Power amps: continuous output power measurements

[StigErik]

When amps are tested here at ASR as well as other places, output power is one of the things measured. I can't however find any information about the duration of such a measurement. They are also conducted at 1 kHz only, if I'm not mistaking?

Now - consider we want our amps to drive home theater subwoofers covering the 10-80 Hz range. Typical action movie soundtracks contain a lot of more of less continuous low frequency signals at very high levels. An example I like to use is the scene from Interstellar where Matthew McConaughey's character fly into the black hole. In that scene there's extremely high level in the LFE channel around 25-30 Hz (I think - will check this), and it literally lasts for minutes. The first time I saw that movie at home, I had to pause it just to check if nothing had caught fire!

The question is - can our amps deliver high power at low frequencies for an extended period of time, which will be the requirement for home theater subwoofers?

I set out to do some experiments on this topic and get some answers. In the next post I will describe a proposal on how it could be done. Comments, corrections, input of any kind is more than welcome.

 

[Sokel]

When amps are tested here at ASR as well as other places, output power is one of the things measured. I can't however find any information about the duration of such a measurement. They are also conducted at 1 kHz only, if I'm not mistaking?

Now - consider we want our amps to drive home theater subwoofers covering the 10-80 Hz range. Typical action movie soundtracks contain a lot of more of less continuous low frequency signals at very high levels. An example I like to use is the scene from Interstellar where Matthew McConaughey's character fly into the black hole. In that scene there's extremely high level in the LFE channel around 25-30 Hz (I think - will check this), and it literally lasts for minutes. The first time I saw that movie at home, I had to pause it just to check if nothing had caught fire!

The question is - can our amps deliver high power at low frequencies for an extended period of time, which will be the requirement for home theater subwoofers?

I set out to do some experiments on this topic and get some answers. In the next post I will describe a proposal on how it could be done. Comments, corrections, input of any kind is more than welcome.

They are tested at other freqs too:



Tests lasts seconds,however some datasheets do state max duration.

 

[StigErik]

They are tested at other freqs too:

View attachment 379723

Tests lasts seconds,however some datasheets do state max duration.

Ah yes, that is correct.

But this one is only done at 1 kHz I think:

 

[amirm]

The analyzer will drive the amplifier long enough to get a stable reading. Likely tens of milliseconds at most. It will then drive it at higher and higher levels though. The gaps will be small. It will even push it into limiting as you see in the clipped portion of the graph. So the amplifier is being stressed.

I could just drive it continuously with a sine wave but that can damage the amplifier and possibly my dummy load.

Keep in mind that the crest factor of sine wave is also lower than what you get in music or even effects. What you perceive as bass lasting a while does not translate into continuous max power.

 

[amirm]

But this one is only done at 1 kHz I think:

It is but that test is a subset of the power sweeps at different frequencies. For sub usage, note the 20 Hz frequency:

 

[Sokel]

Ah yes, that is correct.

But this one is only done at 1 kHz I think:

Here you are:

 

[StigErik]

The analyzer will drive the amplifier long enough to get a stable reading. Likely tens of milliseconds at most. It will then drive it at higher and higher levels though. The gaps will be small. It will even push it into limiting as you see in the clipped portion of the graph. So the amplifier is being stressed.

I could just drive it continuously with a sine wave but that can damage the amplifier and possibly my dummy load.

Keep in mind that the crest factor of sine wave is also lower than what you get in music or even effects. What you perceive as bass lasting a while does not translate into continuous max power.

@amirm Thanks a lot for that information.
I am planning on making myself a 600 W dummy load to do such measurements!

I will show you the soundtrack from Interstellar as a frame of reference a bit later.

 

[amirm]

I am planning on making myself a 600 W dummy load to do such measurements!

Note that such loads (e.g. heater elements and such) tend to not be accurate ones which is another reason I don't use them (they will have distortion products of their own). If you are just measuring power, it would be fine. But not for high precision power vs distortion. That much heat also has to go some place.

 

[amirm]

I will show you the soundtrack from Interstellar as a frame of reference a bit later.

You will need to pull them into an audio workstation software to analyze.

 

[staticV3]

When amps are tested here at ASR as well as other places, output power is one of the things measured. I can't however find any information about the duration of such a measurement.

CEA-2006 Peak Power contains Duration specs:

They are also conducted at 1 kHz only, if I'm not mistaking?

Mistaken:

 

[StigErik]

Note that such loads (e.g. heater elements and such) tend to not be accurate ones which is another reason I don't use them (they will have distortion products of their own). If you are just measuring power, it would be fine. But not for high precision power vs distortion. That much heat also has to go some place.

Sure - it will be purely for power measurements and nothing else.

 

[StigErik]

Disclaimer: this method of mine is not very scientific, nor can any greater precision be expected.

So - what I have done so far is this:
- Mockup 4 ohm dummy load that can take about 200W. I have plans for a 600W+ dummy load.
- DAW software (Samplitude Pro X2)
- Focusrite Scarlett 2i2 USB interface

I created a recording project in my DAW where I have one track for playback and one track for recording the amp's output at each of the following frequencies: 10,15,20,30,40,60,80 Hz. A total of 14 tracks. Each playback track has a sinewave signal with 30 seconds duration followed by a 60 second gap to let the amp cool down. The signal level starts at -10 dBFS, and increases by 1 dB for each time, ending at 0 dBFS.

0 dBFS must be calibrated so it corresponds to the max output power you want to measure. I simply measure this with an accurate AC volt meter without the dummy load at the amp output. If I want 0 dBFS to be 100W @ 4 ohm, I simply adjust the master output volume in my DAW to reach 20V at the output. 20V^2 / 4 ohm = 100W

This is what the DAW project looks like. I'm currently testing an ICEpower 125ASX amp, which actually can deliver full power for 30 seconds in 4 ohm at both 10 and 80 Hz.

 

[StigErik]

Tests lasts seconds,however some datasheets do state max duration.

They do in some cases, but not in a very useful way I think.

This datasheet states different durations, but they are extremely short. 20 ms is just one period of a 50 Hz sinewave, which you probably won't hear as anything else than a short "pop".

I'm going to measure this amp to show what it actually can do.

 

[Sokel]

They do in some cases, but not in a very useful way I think.

This datasheet states different durations, but they are extremely short. 20 ms is just one period of a 50 Hz sinewave, which you probably won't hear as anything else than a short "pop".

I'm going to measure this amp to show what it actually can do.

View attachment 379741

Here's a similar of the ones I have to drive my woofers (1 mono for each one) :



Amir tested the module (crippled by an audiophile buffer) and did pretty decent at 20Hz



It states 90 sec duration at max at 1Khz at the stated conditions (I use the same heatshink at mine) which is ok I guess.
But I would like to see that duration at 20Hz from ice,even better from a third party.

 

[StigErik]

Here's a small portion of the LFE channel from the film Interstellar, the sequence where he flies into the black hole. It lasts for roughly 50 seconds, and the RMS level is about -3 dBFS. The energy is centered around 30 Hz. This is simply INSANE !

 

[StigErik]

It's temping to actually use that track for my power tests in addition to the pure sinewaves of course.

 

[Bjorn]

The Vera Audio amps are rated with continous output. I have to check, but I think the P400/1000 was tested continually for several hours.

One of the reasons I haven't considered some of the ICEpower modules with built in PSU is exactly their weaknesss in power over time.

As a side note. Not uncommon for amps to have a rise in distortion in the lows with a lot of power also by the way. As seen here with Amir's measurement of one.

Doesn't happen here, though high frequencies doesn't look great with 500W:

 

[Sokel]

One of the reasons I haven't considered some of the ICEpower modules with built in PSU is exactly their weaknesss in power over time.

That's one of the reasons I did get them for lows,anything else I checked at this time (sirca 2019) was either vague or vastly lower (time) .

Not that it matters a lot,but I did torture one of them with a cheap horrible 15" sub at 20Hz while making rounds around my house (no attached neighbors) to see if it was audible and the test was lasted a while.
It did get hot though,it's heatshink (the one specified by ice) must have been way over 50° but strangely no clipping kicked in with the sine,nor it shut down,despite the power I used (poor sub was about to explode)

The one spec'ed in the post above mine don't think is Ice,or is an old one?
Edit:It can be,the test is at combined 50Hz/1Khz,so makes sense.

 

[StigErik]

Here's a similar of the ones I have to drive my woofers (1 mono for each one) :

View attachment 379743

Amir tested the module (crippled by an audiophile buffer) and did pretty decent at 20Hz

View attachment 379744

It states 90 sec duration at max at 1Khz at the stated conditions (I use the same heatshink at mine) which is ok I guess.
But I would like to see that duration at 20Hz from ice,even better from a third party.

I have several amps with the same module. To my surprise, it did pretty decent in my 10Hz sinewave torture-test, right before my dummy-load burst into flames..... Not 1250W per channel though. It started to go slightly into protection around 1000W/4ohm, and it shows up on the captured waveform like this:

 

[Sokel]

I have several amps with the same module. To my surprise, it did pretty decent in my 10Hz sinewave torture-test, right before my dummy-load burst into flames..... Not 1250W per channel though. It started to go slightly into protection around 1000W/4ohm, and it shows up on the captured waveform like this:

View attachment 379768

They were strongly suggested by my installers when I asked for strong protections and good reliability record (their main task is big places,my room was kind of a -paid- favor) .
They state 1200W as mono,I don't actually need that kind of power nor 20Hz (f3 of speakers is 31Hz) but they were ideal cause I hooked up a 300a on each of them to drive the mid/tweeter.
They are ok for now,5 years after.