A simple DAC+AMP idea

[MWP]

Hi all,

Is there any reason why i couldnt do this?
Itll be simple to build and should give rather HQ results (well im hopeing).

The only thing im not sure about is how the DAC will handle the value of the feedback resistors being ~6K ohms.

And, yes, no volume control, itll be done digitally before the DAC.

[Pedja]

I am thinking about something like this a few months already. You’ll need a unity gain stable opamp (or maybe resistor to ground to set the noise gain). A few things more should be solved but they are IMHO solvable.

In fact I was intending to try to implement something like this but without a digital volume control. I (still) hope this could be done using TDA1545A (non-o/sed, of course).

Pedja

[MWP]

You’ll need a unity gain stable opamp (or maybe resistor to ground to set the noise gain).

For what/where?
Another stage before the power op amps (LM3886)?

A few things more should be solved but they are IMHO solvable.

Why?
What else is wrong with the idea?

[Pedja]

Originally posted by MWP
Another stage before the power op amps (LM3886)?

No, this applies to the (one) used chip. The opamp that has minimum stable gain of 10 is generally not usable as an opamp I/V converter. It would be good either to use some unity gain stable chip (like BBs) or to use a resistor from the inverting input to ground to ensure the noise gain of, at least, 10.

What else is wrong with the idea?

Hmm, I wouldn't say that there is something really wrong... Say, some low passing would be good (soundwize) before the signal enters the opamp. Power opamps are note state of the art of the slew rate.

Pedja

[ojg]

This idea is fine in theory, but in practice there are several things that has to be considered to make it work as you intend.

1. Study the 1738/1792 datasheet carefully. Its current output has a large DC-component, and you will need to adjust for this in some way. (Hint: Study the AD1955 datasheet)
2. The amplifier doing the IV conversion needs to be able to handle signals in the MHz range. Like Pedja said, the LM3886 may not be ideal for this.
3. You should bypass the feedback resistor with a capacitor to limit the high-frequencies going to the speakers. Maybe also add a small inductor at the output.
4. A 6k feedback resistor is fine.

[MWP]

Ok, thanks for the tips guys.

I forgot about the DC component, the PCM1792 data sheet doesnt make any mention of it, but the PCM1730 shows how to remove it (same method as the AD1955 datasheet).

Not sure if ill use the PCM1792 or the PCM1730 as yet.
I have parts of each, but im not sure i want to build up a uC to setup the PCM1792's registers.
PCM1794's are far to difficult to get

[ojg]

PCM1792 has a DC current component, but it is referenced to ground.

PCM1730 has no DC component current, but is referenced to VCOM2 = 2.48V DC

Since you are running in bridged mode this DC compoenent will not reach the speaker, but it will limit the output voltage range of the amplifiers, since it will be operating "off center"