

Quote:

Originally Posted by craig.p

I'd not trust any power measurement I made without at least a scope so that, as John K has already said, I could see where the sine wave begins to ""go wobbly." In this context I wouldn't trust a voltmeter alone for jack.

Exactly, a scope would be the bare minimum requirement; a distortion analyser would be even better.

When I test an output, I hook up the dummy load with an ammeter in series (to get current) as well as the voltmeter (to get voltage). This takes the changing resistance of the dummy load (as it heats up) out of the equation. I connect a signal generator and I crank the amp up, I watch the scope, I see the signal clip and i record my current and voltage reading. Then P=V times I.

However, if I had no way of knowing when the signal was starting to clip or distort, I could easily keep on, overdriving the amp--meaning that any numbers I had then would be totally meaningless.

To explain this for those readers who are not familiar with amps: Amplifiers take an incoming voltage signal and multiply it (the amplifier's gain). The incoming signal voltage times gain equals the output voltage. The amplifier's maximum output is reached when the limitations of the output devices (whether tubes or transistors) are reached, or the limitations of the amp's internal power supply are reached, whichever comes first.

This limit isn't necessarily found when the Volume knob is cranked all the way up, it's usually reached well before that, when the incoming signal has become strong enough so that the incoming signal times gain has reached the maximum output level. It's therefore dependent on the incoming signal strength--as well as the Volume knob position.

johnk_10 · #10 07-27-2011, 10:07 AM

the OP's question was if there was a device that can measure the power of an amp, and there is. without buying a scope, the compubias with their 'watt probe' does it perfectly, and it measures it without any distortion. you'll still need a signal source (sine wave) and a dummy load, but you could use the speaker cab and a signal generator to do it.

nashvillebill · #11 07-27-2011, 10:35 AM

Gotta respectfully disagree, the compubias does NOT do it perfectly--far from it. I have read their watt probe instructions in their manual.

They assume you are not running the amp into clipping, which they admit the meter has no way of knowing. You could crank the amp well into clipping and the meter would not know that, it would just keep indicating "more power."

They also cannot take into consideration the changing impedance of the speaker's reactive load. They describe how they are measuring AC voltage (filtering and rectifying it) and it appears from their description they are using a voltage divider network to derive a guess at the speaker load, then using that derived number to calculate power. Hardly accurate by any stretch of the imagination.

Again, measuring an amp's output without knowing when it is clipping is meaningless.

subdude67 · #12 07-28-2011, 02:01 AM

thanks to all for your time, awesome crowd here on TB!

allthough i basically understand the procedure, allow me a few more questions.

what do i use as a signal generator? do i feed a certain frequency or do i send pink noise into the amp?

what do i use as a dummy load?

where do i measure the voltage once the dummy load is connected?

thanks again and have a wonderful day.

johnk_10 · #13 07-28-2011, 02:14 AM

Quote:

Originally Posted by subdude67

what do i use as a signal generator? do i feed a certain frequency or do i send pink noise into the amp?

you just need a clean sine wave, and it's good to have a few frequencies, but 400hz-800hz will work fine.

Quote:

Originally Posted by subdude67

what do i use as a dummy load?

either a bank of non-inductive resistors, at the impedance that you amp is safe to run at, that can handle you amp's full output, or even your speaker cab, but it can get loud doing it that way.

Quote:

Originally Posted by subdude67

where do i measure the voltage once the dummy load is connected?

across the + and - leads on the output jack.

Razman · #14 07-28-2011, 07:06 AM

I would be more interested in testing while playing with a peak multimeter to see how much of our SVT-6's power is being used. It replaced my Trace rig (and we all know that Trace watt ratings are conservative) but we don't go much louder.

Downunderwonder · #15 07-28-2011, 05:56 PM

You only need a dummy load if your amp is a valve amp or some weird other amp. Being a v-amp squad member I'm guessing yours is a solid state amp? If so. you don't need anything more than a sinewave and voltmeter to measure.

So long as the amp isn't clipping, the volts translate to watts with a square root function. I've forgetten it but you can find it on TB easily!

Once the amp clips the pure signal it changes the maths, it's no longer a RMS of a sinewave, once the wave is fully clipped it's 1.4 times the RMS from the same peak of voltage. Depending on how fancy your voltmeter is you could get the right wrong answer or wrong right answer if it has "true rms". Good luck with that.

Anyway, you can get a sinewave signal from Audacity computer programme into your amp input from computer soundcard or make a CD of the tone. I'd use 150Hz.

Downunderwonder · #16 07-28-2011, 06:02 PM

btw, just how big is your amp? 300W is about 35V RMS at 4ohms if I remember right.

Be careful with the exposed voltage at your speaker cable or wherever you are measuring. You don't want to short across the voltage as it will do very bad things to the amp, making sparks and letting out the magic smoke.

nashvillebill · #17 07-28-2011, 07:23 PM

Here's some "Gotcha's" that makes it difficult to measure power, especially if you are just measuring voltage or trying to measure as you play.

The formula for power is P = (V squared) divided by R where V is voltage, R is resistance, and P is watts.

If you aren't really sure where clipping is , and you just run it up somewhere without a scope or distortion meter...let's say you aren't sure if it started clipping at 30 volts or 35 volts, into a 4 ohm load. Doesn't sound like it would matter much does it? But 30 times 30 is 900, divided by 4 is 225 watts. Whereas 35 times 35 is 1225, divided by 4 is 306 watts. That's rather a significant difference wouldn't you say?

For true power readings, a constant frequency, constant amplitude sine wave is used. But if you are trying to measure power instantaneously, the amplitude can change drastically in a few milliseconds . Whether a meter with an old-fashioned needle, or newer LED's, no power monitoring circuit can react THAT quickly. So you might think you're cruising along not using much power, but the transient peaks are hidden out of reach of the power meters.

And a problem if using speakers. They are reactive loads. Dummy loads are purely resistive, they present the same load regardless what frequency is presented to them.

But a speaker has inductance plus mechanical loading. It's impedance is the DC resistance plus the reactive component (the inductance) which is not constant for a given frequency, and the mechanical loading may not be constant for varying volume levels. Look at a speaker's impedance curve, it goes all over the place. The Eminence Delta 12B, a 16 ohm nominal woofer, has a DC resistance of 11.4 ohms. But its impedance goes from 14 at 20 Hz, to 170 at 58 Hz, then back down to 8 at 225 Hz, then back up to 30 and rising at 2000 Hz. Hardly consistent enough to use for measuring P.

Bass_Pounder · #18 07-28-2011, 07:28 PM

Quote:

Originally Posted by Razman

I would be more interested in testing while playing with a peak multimeter to see how much of our SVT-6's power is being used. It replaced my Trace rig (and we all know that Trace watt ratings are conservative) but we don't go much louder.

It takes 10X the power to get twice as loud.................

If there is not a very large difference in the ratings of your Trace and the Ampeg, it will not make that great a difference in your percieved volume.

BassmanPaul · #19 07-29-2011, 11:42 AM

Quote:

Originally Posted by Razman

It replaced my Trace rig (and we all know that Trace watt ratings are conservative) but we don't go much louder.

As far as I can see Trace watts, as in the model number, are doubled. An amp with 300 in its model number has actually got an output of 150W RMS.

Downunderwonder · #20 07-29-2011, 08:27 PM

Quote:

Originally Posted by BassmanPaul

As far as I can see Trace watts, as in the model number, are doubled. An amp with 300 in its model number has actually got an output of 150W RMS.

Eh? No way Hoss.
All the older solid state ones put out RMS "model number wattage" at a 4 ohm load. The transient peaks get produced verrrrry well though, hence the reputation for punching above their weight. Trace rated them for peak power at 2x the RMS on the back panel.