JakofClubs wrote: ↑Wed Dec 29, 2021 3:27 am
steelhorse545 wrote: ↑Tue Dec 28, 2021 11:37 pm
For a 12v supply, that’s 8.5v rms less fet drop,
I thought the RMS value (0.707 multiplication factor) was only applied to AC waveforms to get the DC power equivalent, and the 12v PS output power is DC? Or am I getting confused?
Not an expert, just trying to understand better
You’re partly right with the first bit. I’m a little sketchy on some details how class D amps work, but I’ve worked backwards from the quoted power vs supply rail, and it kinda ties up with one assumption.
Starting with a normal typical push- pull amplifier (eg instrument or audio amp.) cos it’s easier to think of (hopefully!), you’ve actually got a positive and negative supply (say 12v for this example), with the load (speaker) connected to the amp on one side, 0v on the other
If you think of a sine wave, assume the positive excursion goes from 0v to a peak somewhere near the positive supply rail (assume 12v), and the negative excursion goes from 0v to near the negative rail, then yeah, it’s 0.707 x 12v to give rms. Take a look at the classB and class AB diagrams here, the upper transistor is connected to the +ve, the lower to the -ve. (Each transistor drives a varying current which follows it’s part of the sinusoid)
https://en.wikipedia.org/wiki/Power_amp ... s#Class_AB
Class Damplifiers are a little different, since the output transistors are either on or off and the on:off time, when filtered (ie short term averaged ) actually gives a sinusoid (for this example) rather than the original on-off pulses see figure 7 here
https://www.soundonsound.com/techniques ... lification
Now, you might think that for a single 12v supply that’s equivalent to a sinusoid going from 0v to 12V, with the ‘zero’ point at 6v- and that’s true of one type of class D amp.
In the case of the tpa3116, I think it’s a ‘full bridge’ - which has the effect of having a +ve and -ve supply of (in this example) 12v - so as you wrote, yeah, the sinusoid peaks are effectively +12v and -12v, so rms is 0.7x12. Making that assumption (given that I’m a bit sketchy on exactly how they work !), 50w ties up with using a 21v rail.
(Even when hard-on -oooer ! - the amp transistors drop will drop something less than 1V, so it’s actually a little less than 12v that can be seen across the load.)
The 0.707 multiplier only works for a sinusoid - for a square wave or triangle wave or anything else it will be different.
(Edit) That becomes relevant when you try to measure the change in supply current to the amplifier when it starts driving, or the voltage across the load (resistor/transformer for the estim case), cos voltmeters/multimeters assume a sinusoids voltage or current to measure when measuring AC.- and possibly limited to 50-60Hz (ie your country’s AC mains frequency).
IDK what the ‘2W’ really relates to - you can’t actually measure power as such without replacing the load (resistor/transformer) with some test equipment, so all you can do is measure load current and voltage, or one of the two and make some assumptions on what the load resistance is.
Err..hope that helps - possibly too much information ! - but any questions, feel free to ask and I’ll attempt to answer..haven’t done any analogue electronics for a looong time :)
