How to properly connect L, R, GBUF to external amplifier

[shubhamgandhi]

Hello VLSI support team

I'm asking because I searched through the forum and wasn't able to find a clear answer.
I'm attempting to create the exact same application as the VS1000D Button Cell Player, except play sound to a 0.5W, 8-ohm speaker instead of headphones. I'm using a 3.7V LiPo battery, with a TPA2005 amplifier. Here is the schematic of my setup so far, based on the buttom cell player: http://i.imgur.com/H0VHQys.png

Can you please critique my schematic and help me answer the following questions:
- I want to combine the audio output of VS1000D to mono, is summing them with 1k resistors the right way to do it?
- Do I need a DC blocking capacitor at IN+ input of amplifier?
- I'm using the following 1.4W class D amplifier. Is there a recommended amp I should be using to drive a 0.5W speaker with 3.7V battery as supply voltage?
http://www.ti.com/lit/ds/symlink/tpa2005d1.pdf

Thank you for your support.

[Henrik]

Hello,

to see how to implement a line output, have a look at Panu's example schematic here:
viewtopic.php?f=9&t=69
Particularly take note of the the Sigma-Delta Reconstruction Filter: you will need it or there may be excessive noise.

For mono output, one option is to turn on the differential output on on the VS1000, by saying somewhere in your application:
audioPtr.rightVol = -audioPtr.leftVol; /* Activatees differential output */

Then you can connect LEFT and RIGHT to the differential inputs of your mono amplifier (GBUF is then not connected to the amp at all, but remember to still connect it to GND through R15 and C32).

If you cannot change the source code, then connecting LEFT and RIGHT through a resistor might be the best idea.

Kind regards,
- Henrik

[shubhamgandhi]

I cannot change the source code, so I'll be summing them at the + terminal of the amplifier through 1k resistors as in my schematic.

Should GBUF be connected to the - terminal of the amp?
Does GBUF still need to the passive components for the sigma-delta reconstruction?

Thanks

[Henrik]

I cannot change the source code, so I'll be summing them at the + terminal of the amplifier through 1k resistors as in my schematic.

Should GBUF be connected to the - terminal of the amp?

Yes. Make sure, though, that the amp is a differential amp, because GBUF is not a hard ground; it's a virtual ground (absolute level is about 1.2 or 1.6 V, depending on settings).

Does GBUF still need to the passive components for the sigma-delta reconstruction?

Actually, if you look closely, Sigma-Delta reconstruction components apply only for LEFT and RIGHT. Note, however, that GBUF needs the High-Frequency Grounding Path components.

Kind regards,
- Henrik

[shubhamgandhi]

Would it make a difference if I was to connect (-) terminal of the differential amplifier to ground instead of GBUF? What purpose does GBUF serve?

Also, I'm basing my schematic off the button cell player scheamtic:
http://www.vlsi.fi/fileadmin/evaluation ... 2b-sch.pdf
I noticed that it does not have the Sigma Delta reconstruction passives as you suggested? Is there a reason for this?

Thanks

[Panu]

Hi!

The GBUF is fairly well explained in many topics in this forum, take a look for instance at:
viewtopic.php?f=8&t=63#p258

Would it make a difference if I was to connect (-) terminal of the differential amplifier to ground instead of GBUF?

It would not work.

What purpose does GBUF serve?

It provides a fixed voltage of 1.25 volts. The analog signals are centered around this voltage. There are no negative voltages and no negative power supply required for VS10xx.

the button cell player scheamtic ... does not have the Sigma Delta reconstruction passives as you suggested? Is there a reason for this?

Headphones don't require the reconstrucion filter (their inductance functions as the reconstruction filter). Only amplifier inputs require the filter.

-Panu

[shubhamgandhi]

Thank you for the support. One last question:

Do I need the DC block for output?

[shubhamgandhi]

I put in the DC blocking RC as well. Would you mind taking one last look at the schematic and seeing if anything glaringly incorrect pops out?

http://i.imgur.com/e9k14fi.png

Thanks

[Panu]

Hi!

A few remarks...

Since your connection to the amplifier is differential, you should not use the DC blocking capacitors. And speaking of pop's you could use the /SHUTDOWN signal of your amplifier to reduce any power-on pop sound. I don't know if there is any digital signal available to control this pin, I would perhaps put an RC delay from AVDD to /SHUTDOWN so that the amp would remain silent for some fraction of a second after the AVDD is switched on. For more precise result, with TPA1006D1, the /SHUTDOWN control might even come from the GBUF signal (maybe delayed for a few milliseconds by an RC). Anyway, if you can, try it before you make the PCB.

Note that the POWBTN pin is not 5V tolerant - I don't know what you intend to connect to JP3.

There are a few unconnected IO pins in the design, since VS1000 is pure CMOS IC, these should be pulled low, high, or tied to other pins. And I wouldn't use 1MOhm resistors for anything except the USB biasing - in dry weather the voltage is unpredictable, too much static electricity in the air.

-Panu

[shubhamgandhi]

Hi Panu, thanks for the pointers.

I have removed the DC blocking caps. Regarding the resistors, I changed all the 1M to 100k, except the oscillator and USB biasing resistors.

I'm going to be controlling the shutdown pin, and the POWBTN pin using an external microcontroller.

How come in the button cell player, the POWBTN pin is connected to USB 5V through a pushbutton?
I'm planning on giving a long enough 5V pulse at JP3 that turns on the VS1000D from sleep.

I added a pull down to all the unconnected pins through a single resistor. Is this allowed?