CC2564C: CC2564C

Hi Tom,

What is the host platform for the CC2564, is it an MCU or is it a processor running Linux? Also, are you attempting to use an assisted mode such as wideband HFP or assisted A2DP?

While the wiki is no longer online, I have backups of the wiki pages that I can provide. For using the PCM interface, the VS command documentation might be helpful: https://tidrive.ext.ti.com/u/-9MRXEDTEUnp0xFd/98ec7908-5d1e-4d80-a27b-a2456f7feda2?l

Regards,
Michael

Hi Michael!         (sorry for the latency.. Your responding time shows me: you sit in DE )

Thx. That PDF is in use here already, of course.

Yep. host platform is Linux, self-made one, so far.   Main use case is HSP - but enabled now HFP as well  -- and  A2DP in addition, - but only for later "juwelry on nightdress", for audio through the air, coming from an AudioPlayer and FMradio on the mobile phone.

Regards,
Tom

Hi Tom,

Have you taken a look at the demo applications for those audio profiles provided along with the Bluetopia linux releases? They should demonstrate how to use the PCM interface and configure it for use with an external codec.

Regards,

Michael

Hi Michael,

What exactly do you mean, where to find? Can you provide me a link to that, please?

Regards from DE to ?,
Tom

Hi Tom,

I'm actually based in the Dallas HQ.

In the Bluetopia Linux addon https://www.ti.com/tool/TI-BT-4-2-STACK-LINUX-ADDON

There are a variety of demos in the sample directory that each show off one of the supported profiles and how you'd make a user application that interfaces with the stack to implement the desired functionality. One of the demos is the LinuxHDSM_AG demo, which should demonstrate an HSP profile implementation.

Regards,

Michael

Ok, so -7 hrs to my location close to Berlin, Germany. I saw your 1st response at 7 AM my time. That's why.. So you work a bit like me, - at night or longer (here since 6 yrs every day until 3..6 in the morning)

Thanks, Michael!  Will "lurgen" like Swiss say (have a look at).
Later I perhaps will also provide my bluetoothd log to show that the basics should be ok..

Just and "only" the codec in the CC seems to be the remaining problem. Unfortunately no chance to watch the four pads, no test points there on the PCB. 
That my wildcat version of hcitool is principally working - proved the use of testing cmds ("Continues Wave") in the certification tests by a lab in Shenzhen. Just finalized for V2 of the phone. I provided that to the CN by MMI codes.

Thank you very much!
Till then...

Hi Tom,

For sure, let me know if the code in the sample applications is of any help or if you need further assistance on getting the PCM interface on the CC2564 working.

Regards,

Michael

No, Michael, cannot work.

Compiled it - but cannot link. Reason: Everything using VFP, armv7ahf in this case. BUT: Our thing is an armv5, no hf.

Question is so slowly: Is the "hardware codec" in the CC2564 (Hah! just firmware) really standalone usable - or does its use depend on much more things in the applicatice environment, - made by special software on the host, like Bluetopia? Looks like..
Special CC initscript for Bluetopia in the kit seem to prove this as well.

Regards,
Tom

Speaking of the objects libraries - to be linked with at the end.

Everything in there had been compiled for an ARM w/ floating point coprocessor, here with VFP in form of NEON.

For a stone age applicator like me w/ ARMv5 .........
Typical "ARM fun"..  https://embeddedartistry.com/blog/2017/10/11/demystifying-arm-floating-point-compiler-options/

Enjoy the weekend!
Tom

Well .., I'm pretty much at the end ..

At least one positive effect by the sample application kit: checked if the format of my HCI output for CODEC_CONFIG etc. is ok regarding OGF, OCF, byte order.   Seems to be:

My hcitool script:

# CODEC_CONFIG
hcitool cmd 0x3f 0x106 0x22 0x00 0x08 0x01 0x40 0x1f 0x00 0x00 0x01 0x00 0x01 0x00 0x00 \
0x10 0x00 0x01 0x00 0x01 0x10 0x00 0x01 0x00 0x00 0x00 \
0x10 0x00 0x11 0x00 0x01 0x10 0x00 0x11 0x00 0x00 0x00

# SCO CODEC PCM  (juwelry on nightdress)
hcitool cmd 0x3f 0x210 0x04 0x00 0x00 0x00 0xff

=== CODEC_CONFIG AIF2
< HCI Command: ogf 0x3f, ocf 0x0106, plen 35
22 00 08 01 40 1F 00 00 01 00 01 00 00 10 00 01 00 01 10 00
01 00 00 00 10 00 11 00 01 10 00 11 00 00 00
> HCI Event: 0x0e plen 4
01 06 FD 00

=== SCO CODEC
< HCI Command: ogf 0x3f, ocf 0x0210, plen 5
04 00 00 00 FF
> HCI Event: 0x0e plen 6
01 10 FE 00 B4 04

#-----------------------------------------------------------------

Exerpt from btpmvs/cc256x/BTPMVS.c   PM_VS_PCM_Codec_Config()  in the kit
Output:

Configure CODEC slave = 1
OGF=0x3F OCF=0x0106 CommandLength=34
0x00 0x08 0x01 0x40 0x1F 0x00 0x00 0x01 0x00 0x01 0x00 0x00
0x10 0x00 0x01 0x00 0x01 0x10 0x00 0x01 0x00 0x00 0x00
0x10 0x00 0x11 0x00 0x01 0x10 0x00 0x11 0x00 0x00 0x00

Made a file with all logs from hciattach, hcitool, bluetoothd but - Huh?! Can't be attached here?
https://owncloud.antarus.com/index.php/s/TdB6PaAOPsUxiVF  (no worries, self-signed cert, signed by me )

Unfortunately.. No chance to watch the 4 PCM/I²S lines by scope. But I once checked on a PCB w/o the Panasonic module that both clock lines and ADCDAT on WM8958's AIF3 (BT module connected to) are working.

Best
Tom

Hi Thomas,

Yes, unfortunately the TI Bluetopia stack is only built for certain ARM architectures, and the ARMv5 is not one of them.

That being said, the PCM interface of the CC2564 is not dependent on using that stack - all the stack is doing for you is calling the same VS HCI commands that you'd use anyway.

I'll take a look at your logs tomorrow, didn't have time to do so today unfortunately.

Regards,

Michael

Hi Michael,

"Nur die Ruhe.." like Germans say ("Just take it easy ..")

Tom

Made a "Logs Ragbag" directory w/ the above and two more log files:  https://owncloud.antarus.com/index.php/s/FcTPn15LgCMtxGE