TAS2563: Tone Generator & Complete Datasheet

Hi Steffen,

I can confirm the beep generator is not configurable on TAS2563 the same way it is for TAS2562. My observation is:

• Beep is only working by setting register 0x3f on book 0 page 0 to 0x40. Other settings are not working to trigger the beep tone.
• Beep tone is triggered by SDIN pin, this means if I2S input has dither noise it can trigger the beep constantly
• Beep tone is triggered for 100ms
• Beep tone can also be triggered by toggling Class-D software shutdown
  • Write 0x02 to register 0x02 on book 0 page 0
  • Delay 1ms
  • Write 0x00 to register 0x02 on book 0 page 0

Some changes are applied to the configuration scheme of TAS2563, this is the main reason why some features are not included or were changes on later days as compared to the documentation.

I'll check what additional content we can share in terms of register map, it may be limited to the first few pages on book 0, most of the other books/pages are used for coefficient memory of the processing features.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

Thanks for clarification.

About enabling the tone generator, I was aware of software enable (writing 0x40 to reg 0x3F), and this is the only mode I use.

Alas, assuming that the basic structure of TAS2563 tone generator is similar to TAS2562, there is a couple of additional parameters that need to be set in advance, i.e. frequency and amplitude. These settings also appearing in PPC3 Device Control, in ROM mode, make me confirm this assumption.

Analyzing PPC3 export headers after setting various tone generator parameters make me assume there are registers in page 5 of book 0, structured pretty much similar to TG1_FREQ und TG1_AMP of TAS2562. The TAS2562 registers are located on page 2 though. But writing to any of those adresses (neither those from PPC3 export nor those from TAS2562 datasheet) fails to set up the tone geneator on our target board.

(In general I am not happy with these reverse-engineering methods and part equivalence assumptions we are applying, but I need to implement this feature and the documentation is lacking. So any additional content about the extended register map you could share with us would be greatly appreciated. If an NDA is required, please let me know.) 

Could you identify the correct tone generator parameter register addresses of TAS2563 please?

Hi Steffen,

Attached is a table with the required calculations for each of the parameters, including an example script to show the relevant register addresses.
I'm getting feedback on the extended register map, will let you know about it as soon as possible.

w 98 00 00
w 98 7f 00
w 98 00 05
w 98 24 7B A3 75 1D
w 98 2c DE DF 04 7D
w 98 40 00 00 00 18
w 98 44 03 37 18 4E

w 98 00 00 
w 98 7f 00 
w 98 00 00 
w 98 02 02
d 01
w 98 02 00

ToneGen_2563.xlsx

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

Thanks for providing tone generator parameters.
I do confirm that these register values set the tone generatot to 2000 Hz (given that register 0x3F is set to 0x40 to enable output). The spreadsheet comes handy for calculating different frequencies and amplitudes.

The drawback is, this tone generator mode appears to be audio level triggered. Failure to provide an audio signal results in no tone. This is not what we are looking for. In our application, the tone is intended to provide keypress feedback to the operator, so we need the tone to be output regardless of audio level (and also regardless of PDM clock being present) please.

Also, currently, when enabling the tone generator, this appears to be adding heavy DC offset on the speaker coil.

If there is audio-level enabled tone mode, there likely is independent tone mode as well, isn't it?
Please also have a look at DC offset. Could there be something wrong with calculation of TG1_FREQ3? In the formulae from data sheet of TAS2562, there is a subtraction of 1 prior to setting this register, which I cannot find in the spreadsheet's formulae?

Hi Steffen,

What would be the best way to trigger the tone? By toggling GPIO pin? Would you be able to toggle any other pin?

I'll double check about the DC offset, I didn't notice that during test, will let you know some feedback as soon as possible.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Steffen,

I do not observe DC offset at the outputs either before or after enabling the beep generator. Can you confirm if the DC offset is observed on both OUTP and OUTM signals so that the resulting difference is nulled? I think that may be the case.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

Yes I could toggle a GPIO pin in order to switch on the tone generator. Please feel free to choose any pin, we can adapt our layout accordingly.
Please keep in mind that the tone generator needs to operate in absence of audio clocks, too. Comparing to TAS2562 one more time, if I have understood correctly, the latter has an option to use an internal oscillator for clocking the DSP to generate tone in absence of I2S clock. Hopefully TAS2563 can, too.

About DC offset, you may be correct indeed. I will recheck.

Hi Steffen,

I've been trying different configurations but so far I'm not able to trigger the tone generator similar to TAS2562, I'll review this with the rest of the team and let you know any updates as soon as possible.

Best regards,
-Ivan Salazar
Applications Engineer