Just wanted to interject with a video link that has an amp technician walking through to explain this technology.
Also wanted to raise an interesting aspect that relates to both subwoofers and amplifiers, that IMO makes the design of these "constant power" amplifiers actually even more remarkable:
I admit amplifier design is not my specialty...
Non-regulated power supplies will allow the rail voltages to increase or decrease as the battery voltage is increased or decreased...
Regulated power supplies will hold the rail voltages constant when the battery voltage changes...
I was going to write a big thing here but Manville Smith of JL Audio wrote a sweet post over at DIYMA about ten years ago.
How does RIPS (JL Audio) work? post #21
Please read Manville's post, it is really well written and it goes into a lot of really good detail.
In short, the output stage is usually made up of transistors that can supply the speaker with voltage up to the rail voltage limits that the power supply has. The speaker will present an impedance (or a load) and then current will flow.
To add complexity to the challenge - speaker's impedance isn't a constant, it's naturally different at different frequencies, and at a speaker's natural resonance can rise surprisingly high - let's say a 4 ohm subwoofer has an Re of 3.5 ohms and an Fs of 24hz, like this completely random sample I found just because it had a plot...
You can see that impedence rise at Fs - up to 100ohms, maybe more:
View attachment 10572
Want another wrench to throw in the works? That above plot - it completely changes as soon as you put a subwoofer in a box. In a sealed box, it shifts based on the size of your box. In a ported box, the vent has an impact on cone motion which in turn impacts impedance. You could change the height or width of that resonance, or you could add a second high-impedance spike on the plot, again depending on box size and tuning.
The point relative to this "constant power" amp design is - a sub's impedance is all over the place, dynamically, and (from the amp's perspective) unpredictably. That amp is designed to drive one sub in a box as well as another totally different sub in a totally different box, and not impart any false coloration on the frequency response of that sub playing.
Continuing with the car analogy, a given car can drive 65mph on the freeway if the road is flat. The same car will go much slower up a steep hill, even with the gas pedal on the floor. The same car can go much faster than 65mph if you're driving down a steep hill, so fast that it can easily loose control and crash into a fireball and cause the road to be closed and inconvenience literally everyone else in a supremely selfish act of dumbassery.
Relating the car to the amplifier: an amplifier can supply 100watts of power if the impedance of the speaker is a good match (like driving the speed limit on a flat highway). The same amplifier, using the same rail voltage, can supply less power if the speaker impedance is too high (like driving up a steep hill). The same amplifier can supply much more than rated power if the speaker impedance is too low (like driving down a steep hill) and if the impedance is too low then the amplifier can overheat and blow up (like the car going too fast and crashing).
I love the analogies. Definitely driving your amp at too low of a load can cause it to die... I'd call that a car wreck.
It might be fun to take a deeper dive into this "impedance rise" topic, as a lot of people refer to it these days...
I see a lot of SPL competitors these days are actually becoming aware of their impedance plots, and are saying to themselves "Wait up... if my box is tuned to 60hz, and I measured my 1 ohm sub's impedance rises to 16 ohms at 60hz in this box - if my 2000w amp is only making that power at 1 ohm... does that mean my amp is only making... hang on... 2 ohm... 4 ohm... 8 ohm..16 ohm... 125 watts when I'm burping?"
So lately I've seen people wiring to very, very low impedances - 1/4 ohm, I'm sure some people are trying less - so that their amp makes the power their sub can handle at (and ONLY at!!) their burp frequency.
Would be fun to take a deeper dive into the negative consequences - That they've literally created a one-note wonder that would fry it's amp if you fed it anything other than that specific test tone (do people still use "burp buttons"?), and to compound that problem - that they've frequently got to solder across their protection resistors to totally eliminate amp protection (for those amps of this class that have protection) for this to even work at all. But maybe a theoretical dive into the negatives from the speaker perspective?
Constant Power
To get constant power out of an amplifier at 1Ω and 2Ω and 4Ω, you need two main things: 1) enough voltage amplifier power supply so that the power can be realized on the highest impedance load and 2) a current sensing/limiting technique so that the amplifier doesn't overheat or fail when the lowest impedance load is used.
An example amplifier might claim:
100w @ 4Ω
100w @ 2Ω
100w @ 1Ω
To get 100w at 4Ω the rail voltage would have to be
Sam from BareVids is an awesome guy for anyone who wants to understand amplifiers better. I subscribe to his vids, watch them while I'm working - he's a huge distraction, honestly.
In this particular link, he looks inside a Taramps amp that has a "smart" circuit like the JL RIPS amps.
It's important to know that - for the reasons Justin already described - there's limits to this "smart" technology just like there are limits to normal amplifiers that don't have this kind of design.
So these amp ratings would look like BS to me:
100w @ 4Ω
100w @ 2Ω
100w @ 1Ω
...because that's a big range to actually maintain constant power across.
I'd be betting it would really dyno out more like this:
40w @ 8Ω
80w @ 4Ω
100w @ 2Ω
100w @ 1Ω
Check out Sam's dive into this smart technology - he brings up the waveforms on his scope, so you can actually see the technology work, can see the voltages change, and power change.
It sounds like a deep dive, but Sam does a great job explaining it at a layman level. Definitely worth a watch.
https://www.youtube.com/watch?v=DxMGPhNASl4&t=440s
Note that link is actually the 7:25 mark on a Williston Labs review - another guy worth subscribing to IMO - but I subscribe to Sam as well. His channel is "barevids" - don't worry, it's not what that looks like. Nothing NSFW.
Well - to me. 