Bruno is using a sinewave and I bet he is clipping the amp. 😄
Under those circumstances, it becomes possible. 🤷🏻
Hehe maybe, but not more then 1% or so because I couldn't hear clipping.
I know how clipping sounds, Im like listening to 20w on a 100w amp to have clean signal.
Anyway, this was more of a " torture test " on the amp.
I don't really care what power the amp measures as long as my friend seems happy, and there's plenty of power !.
If you can’t tell when clipping starts (quite possible without a scope), then what you need to measure is peak power. Make a peak detector out of a 1N400x diode and a 0.1 uF capacitor. Connect that to the output and measure DC. That will be accurate (enough) and can’t read over what it is supposed to. When you go beyond clipping the reading will start to decrease because the power supply continues to drop voltage the more load you put on it. Just pick off the max you get, but remember that pesky factor of two between peak and average (“RMS”) power. Yeah, you could add 0.6 volts for the diode drop. Want real accuracy instead of a 0.6V fudge factor, use an op-amp based peak detector and attenuate ahead of it to keep the signal inside it’s valid range (+/-12 volts or so). You wouldn’t have to resort to cheating like this if you had a scope.
If you do the peak power test with full range music it will read a bit higher, as the power supply hasn’t dropped as much as with a continuous sine wave. The “RMS” you calculate from that is close to the old IHF power rating. With a 350 VA toroid that difference between real power capability and IHF will be about 0.7dB or 85%. Most store bought equipment will use a LOT smaller transformer and the drop between 1 and 2 dB! And it would take doubling it or more to get the drop down in the 5% range. At least you are using the right size capacitor. 10,000 uF is correct. 20,000 for stereo. That will for practical purposes maximize your 20 Hz power capability. More would only result in a hotter transformer and no more power.
If what you friend is used to are little plastic bluetooth speakers and those underpowered KLH subs that burn up all the time this will sound like a million bucks. Just shows how far we’ve fallen.
If you do the peak power test with full range music it will read a bit higher, as the power supply hasn’t dropped as much as with a continuous sine wave. The “RMS” you calculate from that is close to the old IHF power rating. With a 350 VA toroid that difference between real power capability and IHF will be about 0.7dB or 85%. Most store bought equipment will use a LOT smaller transformer and the drop between 1 and 2 dB! And it would take doubling it or more to get the drop down in the 5% range. At least you are using the right size capacitor. 10,000 uF is correct. 20,000 for stereo. That will for practical purposes maximize your 20 Hz power capability. More would only result in a hotter transformer and no more power.
If what you friend is used to are little plastic bluetooth speakers and those underpowered KLH subs that burn up all the time this will sound like a million bucks. Just shows how far we’ve fallen.
@wg_ski Of course I could tell when the clipping starts, not so good as a scope would , but distortion sounds disgusting especially on mids highs. on bass not so obvious maybe. Nice sugestion for that peak detect, reminds me when I was a kid I put a bridge rectifier and a cap on a pc fan on my map to spin it on bass LOL.
True but with a small caveat.
You will be reading audio signal peak (good) plus superimposed ripple peak.
Normal "scoped" reading reaches just top and bottom peaks and stops there.
If you drive amp hard into flat topping, you will see a 100 Hz signal modulating peaks.
As much as 5 V peak, depending on supply filtering.
Meaning such reading is better than nothing (I used it in my poor student scopeless days) but in general "optimistic"
Oh well, the good old days 🙂
Plan B: improving ear sensitivity:
I connect a Piezo speaker in parallel with load and hold it near my ear.
Start of clipping becomes VERY noticeable, as an unbearable "mosquito buzz".
Try it , you'll be surprised.
If having a scope, try it rising and lowering volume through start of clipping point, start of buzzing can clearly be related to clipping seen on screen.
Last edited:
The supply ripple problem is mitigated by using the right size filter capacitor. 10,000 uF and 4 ohms is a low frequency pole of 4 Hz. “5X away from an STC time constant” and the attenuation is half A dB. You could be losing at most 10% to power supply ripple. If there is any inductance or resistance in the power supply secondary you lose even less as the charging pulses get spread out (improved power factor, and with it a drop in the DC volt). Higher than 20 Hz and you lose less. That’s a “close enough for government work” approximation. You want more detail you simply have to spring for a $1000 oscilloscope.
Far too many sub amps on the market dont have enough capacitance. They make these awful “BVVVWOKKK” sounds when driven to clipping with barely-audible LFE signals. Nasty enough to make you think the woofer is bottoming out.
When comparing specs on modern pro amps these days they use an “EIA” rating - driven to 1% THD+N at 1kHz rather than the onset of clipping. The amp is visibly clipping, and there is power supply ripple superimposed you see it. Many of those amps dont have enough capacitance either. That causes them to wear quickly too, as they are usually run far beyond ripple current ratings as well (even when loaded to 1/8 power) due to the undersizing.
Far too many sub amps on the market dont have enough capacitance. They make these awful “BVVVWOKKK” sounds when driven to clipping with barely-audible LFE signals. Nasty enough to make you think the woofer is bottoming out.
When comparing specs on modern pro amps these days they use an “EIA” rating - driven to 1% THD+N at 1kHz rather than the onset of clipping. The amp is visibly clipping, and there is power supply ripple superimposed you see it. Many of those amps dont have enough capacitance either. That causes them to wear quickly too, as they are usually run far beyond ripple current ratings as well (even when loaded to 1/8 power) due to the undersizing.