Hello everyone! First of all sorry for my english, I'm not native speaker) I have a question about bridge-tied load (BTL) mode in IC audio power amplifiers (for example TDA2822M, NJM2073, etc.). I spent some time trying to understand how it works but I'm confused about its schematics and operation as a result. As it can be seen from datasheets such mode is obtained by simply connecting both inverting inputs with a capacitor of 10uF and connecting the load between the 2 outputs of the IC.
My questions are:
1. How can this even work if negative feedback resistor dividers of both amps dont have a connection to the common (ground) wire? They are connected with each other at the point each one is connected to ground in single-end (stereo) mode. So how do such dividers divide output voltage to get feedback signal? They don't have zero-potential reference point in my understanding.
2. What's the use of 0.01uF cap from inverting input of 2nd amp to ground? Is it an accident that its value is equal to 1/1000 of the coupling capacitor (10uF)? Why this cap doesn't form an almost 100% negative feedback to inverting input of 1st amp thus making it almost unity gain? 0.01uF has impedance of tens kiloohms on signal frequencies making inverting input being connected to ground via this impedance.
3. Does BTL-mode provide gain equal to single-end stereo mode? I thought it's gain is doubled but having built a test circuit in multisim I see that it isn't. I got 700mV RMS output with 70mV RMS input both in BTL and single-end mode with equal feedback resistors. Then what's the use of such mode? Does it simply give an opportunity to get a larger output swing but due to increase of input voltage, not but by doubling gain? Or have I built my circuit wrong and the gain must be doubled?

 

Your attachments do not show their negative feedback.
Your "BTL1" circuits are not power amplifiers.

One power amplifier has its own negative feedback and feeds ONE speaker wire with its output in phase with its input.
The other power amplifier also has its own negative feedback and feeds the other speaker wire with its output out of phase (inverted) with its input.

IF the amplifier input is 1V peak and their gains are 10 times then one amplifier output reaches +10V when the other amplifier output reaches -10V at the same time for a total output of 20V.

The input to the inverting amplifier can be taken from the output of the non-inverting amplifier.

 

In my BTL1 circuit I threw out current sources and output stages for simplicity. Let’s call them voltage amplifiers then. But the datasheet circiut for njm2073 is a fully working circuit, isn’t it? Yes, feedback resistors are inside, providing maximum gain without external ones. But it works. Connecting inverting inputs with a cap makes everything work, but I don’t understand how it can be truth) I know about circuits when 2nd inverting amp takes signal from the output of 1st inverting amp through a resistor equal to feedback resistor, BTL from datasheet is obviously not this case.

 

Digikey has a datasheet of the old NJM2073 with more detail but its English is poor.

 

Digikey has a datasheet of the old NJM2073 with more detail but its English is poor.

Could you please get and attach it here if you can access digikey site? It returns an error for me for some reason. Thank you very much in advance. I can only find datasheets by NJR on other sites, all are the same.

 

Digikey says the NJM2073 is obsolete and have none in stick.
Here is their 20 years old datasheet. New Japan Radio recently changed their name to Nisshinbo Micro Devices Inc.