How to determine output wattage of an amp

[wsscott]

Let's say you have an amp that is supposed to have an output at the speaker of 15 Watts.  It's referred to as a "15 Watt Amp".  Like the Fender 5F1 Champ is supposed to be a 5 Watt amp.

How do you test the amp to see if the amp is in fact putting out 15 watts or 5 watts, or whatever in fact the amp is producing?

[scstill]

https://www.tdpri.com/threads/measuring-amp-wattage.551914/
in this link there is a YouTube video of a guy using a voltmeter across speaker as he plays

[kagliostro]

Conditions needed to use a non-true-RMS multimeter:
The output signal must be a pure sine wave
→ No music, only a test tone at 1 kHz (from a signal generator or a phone app).

The signal must not be distorted
→ Raise the volume just before clipping. A distorted wave gives inaccurate readings.

Use a known resistive load (e.g., 8 ohms)
→ A real speaker complicates things due to varying impedance with frequency.

 Power calculation with voltage measured from a multimeter:
Measure the AC voltage at the amplifier output (connected to the dummy load). Then use this formula:

P (W) = (V_RMS × V_RMS) / R
or simply:
P = V² / R

Where:

P is the power in watts

V_RMS is the root mean square (effective) voltage in volts

R is the load resistance in ohms (Ω)

Franco

[wsscott]

Franco-thanks.

I have a simple digital oscilloscope that reads RMS voltage.  Could I connect that to the output speaker jack on the amp, and when the sine wave starts to clip, see what the RMS voltage reading is?  Then I would just square that number and divide it by 8 or 4, etc?

Does it matter what the 1K test tone's volume level and the amp's input jack is?

[Merlin]

I have a simple digital oscilloscope that reads RMS voltage.  Could I connect that to the output speaker jack on the amp, and when the sine wave starts to clip, see what the RMS voltage reading is?  Then I would just square that number and divide it by 8 or 4, etc?

Does it matter what the 1K test tone's volume level and the amp's input jack is?

Since your scope is giving you true RMS, you can measure the voltage across the dummy load at any level you like, clean or clipped, and the power calculation will be correct for that level (for a resistive dummy load).

A non-RMS meter will only give correct readings on a fairly good sinewave, and will become innacurate with a very clipped signal

[glass54]

+1 for Franco and Merlin.
I like to use typically 25mV RMS (max of 50mV) at input, making sure that I have no distortion at V1 output, prefer to have TCs "at neutral" ie mid position, then proceed as Franco said. I have a nice digital Oscilloscope with True RMS, so i/p and o/p RMS is easy.
I also like to observe max Power Out at 100Hz and 5kHz, if nothing more than academic BUT its good to know you have a full power band for your guitar input signal. You can also check feature  like max Bass out, etc.
By the way, many later Fender Sch indicate AC signal levels (1kHz) and you will note that the input is around 25-30mV RMS on many.
Regards
Mirek

[wsscott]

I wanted to test my amp's output given this information using a "Dummy Load" with a Scope output jack that I built from D-Lab's video.  Attached is a link to the section of that video showing the schematic: 

&t=412s

But when I use the scope with this box, I get a beautiful sine wave and as I increase the volume of the amp it stays beautiful and the Vrms increases, but the Vrms of the Output without clipping maxes out at either 4 or 8 watts, based upon which amp and impedance of the speaker in the amp.  And these amps have a much higher rated output than what the Vrms on the scope shows.

What's going on?

So, do I have to connect the scope's probe directly at the speaker jack?  Is the circuit in this Dummy Load box such that I can't use it with the scope for measuring Vrms even though I get a proper sine wave?

[shooter]

these amps have a much higher rated output than what the Vrms on the scope shows.

clean watts n max output/power are 2 deferent measurements, so make both measurements.
most of my "12 w" amps top out at 4-6W "clean"


 "power out" ratings are used as a "selling point", to skirt around lying companies will list power out in peak-to-peak, cuz, well, it's a bigger number 

[stratomaster]

I wanted to test my amp's output given this information using a "Dummy Load" with a Scope output jack that I built from D-Lab's video.  Attached is a link to the section of that video showing the schematic: 

&t=412s

But when I use the scope with this box, I get a beautiful sine wave and as I increase the volume of the amp it stays beautiful and the Vrms increases, but the Vrms of the Output without clipping maxes out at either 4 or 8 watts, based upon which amp and impedance of the speaker in the amp.  And these amps have a much higher rated output than what the Vrms on the scope shows.

What's going on?

So, do I have to connect the scope's probe directly at the speaker jack?  Is the circuit in this Dummy Load box such that I can't use it with the scope for measuring Vrms even though I get a proper sine wave?

Which amp is it?  If you put a large enough signal in you'll have preamp clipping that will obscure the output power measurement.  If your amp has a cold clipper, the effect is even worse. You'll have to inject signal later in the path to evaluate the output power.

I try to keep my input signal no greater than 120mV. I also look at 400Hz and 1kHz.  The measurements at these two frequencies are often close, but are different much of the time.

[wsscott]

It's a Fender Champ 5F1 that measures 4 watts clean, and a Twighliter Mod that measures 8 watts clean.

The Fender has a 4 ohm speaker, and the Twighlighter has an 8 ohm speaker.

The signal is around 150 Vrms at 1K.

[stratomaster]

It's a Fender Champ 5F1 that measures 4 watts clean, and a Twighliter Mod that measures 8 watts clean.

The Fender has a 4 ohm speaker, and the Twighlighter has an 8 ohm speaker.

The signal is around 150 Vrms at 1K.

mV you mean?

What is the actual voltage you measure across your dummy load?

[pdf64]

... do I have to connect the scope's probe directly at the speaker jack?  Is the circuit in this Dummy Load box such that I can't use it with the scope for measuring Vrms even though I get a proper sine wave?

Dunno, who can be bothered to watch a feckin video, please provide a schematic of the load box if you want that answering.

Whatever, measuring at the amp's speaker output avoids doubt / uncertainty. So why not do that?

[scstill]

Very interesting approaches and good technical discussion which I appreciate.

But my question is kind of "so what" if we have measured our amps output power.
Other than having a precise metric what would we do differently? change the design? have bragging rights?

And what would be the difference with the power measurement, calculating power via Load line or even the value provided on the tube datasheet.

[pdf64]

..  my question is kind of "so what" if we have measured our amps output power. ...

It might be seen to be the difference between taking this seriously and living in la-la land 

[scstill]

I have designed several amps from scratch and have another (Prew42) currently in design.
Most sound pretty good with good operating parameters. Have never measured output wattage.
Guess someday I should do so.

BTW - has anyone used a switch on the OT to change the primary load resistance (my OT has two available) and possibly the cathode resistor if necessary? Could label the switch Power - More Power
I don't think there will be a noticeable change because power doesn't necessarily mean more volume.

[Merlin]

It's a Fender Champ 5F1 that measures 4 watts clean,

That's normal for a Champ.  I have no idea what a Twighliter Mod is.

[passaloutre]

Just a warning so you don't make the same mistake I did: a 50W "attenuator" will not last long with a 50W continuous square wave coming out of your power amp when trying to measure maximum output power. It was fine for measuring clean power, but it gave up the smoke with the volume dimed. Now I use a 200w resistor.

Be sure to overbuild your dummy load, or pulse your input signal only long enough to make the measurements you need.