Guess you haven't messed around with BTL configurations very much.
No, not really, haven't needed to. But since I spent about half of last year working on the design of an active speaker project at work, I had to immerse myself in the topic. And I think you dismissed my comment without actually having carefully read what I wrote.
It is true that both the - and the + are active and not ground referenced exactly and the speaker is definitely not connected to ground but 'across' the class D switched mode outputs and inductors. If there is no differential between +L and +R and no differential between -L and -R than you can in fact tie the outputs together and get the same voltage but twice the current for a 3Db power gain. This would be paralleling the outputs and summing two equal amplitude in phase signals. In this case you could drive a 4 ohm speaker safely when each output was designed to drive 8 ohms independently.
[snip section below that goes into a different topic before you come back to this one]
Again, if the original poster sees no differential between +L and +R and no differential between -L and -R he can parallel the outputs and use a lower impedance speaker and take advantage of a 3db power gain. On the switching side of the inductors it makes no difference if the switching frequency is in phase or not, it only matters that the audio frequency after the inductors is equal and in phase and if that is true there is no reason the outputs can't be safely paralleled. The Carver PM2.0-T is bridged parallel for over 1000 watts at 4 ohms although it consists of two independent 'stereo' 500 watt channels at 8 ohms each. It uses a common ground which also happens to be tied to chassis as is the + and - power supply returns. The QSC units are different as both the - and + outputs move in opposite directions as measured to ground and the speaker is across these outputs and in bridge mono one of the channels gets an inverted input signal and the load gets connected +L to +R. The QSC amplifier output stages resemble the original posters amplifier but in his case he will be driving an in-phase audio signal and paralleling the outputs.
Right, that's classic paralleling. The thing is, that's all fine and true in theory, but modern class-D amp ICs use all sorts of clever sorcery to eke more power out of less, in particular, weird proprietary modulation schemes. So while you definitely can parallel the smoothed output of
two completely independent amplifiers, you definitely cannot make that assumption about two amplifier blocks
within the same IC. Some class-D amplifier ICs today have explicit PBTL support which you must enable in order to use them in that configuration (for example, the
TI TPA3156D2).
There is another way that big power amps can be bridged. Take the case of a stereo amplifier designed to be 500 watts at 4 ohms each side. There is a bridging mode where you input a mono signal, flip a rear panel switch and one channel will get an equal but inverted input signal. Then you connect a speaker of double the impedance (no longer 4 ohms on each side but a single 8 ohm) across the outputs from +L to +R and get 1000 watts at 8 ohms.
So… bridge tying the load. Which as I explained, in the TDA7492P (like the TPA3156D2) the outputs are already configured as a BTL. As I understand it, you can't
bridge the bridged outputs
again, because you're already at the limits of the power supply voltage x2 (p-p).
Apparently the majority of posters here don't use smartphones
No idea what the stats are, but I'm definitely far from the only person to access the forum by phone. I, of course, use the forum on all my devices (computer, tablet, phone), just depends on the situation. How the forum gets rendered on a given device will vary; I think I remember others saying it's also really annoying on the mobile forum skin and on tapatalk.
Apparently the majority of posters here […] also use hard coded line breaks.
Absolutely not. There's a tiny handful (like… 2 or 3?) of intransigent die-hards who refuse to stop putting in line breaks even after the problem has been explained to them, but the overwhelming majority of forum members do not use hard line breaks every line. They use line breaks the way they're intended to be used on the web and in word processors: for semantic grouping, i.e. separating paragraphs.
Thank you for being receptive to the feedback, it's much appreciated!!
you are the first to mention it.
I may be the first to mention it to
you. But I know the topic has come up a few times elsewhere on the forum (not regarding you specifically, but others who've done the same).