Add TPA3251

[evanshultz]

http://www.ti.com/lit/ds/symlink/tpa3251.pdf

Pins 9 and 10 were a little hard to figure out their type, but based on page 23 I chose Input.

I've generated a scripted footprint, but haven't submitted it yet pending the resolution of KiCad/kicad-footprints#955.

---

All contributions to the kicad library must follow the KiCad library convention

Thanks for creating a pull request to contribute to the KiCad libraries! To speed up integration of your PR, please check the following items:

• Provide a URL to a datasheet for the symbol(s) you are contributing
• An example screenshot image is very helpful
• Ensure that the associated footprints match the official footprint library
  • A new fitting footprint must be submitted if the library does not yet contain one.
• If there are matching footprint PRs, provide link(s) as appropriate
• Check the output of the Travis automated check scripts - fix any errors as required
• Give a reason behind any intentional library convention rule violation.

[evanshultz]

The CI failures are OK, I believe. AVDD and DVDD are internal regulator outputs that cannot source current and just need a cap to ground, making Passive a fine choice.

[cpresser]

A few questions & remarks:

• Why did you split it this exact way? I would rather expect 3 units. A: all Controll-Stuff, B+C one channel each. Or just make one single unit. But having control pins split over two units seems weird.
• I agree for the pin-types of AVDD and DVDD. Also the position bottom-left is good.
• OSC_IOP and FREQ_ADJ are outputs
• OC_ADJand OSC_ION are bidir,
• FAULTand CLIP_OTW should be on the right side.
• I would swap OUT_Band BST_B so they have the same order as *_A. Thats just my personal opinion that this looks better when wiring it up with the bootstrap capacitors.

[evanshultz]

1. The logic is the unit A has IC-to-IC interface and 'passive' pins, while unit B has pins that interface with an external processor (or circuit). So would you prefer two units with just audio input, output, and power, with a third unit just for interface and 'passive' pins? Would you then prefer all interface pins on the same side? I'm trying to seek answer and explain my reasoning at this time. You may look at the IR4302 class D IC which has pins split unevenly.
2. Pages 15 and 16 intimates that OSC_IOP are bidirectional. Pin type changed. FREQ_ADJ is actually connected just to a resistor for setting the freq as we've used Passive for that. Probably the pin table shows as 'O' because it's a current source inside the IC.
3. OSC_ION pin type changed. OC_ADJ is same as FREQ_ADJ.
4. In this case I thought keeping all the pins that would go to an external process together was best. So they're all on the left side. (But yes, split across two units.)
5. I thought it was more clean the way it is so that the class D filter can be between the outputs without a bootstrap cap in the middle. But I agree this is rather subjective. No change made for now.

[cpresser]

I re-checked the pin-types of the 'controversial' pins. The block-diagram arrows and the pin-table don't always match. And then there is also the example application (e.g. 10.2.1).
Anyway, I don't want to spend to much time discussing pin types. The current types are good enough.

Regarding the layout:
As suggested I looked at IR4302. That did help me understand why you did split the symbol this way. I agree that this spit does follow sane rules. Just because I would have done it different does not mean it needs to be changed 😄

So in conclusion, I am happy to merge as is. Or is there anything else?

[evanshultz]

Thanks for looking again!

I can change things. It doesn't have to be this way. Maybe IR4302 isn't a good example and no symbols should look like this. If you're OK so am I, but I'm not staunchly opposed to a change.

The footprint needs to be merged before this. And that depends on KiCad/kicad-footprints#955. Sigh.

[evanshultz]

Submitted the footprint at KiCad/kicad-footprints#2468.

[myfreescalewebpage]

Closing/opening to refresh the Travis test.

[myfreescalewebpage]

@cpresser the footprint has been merged

[evanshultz]

I fixed the footprint name referenced by the symbol, so the only remaining Travis errors are fine and described above.