Database Info

[Q] choose array-elements with certain probability

hello,
I'm writing a Android App (java) - but I guess this question is pretty general, and isn't java-specific:
so I have an Array of Elements
and I let a random-number-generator pick one array-element randomly and hand it to me.
now I've also built in a "score"-field into each array Element
so what I want to do now, is that the random-number-generator takes the array's "score" in consideration, and gives me the array-elements with the higher/lower scores with a higher/lower probability
I dont want it to ALWAYS/NEVER give me the elements with the highest/lowest score - just with a higher/lower probability
I hope I could describe my problem in a proper way....
does anybody know how to achieve this?
(as I said, i use java, but i guess code in any language - or even pseudo-code would help me out)

*bump*
anybody?

Try this, it's in C# but it's pretty close to Java.
You cannot directly weight the random function so you have to use a different method.
By applying a weight to each item in the array, then using a random function to select using the weighting values, the results of the selection can be swayed.
The weighting values are relative to each other.
Using the values in the code, 'B' should turn up about three times more often than 'A'
The only object that may need some explanation is Random:
http://msdn.microsoft.com/en-us/library/system.random(v=VS.80).aspx
Code:
using System;
namespace RandomWeighting
{
class Program
{
static void Main(string[] args)
{
char[] Select = new char[10] {'A','B','C','D','E','F','G','H','I','J'};
int[] Weight = new int[10] {10,30,25,60,20,70,10,80,20,30};
int[] WeightSum = new int[10];
int i,j,k;
Random Rnd = new Random();
WeightSum[0]=Weight[0];
for (i = 1; i < 10; i++)
WeightSum[i] = WeightSum[i - 1] + Weight[i];
for (j = 0; j < 70; j++)
{
k = Rnd.Next(WeightSum[9]);
for (i = 0; k > WeightSum[i]; i++) ;
Console.Write(Select[i]);
}
Console.WriteLine();
}
}
}
Output:
Code:
HEFIBHHCCFBCAEFFDHACHBEJHHFDFIDFEDFFCHHDJBIDJEHHFHCJJJBHJGBDDGFDDFHHHB
Note the low density of A and G as opposed to H
It is just a sample at random, but 2 'A's and 7 'B's roughly matches the conjecture above, but over a million selections, the distribution is as follows:
Code:
A 30664
B 84187
C 70648
D 168481
E 56529
F 197311
G 28145
H 225764
I 56613
J 81658
If your code changes values in the Weight array then the WeightSum array must be recalculated.

wow, thats a lot!
The only thing I dont understand about the code is this line:
k = Rnd.Next(WeightSum[9]);
for (i = 0; k > WeightSum; i++) ;
what exactly happens here?
what kind of a for-loop is this?
why is there no body?
and strangely, while debugging this line, i noticed that the value of i jumps to some random number in this line - and I have no idea why and how
------
by the way: I've already tried it out, it works pretty good.
although I noticed that the first element always gets picked ALOT, no matter how low the weight.
I think thats a flaw in the algorithm

After the following code:
Code:
for (i = 1; i < 10; i++)
WeightSum[i] = WeightSum[i - 1] + Weight[i];
The WeightSum array contains the following values:
Code:
[0] 10
[1] 40
[2] 65
[3] 125
[4] 145
[5] 215
[6] 225
[7] 305
[8] 325
[9] 355
The Random.Next() function that takes a single integer as an argument is defined as:
-------- Courtesy of MS VS .NET Help ------
public virtual int Next(int maxValue)
Parameters maxValue Type: System.Int32
The exclusive upper bound of the random number to be generated. maxValue must be greater than or equal to zero.
Return Value Type: System.Int32
A 32-bit signed integer greater than or equal to zero, and less than maxValue; that is, the range of return values ordinarily includes zero but not maxValue. However, if maxValue equals zero, maxValue is returned.
-------------- End of Help ----------------
So, we are asking for a random value between 0 and 354,(element WeightSum[9] above). The following code then finds which is the lowest element of WeightSum which holds a value greater than the one we have been given. When this happens 'i' contains the index to use for the selection.
Code:
for (i = 0; k > WeightSum[i]; i++);
The standard construction of a for loop in C is
for(One or more initialisation statements; Do_While_True Condition; One or more iteration statements)
{
loop processing code;
}
If there is nothing to do in the loop body, you don't need it, and you can replace it with the ';' at the end of the for statement. In effect, it is identical to the following code:-
Code:
i=0;
while(k > WeightSum[i])
{
i++;
}
As regards the first element being picked more than the others, have a look at the distribution table in post #3. It is what you would expect for the values given. I assume the difference is either the Random function in Java or some different implementation between C# and Java.
You may have to change some of the code slightly, i.e. change the for() loop to the while() loop above and step though it in the Java debugger to get to the root of the problem.
You can't debug a for loop followed by an immediate ';' The entire for loop is executed to completion when you use debug and try and step through it. To debug it place a dummy statement in the for loop. A breakpoint may now be set on this line.
Code:
int i,z;
for (i = 0; k > WeightSum[i]; i++)
{
z=0;
}

ActionScript and Javascript versions ...
Thanks for the great write-up! In case anyone is interested, I've adapted this into a javascript and ActionScript class. If anyone is interested, I've attached the code. For a more in-depth post, check out blog.teamthinklabs.com.
Cheers!
Kipp Ashford
Art Director
Seven2 Interactive

TI FM Radio

Preface
Hi !
In my reading of various threads it's become apparent that there are MANY people who passionately
desire a "real FM radio" app. Not all of us live in data dense areas, or can afford the costs of
streaming audio. And not all of us have given up on RF broadcasts, either due to our tastes,
local available programming, lack of commercials or whatever. And many of us very much desire
the ability to transmit on the FM band for various reasons.
I want this thread to deal with TI FM radios on all devices that contain TI chips. My observation is that
there is a great deal of commonality in the TI chips supporting FM in the last 4-6 years.
I don't generally want to deal with the Broadcom FM chips here, but the audio routing issues will be similar.
I also WOULD like to create a list of devices containing the FM chips they use, so people can more quickly
determine their FM chip type.
To non-devs who want FM yesterday: Yes, I know. No need to post about it. I'll do my best to create some
kind of app ASAP, but first some fundamentals need to be figured out. I'd think and hope that others will also
be able to create FM radio apps from the info here and elsewhere.
This thread is in a developer forum, and as such it would be preferable to limit discussion to technical
aspects, preferrably by those who are developers or thereabouts in terms of technical skills. It's hard
work sometimes to slog through near 100 page threads such as the one for the Nexus One FM radio.
If you have simple questions, comments, requests, corrections or additions to this, please consider
PMing me directly and I'll do my best to incorporate that into the thread, giving acknowledgement
if you so desire.
That said, I've posted FM Receiver and FM Transmitter scripts below. If you feel you have a reasonable
capability to try these scripts on your device, please do, and report here or via PM on success or failure
if you might be among the first few to try these on your device/model.
I will try to keep these first posts updated with the latest information, so hopefully you won't need to post
questions about whether or not your device works by referring to the "Devices" post.
---------------------
Introduction
I've enabled the FM radio functions on my HTC Legend. It is also known to work on HTC Tattoo.
Scouring the web looking for the magic incantations to enable FM audio I'm finding myself overwhelmed with all the things I don't know.
Much of the information needed is kept under wraps by TI and their customers. To get TI's information requires signing an NDA,
and perhaps other legal documents. Signing such an NDA limits how much you can say publicly, and I'd prefer to not be under
such constraints. I'm not even sure if an NDA would be sufficient for a person not employed by TI or a TI customer.
Thus this thread, to share with you the information I've found, and to ask for your help in correcting it or adding to it.
If there is any similar thread on any site, that is TI specific, but not device model specific, please let me know.
I've seen and read through a number of mega-threads here and elsewhere that are device specific, but much of the information
contained therein is useful for all devices with TI FM chips.
The chips in question are usually named: WL1271, WL1273, WL1281 and WL1283. The first two have WF + BT + FM and the latter two add GPS.
TI also calls these WiLink or BlueLink 5.0, 6.0 and 7.0, as well as BRF6300, BRF6350 and BL6450.
TI also sells various evaluation boards carrying these chips, and some TI partners produce modules, sometimes with similar numbers.
AFAICT there is no FM functionality in some of the predecessor chips such as the WiLink 4.0 chips: BRF6100 (WL1251) & BRF6150 (WL1253).
Just so you better know my knowledge level:
- I'm new to Android, smartphones and post 1995 PDAs, although I've read some on these subjects over the years.
I'm diving straight in to learn as much as I can as quickly as I can. I'm not currently employed but hope to
transition myself to what appears to be the rapidly expanding Android world.
- I've worked in software development on "semi-embedded" Linux VOIP and security appliances since 1997, with a good bit
of low level kernel/driver stuff. Not much low level stuff recently, mostly daemons and command line utilities.
At home I've recently worked on a home Asterisk VOIP box and MythTv and XBMC based HTPCs. I also manage the 5 Ubuntu
PCs our family uses, as well as one lonely 5 year old HP WinXp tablet.
- My background in electronics and computing goes back to the mid 1970's with 8080 and SCMP. I designed and built a variety
of computer, electronics and even RF devices in those days, and can still wield a mean soldering pencil when needed.
-----------------------------------------------------------------------------------------------------------------------------------------------------
Here is some information about which chips being discussed here. Note that the terms BlueLink and WiLink can be somewhat confusing as some chips are both.
TI's Wireless Connectivity Solutions page:
http://focus.ti.com/general/docs/wt...lateId=6123&navigationId=12493&contentId=4637
Note at left that GPS (NaviLink), Bluetooth (BlueLink) and Wireless (WiLink) are represented.
FM radio seems to be the neglected "step child" that gets little mention or notice.
It's a small add-on feature that just happens to come along with Bluetooth sometimes, if at all.
Generally, the WiFi, BT, FM and GPS components in these multi-function chips tend to be independent of each other.
They can be powered up or down individually and usually have seperate control paths. They are called "IP"s, eg.
the WiFi IP or BT IP. I haven't yet determined what IP stands for, LOL.
FM is the exception though; it seems to piggyback on the BT IP. To power up FM you must first power up BT
(although some doc implied BT can then be powered down). FM has it's own I2C control path, but that is usually
not used, in favor of sharing the BT HCI interface.
Some docs I've read discourage the use of FM. Perhaps it can cause issues ?
Note that some of these chips may indicate support for Wireless-N, but that doesn't mean the device manufacturer
enabled it in their stack etc. It might be possible to enable N, perhaps with different Wireless firmware or init
scripts. While an interesting prospect, despite the expectation of vastly increased battery consumption, I don't
want to get into the Wireless issues, except as they might impact FM.
------------------------------------------------------------------------
Predecessor single function products upon which the later integrated products are based:
2004: BRF6100 / BRF6150 = BT only
http://focus.ti.com/pdfs/wtbu/TI_brf6100_6150.pdf
2005: WiLink 4.0 mWLAN (WL1251 and WL1253) = WF only
WL1251 = 802.11 b/g/e/i/d/k
WL1253 = 802.11 a/b/g/e/i/d/k/h/j
http://focus.ti.com/pdfs/wtbu/wl1251_1253_prod_bulletin.pdf
2005: BlueLink 5.0 BRF6300 = BT only
http://focus.ti.com/pdfs/wtbu/ti_bluelink_5_brf6300.pdf
------------------------------------------------------------------------
Combined products. All seem to support FM Rx and Tx:
2007: BlueLink 6.0 BRF6350 = BT + FM
http://focus.ti.com/pdfs/wtbu/ti_bluelink_6_brf6350.pdf
200?: WiLink 5.0 = WiLink 4.0 mWLAN + BlueLink 6.0 = WF + BT + FM
http://focus.ti.com/general/docs/wt...ateId=6123&navigationId=12661&contentId=15402
2010: WiLink 6.0 = WF + BT + FM (Bluetooth (2.1?) Low Energy Specification 4.0 + EDR)
WL1271 = 802.11 b/g/n (2.4 GHz)
WL1273 = 802.11 a/b/g/n (2.4 & 5 GHz)
http://www.ti.com/lit/swmt013
2010.1: BlueLink / WiLink 7.0 BL6450 = BT 2.1 (+EDR) + FM (No WF)
http://focus.ti.com/pdfs/wtbu/BlueLink7_BL6450_swmt014d.pdf
2010.2: WiLink 7.0 = WF + BT + FM + GPS
WL1281 / WL1283 = 802.11 a/b/g/n + BT 3.0 + FM + GPS 3GPP
http://focus.ti.com/pdfs/wtbu/WiLink7_WL1283_swmt016.pdf

Sources
Sources
The source of information I've found include:
- Documents from TI or customers. Usually these contain limited information.
- Source code from TI, TI's customers or other parties, including ROM builders.
- Threads/Posts on forums including this one, as well as TI support forums.
- Miscellaneous random sources such as IRC logs for HTC-Linux.
The most comprehensive and easily useful sources I've found so far are the source codes
for a Linux WL1271 driver being produced by a Nokia employee, and somewhat similar
source codes from TI.
The "Texas Instruments WL1273 FM radio" Linux driver is under development by Matti J. Aaltonen
of Nokia. I believe the Nokia N900 uses a TI chip under Maemo->Meego, and perhaps other Nokia
devices too. Various patches and discussions are underway, and can be googled, and parts of it
are slowly appearing in the latest kernel source. If you want to see the latest, the only easy way
seems to be downloading one of the latest kernels at https://lkml.org/ .
There's an ancient first version from and some discussion from April here:
http://thread.gmane.org/gmane.linux.drivers.video-input-infrastructure/18449
I used http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.37.tar.bz2 and
http://www.kernel.org/pub/linux/kernel/v2.6/patch-2.6.37.bz2 but I see there's a 2.6.38 RC5
there as well.
Finding and downloading TI code has been a pain, but they have a lot there.
TI WL 128x FM V4L2 driver:
There's a git repository for what appears to be an alternate V4L driver at http://dev.omapzoom.org/pub/scm/manju/L24x-btfm.git
Some discussion: http://www.spinics.net/lists/linux-media/msg28310.html
I'm not sure why there appear to be two efforts underway to create FM V4L2 drivers, the one by Nokia and the other by TI.
This appears to be source for TI's fmapp test utility and fm_stack library in a form that can be viewed by browser:
http://git.omapzoom.org/?p=platform...2a9dcca2dced00e724a2eb1dec578152f5beb;hb=HEAD
I managed to download the older 0.12 version of fmapp and fm_stack source code from somewhere, but can't recall where.
The "fmapp" utility has a LOT of functionality for testing just about every exposed FM feature, including RDS.
There is also a recently released "Android Froyo DevKit V2" at:
http://software-dl.ti.com/dsps/dsps_public_sw/sdo_tii/TI_Android_DevKit/02_00_00/index_FDS.html
You have to sign in for that but it should be easy to create an account. I already had one via a previous TI adventure.
The K2 BM6350 module PDFs have some further info:
http://www.ktwo.co.in/index.php?option=com_content&task=view&id=178&Itemid=465
http://www.ktwo.co.in/pdf/K2BM6350_Datasheet.pdf
http://www.ktwo.co.in/pdf/K2-BM6350 StarterKit UserManual.pdf
Forum threads:
[TUTORIAL] Reverse engineering HTC FM Radio for noobs (on EVO 4G)
http://forum.xda-developers.com/showthread.php?t=725870
Decompiled HTC Radio app
http://martinmarinov.info/HTCRadio.rar
Some words about bluetooth....
http://forum.xda-developers.com/showthread.php?t=816019
[Q] FM Radio app, Broadcom BCM4329 chipset
http://forum.xda-developers.com/showthread.php?t=837691
[THINK TANK] Enabling the Nexus One FM radio ...
http://forum.xda-developers.com/showthread.php?t=707404
FM Radio on 2.x ROMs - An Idea
http://forum.xda-developers.com/showthread.php?p=11366697

Devices
Devices
List of devices and FM chips. At this time I'd like to limit this to Android devices, but might consider others.
Would be useful to list limitations here. For example, some Motorola Droid owners were understandably disheartened,
after much work, to find they had no Fm Rx antenna connection, and could not make one without opening up the cans, etc.
on the board. So technically they had Fm Rx, practically, they had none.
Also, some boards may have no Tx antenna, but might possibly work within a few inches of an external FM receiver antenna.
--------------------------------------------------------------------------------
TI FM devices:
--------------------------------------------------------------------------------
HTC Legend: WL1273
HTC Tattoo/Click WL1271?
HTC Dream/Google G1 WL1271?
HTC Sapphire/Hero BRF6300 = WL1271?
HTC Diamond/Raphael/Blackstone BRF6350 (Windows Mobile?)
Motorola Droid WL1271
Motorola Backflip WL1271?
Motorola Milestone WL1273?
Nokia N900 WL1273? (Maemo?)
Barnes & Noble Nook Color WL1273?
--------------------------------------------------------------------------------
Broadcom FM devices:
--------------------------------------------------------------------------------
HTC Nexus One: 43xx
--------------------------------------------------------------------------------

FM Apps and APIs
FM Apps and APIs
Many handset manufacturers provide their own proprietary FM radio apps. Some people have managed to get an
FM radio app meant for another device working on theirs. Most, however, have library or other issues with a foreign app.
AFAICT, Google has not released any sanctioned FM radio API, nor do they intend to. I'd guess FM radio likely
won't bring much revenue to Google or the carriers.
In an ideal world, Android apps would use the same API as on Linux: the "Video For Linux version Two" aka V4L2.
This API makes use of a /dev/radioX device. This is somewhat similar to the /dev/videoX devices that some devices
appear to support for cameras.
If the V4L2 API was available on Android, Android FM radio apps could then be ported more easily from Linux.
Alas, there are relatively few Linux radio apps. GnomeRadio hasn't been touched in over 2 years and Gnome
doesn't run on Android anyway of course. Some command line apps could be ported, but that doesn't make for
an Android app.
So thus far, the defacto "API" for FM radio on Android has been vendor specific commands over HCI, the
Bluetooth interface. This is more or less similar to the way it can be done via I2C, but apparently
most FM chips are not wired via 12C; they use the existing HCI UART. Once again, FM radio is the
poor neglected "step-child".
One advantage of using HCI is that no new kernel drivers are needed. A disadvantage is that some mediation
driver would be required to use bluetooth and FM at the same time; the only alternative being drivers for
both smashed together, but that would be an Android specific hack and is not a good idea.
I've noted that one individual created an API spec and an app for Windows devices a few years back.
I believe it was called GFMRadio and XFMRadio or similar. That project was apparently abandoned.
MIUI released a GPL licensed FM app for some phones based on broadcom chips; HTC Desire and Nexus One.
The source code contains the string "/dev/radio", but AFAICT it doesn't appear to actually use V4L API.
It speaks directly to the broadcom FM chip via HCI.
Since MIUI source is GPL and available it could be used as a base for a TI, or TI and broadcom specific app.
In theory patches could be submitted to MIUI but I'm not sure they are open to that and the language barrier
from English to Chinese and back may be difficult.
Some interesting posts on MIUI here: http://forum.xda-developers.com/showthread.php?t=837691
http://www.miui.com/thread-1687-1-1.html
Using hcitool commands, or similar, one could write a radio app in bash or Perl etc. LOL.
TI has an "fmapp" command line testing utility that relies on libfm_stack.so .
This app won't run on my Legend because it depends on snd_ctl_* APIs in libaudio.
"strings libfm_stack.so" produces lots of interesting detail and embedded BTS scripts.
The source code I've found for TI fmapp, and it's FM stack library does not seem to have all
the functionality I've seen in the binary fmapp I found. So they may have stripped much code
for the publicly released source code.

Audio routing
Audio routing
On HTC phones, FM analog audio routing can be achieved by:
# Default
adb shell 'echo "disable" > /sys/class/htc_accessory/fm/flag'
# Headset
adb shell 'echo "fm_headset" > /sys/class/htc_accessory/fm/flag'
# Speaker
adb shell 'echo "fm_speaker" > /sys/class/htc_accessory/fm/flag'
# View
adb shell cat /sys/class/htc_accessory/fm/flag

Firmware, TI BTS file, HCI and I2S/I2C issues, tools etc.
Firmware, TI BTS file, HCI and I2S/I2C issues, tools etc.
The Nokia V4L driver loads radio-wl1273-fw.bin, although the code does indicate it may not be necessary.
I can't find this firmware file anywhere. As with many other firmwares, it may simply be a known firmware
file renamed. This has been noted with other firmware files for the TI radio.
Information on firmware for these TI chips seems scattered and incomplete. Same with the BTS Bluetooth script files
which are usually important for accessing the FM functionality.
I think one of the reasons for this lack of information is that most parties do not want us messing with the available functionality.
- Device manufacturers do not want their devices used in violation of FCC or other regulatory body rules.
For example, FM transmission at higher power levels or improper frequencies. Also RDS transmissions with bogus data.
- Device manufacturers and carriers want us to buy newer or more expensive products for additional functionality.
They also would rather we use voice minutes instead of FM "walkie talkies".
- Google and carriers want us to stream music via data rather than pick it up for free from over the air.
...
Where do I find utilities to dump/decode/encode BTS files ?
....
HCI is usually used to access FM functions, but I2C might be usable on some devices.
...
fm_rx_init_6350.1.bts
fm_rx_init_6350.2.bts
fm_rx_init_6450.1.bts
fm_tx_init_6450.1.bts
fmc_init_6350.1.bts
fmc_init_6350.2.bts
fmc_init_6450.1.bts
tiinit_0.0.0.bts
tiinit_5.2.34.bts
tiinit_5.3.53.bts
tiinit_6.1.24.bts
tiinit_6.2.31.bts

HCI/I2C Commands
HCI/I2C Commands
Most of this information is gleaned from:
- The Linux WL1271 FM Radio source code written by a Nokia employee.
- TI source code for fmapp/fmstack, etc.
Various forum posts also make it clear there is a bewildering array of commands etc. not referenced in the source codes above.
Some information can also be retrieved by looking inside BTS, firmware, app, utility and library etc. files.
...
Vendor Specific Opcodes for the various FM-related commands over HCI. (_FmcCoreTransportFmCOmmands)
_FMC_CMD_I2C_FM_READ 0x0133
_FMC_CMD_FM_I2C_FM_READ_HW_REG 0x0134
_FMC_CMD_I2C_FM_WRITE 0x0135
?0x136
_FMC_CMD_FM_POWER_MODE 0x0137
?0x138
_FMC_CMD_FM_SET_AUDIO_PATH 0x0139
_FMC_CMD_FM_CHANGE_I2C_ADDR 0x013A
Format of an HCI READ/WRITE command to FM over I2C is:
HCI Header:
- HCI Packet Type: (Added internally by the HCI Transport Layer)
- HCI Opcode: 2 bytes (LSB, MSB - LE)
- HCI Parameters Total Len: 1 byte (total length of all subsequent fields)
HCI Parameters:
- FM Opcode: 1 byte
- FM Parameters Len: 2 bytes (LSB, MSB - LE)
- FM Cmd Parameter Value: N bytes
For "simple" (non-RDS) read commands "FM Parameters Len" is always "2, 0" (2).
...

HCI/I2C Commands/Opcodes, registers, values
HCI/I2C Opcodes, registers, values.
Most of this information is gleaned from:
- The Linux WL1271 FM Radio source code written by a Nokia employee.
- TI source code for fmapp/fmstack, etc.
---------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------
The most important "commands":
----------
0x137 FM_POWER_MODE: FM Core power up (last byte 0=down, 1=up)
Usage:
# FM_POWER_MODE: FM Core power up
adb shell hcitool cmd 0x3f 0x137 0x01 0x01
# FM_POWER_MODE: FM Core power down
adb shell hcitool cmd 0x3f 0x137 0x01 0x00
----------
0x133 FM_READ
Examples:
# FM_READ: POWER (Register 0x20)
adb shell hcitool cmd 0x3f 0x133 0x20 0x02 0x00
# FM_READ: RSSI (Register 0x01)
adb shell hcitool cmd 0x3f 0x133 0x01 0x02 0x00
----------
0x135 FM_WRITE
Examples:
# FM WRITE: POWER: Rx on (This actually seems to be "audio enable" !
adb shell hcitool cmd 0x3f 0x135 0x20 0x02 0x00 0x00 0x02
# FM WRITE: POWER: Rx on plus RDS
adb shell hcitool cmd 0x3f 0x135 0x20 0x02 0x00 0x00 0x03
----------
------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------
The "registers": (some call them opcodes, but they seem to be registers IMO)
0x00 0 WL1273_STEREO_GET FMC_FW_OPCODE_RX_STEREO_GET
0 FM_STEREO_MODE mono (or no signal ?)
1 FM_MONO_MODE stereo signal
0x01 1 WL1273_RSSI_LVL_GET FMC_FW_OPCODE_RX_RSSI_LEVEL_GET
(-128) SCHAR_MIN FM_RX_RSSI_THRESHOLD_MIN See also WL1273_SEARCH_LVL_SET
127 SCHAR_MAX FM_RX_RSSI_THRESHOLD_MAX
0x02 2 WL1273_IF_COUNT_GET FMC_FW_OPCODE_RX_IF_COUNT_GET
# changes: 1, 2, 3, ff, fe, 0
0x03 3 WL1273_FLAG_GET FMC_FW_OPCODE_CMN_FLAG_GET
? = Event masks ?
#define FM_FR_EVENT (1 << 0)
#define FM_BL_EVENT (1 << 1)
#define FM_RDS_EVENT (1 << 2)
#define FM_BBLK_EVENT (1 << 3)
#define FM_LSYNC_EVENT (1 << 4)
#define FM_LEV_EVENT (1 << 5)
#define FM_IFFR_EVENT (1 << 6)
#define FM_PI_EVENT (1 << 7)
#define FM_PD_EVENT (1 << 8)
#define FM_STIC_EVENT (1 << 9)
#define FM_MAL_EVENT (1 << 10)
#define FM_POW_ENB_EVENT (1 << 11)
0x04 4 WL1273_RDS_SYNC_GET FMC_FW_OPCODE_RX_RDS_SYNC_GET
0 WL1273_RDS_NOT_SYNCHRONIZED
1 WL1273_RDS_SYNCHRONIZED
0x05 5 WL1273_RDS_DATA_GET FMC_FW_OPCODE_RX_RDS_DATA_GET
64 FMC_FW_RX_RDS_THRESHOLD (*1 or *3) See also FMC_FW_OPCODE_RX_RDS_MEM_SET_GET
85 FMC_FW_RX_RDS_THRESHOLD_MAX Used as FMC_FW_RX_RDS_THRESHOLD_MAX*RDS_BLOCK_SIZE for mem size.
? Set to 3e 16 ? (15894)
0x06 ? Set to 1 ?
------------------------------------------------------------------------------------------
Codes 6-9 missing
------------------------------------------------------------------------------------------
0x0a 10 WL1273_FREQ_SET FMC_FW_OPCODE_RX_FREQ_SET_GET
base + freq / 50Khz
0 0x000 87500 WL1273_BAND_OTHER_LOW
410 0x19a 108000 WL1273_BAND_OTHER_HIGH
0 0x000 76000 WL1273_BAND_JAPAN_LOW
280 0x118 90000 WL1273_BAND_JAPAN_HIGH
? #define FM_UNDEFINED_FREQ 0xFFFFFFFF
0x0b 11 WL1273_AF_FREQ_SET FMC_FW_OPCODE_RX_AF_FREQ_SET_GET
0x0c 12 WL1273_MOST_MODE_SET FMC_FW_OPCODE_RX_MOST_MODE_SET_GET ! MOST = "MOno/STereo"
0 WL1273_RX_STEREO Stereo according to blend
1 WL1273_RX_MONO Force mono output
0 FM_STEREO_MODE
1 FM_MONO_MODE
0x0d 13 WL1273_MOST_BLEND_SET FMC_FW_OPCODE_RX_MOST_BLEND_SET_GET
0 Switched blend & hysteresis
1 FM_STEREO_SOFT_BLEND Soft blend
Now set to 1 = Soft blend
0x0e 14 WL1273_DEMPH_MODE_SET FMC_FW_OPCODE_RX_DEMPH_MODE_SET_GET
0 FM_RX_EMPHASIS_FILTER_50_USEC
1 FM_RX_EMPHASIS_FILTER_75_USEC
0x0f 15 WL1273_SEARCH_LVL_SET FMC_FW_OPCODE_RX_SEARCH_LVL_SET_GET
7 WL1273_DEFAULT_SEEK_LEVEL
(-128) SCHAR_MIN See also WL1273_RSSI_LVL_GET
127 SCHAR_MAX
0x10 16 WL1273_BAND_SET FMC_FW_OPCODE_RX_BAND_SET_GET
0 WL1273_BAND_OTHER 87.5-108 Mhz North America, Europe, generally rest of world besides Japan
1 WL1273_BAND_JAPAN 76-90 Mhz Japan (perhaps soon US also)
0x11 17 WL1273_MUTE_STATUS_SET FMC_FW_OPCODE_RX_MUTE_STATUS_SET_GET
0 ........ FMC_FW_RX_MUTE_UNMUTE_MODE
bit0 0x01 WL1273_MUTE_SOFT_ENABLE FMC_FW_RX_MUTE_RF_DEP_MODE
bit1 0x02 WL1273_MUTE_AC FMC_FW_RX_MUTE_AC_MUTE_MODE
bit2 0x04 WL1273_MUTE_HARD_LEFT FMC_FW_RX_MUTE_HARD_MUTE_LEFT_MODE
bit3 0x08 WL1273_MUTE_HARD_RIGHT FMC_FW_RX_MUTE_HARD_MUTE_RIGHT_MODE
bit4 0x10 WL1273_MUTE_SOFT_FORCE FMC_FW_RX_MUTE_SOFT_MUTE_FORCE_MODE
Set to one of these:
2 0x02 FM_RX_AC_MUTE_MODE Mute On But enable soft/attenuate ?
0 0x00 FM_RX_UNMUTE_MODE Mute Off
16 0x10 FM_RX_SOFT_MUTE_FORCE_MODE Mute Attenuate
Then optionally logically "OR" ('|') this:
bit0 0x01 FM_RX_RF_DEP_MODE
Optional bits ?
bit2 0x04 FM_RX_HARD_MUTE_LEFT_MODE
bit3 0x08 FM_RX_HARD_MUTE_RIGHT_MODE
? #define FM_MUTE_OFF 0
? #define FM_MUTE_ON 1
? #define FM_MUTE_ATTENUATE 2
? #define FM_RX_RF_DEPENDENT_MUTE_ON 1
? #define FM_RX_RF_DEPENDENT_MUTE_OFF 0
0x12 18 WL1273_RDS_PAUSE_LVL_SET FMC_FW_OPCODE_RX_RDS_PAUSE_LVL_SET_GET
? Set to 5 ?
0x13 19 WL1273_RDS_PAUSE_DUR_SET FMC_FW_OPCODE_RX_RDS_PAUSE_DUR_SET_GET
? Set to 0x0c = 12
0x14 20 WL1273_RDS_MEM_SET FMC_FW_OPCODE_RX_RDS_MEM_SET_GET
64 FMC_FW_RX_RDS_THRESHOLD (*1 or *3) See also FMC_FW_OPCODE_RX_RDS_DATA_GET
85 FMC_FW_RX_RDS_THRESHOLD_MAX Used as FMC_FW_RX_RDS_THRESHOLD_MAX*RDS_BLOCK_SIZE for mem size.
Set to 0x55 = 85 = Max Thresh
0x15 21 WL1273_RDS_BLK_B_SET FMC_FW_OPCODE_RX_RDS_BLK_B_SET_GET
0x16 22 WL1273_RDS_MSK_B_SET FMC_FW_OPCODE_RX_RDS_MSK_B_SET_GET
0x17 23 WL1273_RDS_PI_MASK_SET FMC_FW_OPCODE_RX_RDS_PI_MASK_SET_GET
0x18 24 WL1273_RDS_PI_SET FMC_FW_OPCODE_RX_RDS_PI_SET_GET
0x19 25 WL1273_RDS_SYSTEM_SET FMC_FW_OPCODE_RX_RDS_SYSTEM_SET_GET
0 FM_RDS_SYSTEM_RDS
1 FM_RDS_SYSTEM_RBDS
0x1a 26 WL1273_INT_MASK_SET FMC_FW_OPCODE_CMN_INT_MASK_SET_GET
0x1b 27 WL1273_SEARCH_DIR_SET FMC_FW_OPCODE_RX_SEARCH_DIR_SET_GET
0 FM_SEARCH_DIRECTION_DOWN
1 FM_SEARCH_DIRECTION_UP
0x1c 28 WL1273_VOLUME_SET FMC_FW_OPCODE_RX_VOLUME_SET_GET
880 0x370 ........ FMC_FW_RX_FM_GAIN_STEP ? 35 steps ?
0 0x00 ........ FMC_FW_RX_FM_VOLUMN_MIN
30904 0x78b8 WL1273_DEFAULT_VOLUME FMC_FW_RX_FM_VOLUMN_INITIAL_VALUE
61808 0xf170 ........ FMC_FW_RX_FM_VOLUMN_MAX
65535 0xffff WL1273_MAX_VOLUME ........
- #define FM_RX_VOLUME_MIN 0
? #define FM_RX_VOLUME_MAX 70
? #define FM_RX_VOLUME_GAIN_STEP 0x370
0x1d 29 WL1273_AUDIO_ENABLE FMC_FW_OPCODE_RX_AUDIO_ENABLE_SET_GET
bit0 0x01 WL1273_AUDIO_ENABLE_I2S FMC_FW_RX_FM_AUDIO_ENABLE_I2S
bit1 0x02 WL1273_AUDIO_ENABLE_ANALOG FMC_FW_RX_FM_AUDIO_ENABLE_ANALOG
bit0|1 0x03 ........ FMC_FW_RX_FM_AUDIO_ENABLE_I2S_AND_ANALOG
0 ........ FMC_FW_RX_FM_AUDIO_ENABLE_DISABLE
0x1e 30 WL1273_PCM_MODE_SET FMC_FW_OPCODE_CMN_I2S_CLOCK_CONFIG_SET_GET
0 0x00 WL1273_PCM_DEF_MODE ? I2S protocol, left channel first, data width 16 bits
0x1f 31 WL1273_I2S_MODE_CONFIG_SET FMC_FW_OPCODE_CMN_I2S_MODE_CONFIG_SET_GET
0x0145= WL1273_IS2_RATE_48K(0) | IS2_TRI_OPT(0) | IS2_SDOWS_RF(0x0100) |
IS2_SLAVEW(0x0040) | IS2_FORMAT_STD(0) | IS2_WIDTH_50(0x0005)
0 0x0 WL1273_IS2_WIDTH_32
1 0x1 WL1273_IS2_WIDTH_40
2 0x2 WL1273_IS2_WIDTH_22_23
3 0x3 WL1273_IS2_WIDTH_23_22
4 0x4 WL1273_IS2_WIDTH_48
5 0x5 WL1273_IS2_WIDTH_50
6 0x6 WL1273_IS2_WIDTH_60
7 0x7 WL1273_IS2_WIDTH_64
8 0x8 WL1273_IS2_WIDTH_80
9 0x9 WL1273_IS2_WIDTH_96
10 0xa WL1273_IS2_WIDTH_128
bits0-3 0xf WL1273_IS2_WIDTH 0xf Mask
........
0 0x00 WL1273_IS2_FORMAT_STD (0x0 << 4)
16 0x10 WL1273_IS2_FORMAT_LEFT (0x1 << 4)
32 0x20 WL1273_IS2_FORMAT_RIGHT (0x2 << 4)
48 0x30 WL1273_IS2_FORMAT_USER (0x3 << 4)
........
0 0x00 WL1273_IS2_MASTER (0x0 << 6)
64 0x40 WL1273_IS2_SLAVEW (0x1 << 6)
........
0 0x00 WL1273_IS2_TRI_AFTER_SENDING (0x0 << 7)
128 0x80 WL1273_IS2_TRI_ALWAYS_ACTIVE (0x1 << 7)
........
0 0x00 WL1273_IS2_SDOWS_RR (0x0 << 8)
256 0x100 WL1273_IS2_SDOWS_RF (0x1 << 8)
512 0x200 WL1273_IS2_SDOWS_FR (0x2 << 8)
768 0x300 WL1273_IS2_SDOWS_FF (0x3 << 8)
........
0 0x00 WL1273_IS2_TRI_OPT (0x0 << 10)
1024 0x400 WL1273_IS2_TRI_ALWAYS (0x1 << 10)
........
0 0x00 WL1273_IS2_RATE_48K (0x0 << 12)
4096 0x1000 WL1273_IS2_RATE_44_1K (0x1 << 12)
8192 0x2000 WL1273_IS2_RATE_32K (0x2 << 12)
16384 0x4000 WL1273_IS2_RATE_22_05K (0x4 << 12) ?! No 0x3, 0x6-0x7, 0xb-0xe ?
20480 0x5000 WL1273_IS2_RATE_16K (0x5 << 12)
32768 0x8000 WL1273_IS2_RATE_12K (0x8 << 12)
36864 0x9000 WL1273_IS2_RATE_11_025 (0x9 << 12)
40960 0xa000 WL1273_IS2_RATE_8K (0xa << 12)
61440 0xf000 WL1273_IS2_RATE (0xf << 12) Mask
........
0 WL1273_I2S_DEF_MODE WL1273_IS2_WIDTH_32| WL1273_IS2_FORMAT_STD| WL1273_IS2_MASTER| WL1273_IS2_TRI_AFTER_SENDING|
WL1273_IS2_SDOWS_RR| WL1273_IS2_TRI_OPT| WL1273_IS2_RATE_48K
0x20 32 WL1273_POWER_SET FMC_FW_OPCODE_RX_POWER_SET_GET
0 WL1273_POWER_SET_OFF FMC_FW_RX_POWER_SET_FM_AND_RDS_OFF
! 0 just seems to mute output. RSSI still responds and all registers remain set.
bit0 0x01 WL1273_POWER_SET_FM FMC_FW_RX_POWER_SET_FM_ON_RDS_OFF
bit1 0x02 WL1273_POWER_SET_RDS ........
bit0|1 0x03 ........ FMC_FW_RX_POWER_SET_FM_AND_RDS_ON
bit4 0x10 WL1273_POWER_SET_RETENTION ........
0x21 33 WL1273_INTX_CONFIG_SET FMC_FW_OPCODE_CMN_INTX_CONFIG_SET_GET
0x22 34 WL1273_PULL_EN_SET FMC_FW_OPCODE_CMN_PULL_EN_SET_GET
? Set to 0xff = 255 ?
0x23 35 WL1273_HILO_SET FMC_FW_OPCODE_RX_HILO_SET_GET
0 0x0 FM_RX_IFFREQ_TO_HI_SIDE
1 0x1 FM_RX_IFFREQ_TO_LO_SIDE
2 0x2 FM_RX_IFFREQ_HILO_AUTOMATIC
Set to 1 = FM_RX_IFFREQ_TO_LO_SIDE
0x24 36 WL1273_SWITCH2FREF FMC_FW_OPCODE_CMN_SWITCH_2_FREF_SET
0x25 37 WL1273_FREQ_DRIFT_REPORT FMC_FW_OPCODE_CMN_FREQ_DRIFT_REPORT_SET TI fmc_fw_defs.h error defines as 0x24
0x28 40 WL1273_PCE_GET FMC_FW_OPCODE_CMN_PCE_GET
Set to 0x0f = 15
0x29 41 WL1273_FIRM_VER_GET FMC_FW_OPCODE_CMN_FIRM_VER_GET
Set to 2
0x2a 42 WL1273_ASIC_VER_GET FMC_FW_OPCODE_CMN_ASIC_VER_GET
Set to 2
0x2b 43 WL1273_ASIC_ID_GET FMC_FW_OPCODE_CMN_ASIC_ID_GET
Set to 0x1273 = 4723
0x2c 44 WL1273_MAN_ID_GET FMC_FW_OPCODE_CMN_MAN_ID_GET
Set to 0x17 = 23
0x2d 45 WL1273_TUNER_MODE_SET FMC_FW_OPCODE_RX_TUNER_MODE_SET
0 TUNER_MODE_STOP_SEARCH FMC_FW_RX_TUNER_MODE_STOP_SEARCH
1 TUNER_MODE_PRESET FMC_FW_RX_TUNER_MODE_PRESET_MODE
2 TUNER_MODE_AUTO_SEEK FMC_FW_RX_TUNER_MODE_AUTO_SEARCH_MODE (AUTONOMOUS)
3 TUNER_MODE_AF FMC_FW_RX_TUNER_MODE_ALTER_FREQ_JUMP
4 TUNER_MODE_AUTO_SEEK_PI ........
5 TUNER_MODE_AUTO_SEEK_BULK ........
0x2e 46 WL1273_STOP_SEARCH FMC_FW_OPCODE_RX_STOP_SEARCH
0x2f 47 WL1273_RDS_CNTRL_SET FMC_FW_OPCODE_RX_RDS_CNTRL_SET
1 FMC_FW_RX_RDS_FLUSH_FIFO
------------------------------------------------------------------------------------------
Codes 48-51 (0x30-0x33) missing
------------------------------------------------------------------------------------------
0x32 ? Set to 0xC000 = 49152
------------------------------------------------------------------------------------------
0x34 52 WL1273_SOC_INT_TRIGGER
------------------------------------------------------------------------------------------
Code 53 (0x35) missing
Code 54 (0x36) and up are mostly TX, except:
0x57 87 WL1273_RX_ANTENNA_SELECT ........
and the common/CMN values 100-102, 254, 255
------------------------------------------------------------------------------------------
0x36 54 WL1273_TX_AUDIO_INPUT_LEVEL_RANGE_SET
0x37 55 WL1273_CHANL_SET FMC_FW_OPCODE_TX_CHANL_SET_GET
freq / 10Khz
7600 0x1db0 76000 WL1273_BAND_TX_LOW
10800 0x2a30 108000 WL1273_BAND_TX_HIGH
0x38 56 WL1273_SCAN_SPACING_SET FMC_FW_OPCODE_TX_CHANL_BW_SET_GET
1 WL1273_SPACING_50kHz FMC_FW_TX_CHANNEL_BW_50_KHZ
2 WL1273_SPACING_100kHz FMC_FW_TX_CHANNEL_BW_100_KHZ
4 WL1273_SPACING_200kHz FMC_FW_TX_CHANNEL_BW_200_KHZ
Set to 4 = 200 KHz
1 0x1 FM_CHANNEL_SPACING_50KHZ
2 0x2 FM_CHANNEL_SPACING_100KHZ
4 0x4 FM_CHANNEL_SPACING_200KHZ
0x39 57 WL1273_REF_SET ........
0x3a 58 WL1273_POWER_ATT_SET ........
0x3b 59 WL1273_POWER_LEV_SET FMC_FW_OPCODE_TX_POWER_LEVEL_SET_GET
Set to 4
/* Range for TX power level in units for dB/uV */ ! 122-pwr
#define FM_PWR_LVL_LOW 91
#define FM_PWR_LVL_HIGH 122
/* Chip specific default TX power level value */
#define FM_PWR_LVL_DEF 4
0x3c 60 WL1273_AUDIO_DEV_SET ........
0x109 = 265 ?
0x3d 61 WL1273_PILOT_DEV_SET ........
0x1b = 27
0x3e 62 WL1273_RDS_DEV_SET ........
8
0x3f 63 WL1273_AUDIO_IO_SET FMC_FW_OPCODE_TX_AUDIO_IO_SET
0 WL1273_AUDIO_IO_SET_ANALOG FMC_FW_TX_AUDIO_IO_SET_ANALOG
1 WL1273_AUDIO_IO_SET_I2S FMC_FW_TX_AUDIO_IO_SET_I2S
0x40 64 WL1273_PREMPH_SET FMC_FW_OPCODE_TX_PREMPH_SET_GET
0 FM_TX_PREEMPH_50US FM TX Pre-emphasis filter default ?
1 FM_TX_PREEMPH_OFF
2 FM_TX_PREEMPH_75US
0x41 65 TX_BAND_SET !!??
0x42 66 WL1273_MONO_SET FMC_FW_OPCODE_TX_MONO_SET_GET
0 WL1273_TX_MONO
1 WL1273_TX_STEREO
1 by default
0x43 67 WL1273_MPX_LMT_ENABLE
0x44 68 ........ FMC_FW_OPCODE_TX_PI_CODE_SET_GET
0x45 69 WL1273_ECC_SET FMC_FW_OPCODE_TX_RDS_ECC_SET_GET
0x46 70 WL1273_PTY FMC_FW_OPCODE_TX_RDS_PTY_CODE_SET_GET
0x47 71 WL1273_AF FMC_FW_OPCODE_TX_RDS_AF_SET_GET
------------------------------------------------------------------------------------------
Code 72 (0x48) missing
------------------------------------------------------------------------------------------
0x49 73 WL1273_TX_AUDIO_LEVEL_TEST_THRESHOLD ........
0x4a 74 WL1273_DISPLAY_MODE FMC_FW_OPCODE_TX_RDS_PS_DISPLAY_MODE_SET_GET
0 FMC_FW_TX_RDS_PS_DISPLAY_MODE_SCROLL_OFF
1 FMC_FW_TX_RDS_PS_DISPLAY_MODE_SCROLL_ON
0x4d 77 WL1273_RDS_REP_SET FMC_FW_OPCODE_TX_RDS_REPERTOIRE_SET_GET
0x4e 78 WL1273_TA_SET FMC_FW_OPCODE_TX_RDS_TA_SET
0x4f 79 WL1273_TP_SET FMC_FW_OPCODE_TX_RDS_TP_SET
0x50 80 WL1273_DI_SET FMC_FW_OPCODE_TX_RDS_DI_CODES_SET_GET
0x51 81 WL1273_MS_SET FMC_FW_OPCODE_TX_RDS_MUSIC_SPEECH_FLAG_SET_GET
0x52 82 WL1273_PS_SCROLL_SPEED FMC_FW_OPCODE_TX_RDS_PS_SCROLL_SPEED_SET_GET
0x53 83 WL1273_SOC_AUDIO_PATH_SET ........
0x54 84 WL1273_SOC_PCMI_OVERRIDE ........
0x55 85 WL1273_SOC_I2S_OVERRIDE ........
0x56 86 WL1273_I2C_DEV_ADDR_SET ........
default 0x22 = 34 (Nokia: #define RX71_FM_I2C_ADDR 0x22)
0x57 87 WL1273_RX_ANTENNA_SELECT ........
0x58 88 WL1273_REF_ERR_CALIB_PARAM_SET ........
0x0c = 12
0x59 89 WL1273_REF_ERR_CALIB_PERIODICITY_SET ........
0x5a 90 WL1273_POWER_ENB_SET FMC_FW_OPCODE_TX_POWER_ENB_SET
0 FMC_FW_TX_POWER_DISABLE
1 FMC_FW_TX_POWER_ENABLE
0x5b 91 WL1273_PUPD_SET FMC_FW_OPCODE_TX_POWER_UP_DOWN_SET
0 WL1273_PUPD_SET_OFF FMC_FW_TX_POWER_DOWN
bit0 1 WL1273_PUPD_SET_ON FMC_FW_TX_POWER_UP
bit4 0x10 WL1273_PUPD_SET_RETENTION ........
0x5c 92 WL1273_MUTE FMC_FW_OPCODE_TX_MUTE_MODE_SET_GET
0 FMC_FW_TX_UNMUTE
1 FMC_FW_TX_MUTE
0x5d 93 WL1273_PI_SET ........
0x5e 94 WL1273_RDS_DATA_ENB FMC_FW_OPCODE_TX_RDS_DATA_ENB_SET_GET
0 FMC_FW_TX_RDS_ENABLE_STOP
1 FMC_FW_TX_RDS_ENABLE_START
0x5f 95 WL1273_RSSI_BLOCK_SCAN_FREQ_SET ........
0x60 96 WL1273_TX_AUDIO_LEVEL_TEST ........
0x61 97 WL1273_RSSI_BLOCK_SCAN_START ........
0x62 98 WL1273_RDS_CONFIG_DATA_SET FMC_FW_OPCODE_TX_RDS_CONFIG_DATA_SET
0x63 99 WL1273_RDS_DATA_SET FMC_FW_OPCODE_TX_RDS_DATA_SET
0x64 100 WL1273_WRITE_HARDWARE_REG FMC_FW_OPCODE_CMN_HARDWARE_REG_SET_GET
0x65 101 WL1273_CODE_DOWNLOAD FMC_FW_OPCODE_CMN_CODE_DOWNLOAD
0x66 102 WL1273_RESET FMC_FW_OPCODE_CMN_RESET
0x0f00 = 3840
------------------------------------------------------------------------------------------
Code 103 (0x67) missing
------------------------------------------------------------------------------------------
0x68 104 WL1273_READ_FMANT_TUNE_VALUE ........ TX tuning capacitor value
?
/* FM TX antenna impedence values */
#define FM_TX_ANT_IMP_50 0
#define FM_TX_ANT_IMP_200 1
#define FM_TX_ANT_IMP_500 2
------------------------------------------------------------------------------------------
Codes 105-253 (0x69-0xfd) missing
------------------------------------------------------------------------------------------
0xfe 254 WL1273_FM_POWER_MODE ........
0 FMC_FW_RX_FM_POWER_MODE_DISABLE
1 FMC_FW_RX_FM_POWER_MODE_ENABLE
0xff 255 WL1273_FM_INTERRUPT ........
------------------------------------------------------------------------------------------
?
5 WL1273_RSSI_BLOCK_SCAN_DATA_GET RSSI_BLOCK_SCAN_DATA_GET
------------------------------------------------------------------------------------------
/*
Maximum length of data that may be sent in a single RDS data set command
Once FM FW team removes internal limitations, HCI limitations (much
longer) may apply.
In case a longer RDS data should be sent to the chip, it is divided into
multiple chunks, each chunk being up to FMC_FW_TX_MAX_RDS_DATA_SET_LEN
bytes long
*/
#define FMC_FW_TX_MAX_RDS_DATA_SET_LEN ((FMC_UINT)30)
/*
Defines the max length of data that can be written to FM Hardware register
*/
#define FMC_FW_WRITE_HARDWARE_REG_MAX_DATA_LEN ((FMC_UINT)HCI_CMD_PARM_LEN)
? HCI_CMD_PARM_LEN ?
------------------------------------------------------------------------------------------
Event masks:
bit, 2 hex bytes, (1), (2)
0 0x0001 WL1273_FR_EVENT FMC_FW_MASK_FR Tuning Operation Ended
1 0x0002 WL1273_BL_EVENT FMC_FW_MASK_BL Band limit was reached during search
2 0x0004 WL1273_RDS_EVENT FMC_FW_MASK_RDS RDS data threshold reached in FIFO buffer
3 0x0008 WL1273_BBLK_EVENT FMC_FW_MASK_BBLK RDS B block match condition occurred
4 0x0010 WL1273_LSYNC_EVENT FMC_FW_MASK_LSYNC RDS sync was lost
5 0x0020 WL1273_LEV_EVENT FMC_FW_MASK_LEV RSSI level has fallen below the threshold configured by SEARCH_LVL_SET
6 0x0040 WL1273_IFFR_EVENT FMC_FW_MASK_IFFR Received signal frequency is out of range
7 0x0080 WL1273_PI_EVENT FMC_FW_MASK_PI RDS PI match occurred
8 0x0100 WL1273_PD_EVENT FMC_FW_MASK_PD Audio pause detect occurred
9 0x0200 WL1273_STIC_EVENT FMC_FW_MASK_STIC Stereo indication changed
10 0x0400 WL1273_MAL_EVENT FMC_FW_MASK_MAL Hardware malfunction
11 0x0800 WL1273_POW_ENB_EVENT FMC_FW_MASK_POW_ENB Tx Power Enable/Disable
12 0x1000 WL1273_SCAN_OVER_EVENT FMC_FW_MASK_INVALID_PARAM
13 0x2000 WL1273_ERROR_EVENT !! One of the above is wrong !!
---------------------------------------------------------------------------------------------------------

FM Receiver script
---null set---

FM Transmitter script
---null set---

I think you are doing an awsome job and I really hope that you'll succeded. But the reason why I'm writing this post is to get this thread on the first page again for a while, so maybe more developers will see it, and can contribute !
Good luck!
EDIT: Perhaps it just needs more outstandig title, maybe Unviersal FM radio for android devices with TI WL chips, or something that would get people to read it.

qzem said:
I think you are doing an awsome job and I really hope that you'll succeded. But the reason why I'm writing this post is to get this thread on the first page again for a while, so maybe more developers will see it, and can contribute !
Good luck!
EDIT: Perhaps it just needs more outstandig title, maybe Unviersal FM radio for android devices with TI WL chips, or something that would get people to read it.
Click to expand...
Click to collapse
It's already off the first page, LOL.
My plan has been to post a link to this thread in the various existing threads for different devices using the TI FM chips. I'm sure that will get this thread some notice. I think a lot of devs and dev types stick to the forums for their devices and don't look at this general section.
At least there are so many potentially matching keywords in the first 10 posts that google searches on the subject are likely to link here.
I agonized over the thread title name for some time and "TI FM Radio" is best description I could think of, technically at least. I don't know if I can rename the thread, but it might help to put the names of popular devices with TI chips in the title.
As I posted on the Legend thread, I now have an App Inventor app up and running with functionality to tune, scan, change volume and see signal strength. The audio routing is the last major piece of the puzzle, but it may be different on different devices.
I'll spend a few more days at most to try and get audio routing working, and then, whether working or not, I'll post in a few threads looking for further info and people who want to try the app I'm building.
Would be nice to see the RDS and transmitter working soon too.

Regarding merging the bluetooth and FM drivers, TI's solution appears to be what they call a shared transport line discipline driver. The way this should work, each driver has what appears to it to be dedicated access to it's respective core, and the line discipline driver takes care of any queueing or delaying of commands & such that has to happen to keep them from stepping on each others toes.
Oh, and "IP" is "Intellectual Property" in this case. So the core, generally you'd say the WL127x has a wifi core, bluetooth core, etc. and 128x has a gps core as well. An IP core uses a description language (it used to be VHDL) to describe the layout of the core, so for instance if a company wants to build wifi onto their own chip, they can buy use of the IP core from TI instead of having to buy a phyiscal chip and interface to it.
I've got a debian install wedged onto my Droid 2 Global, I'm going to look into the "ti-st" V4L2 FM drivers, and see if I can get a module that will insert. The kernel can't be replaced on D2G yet, but as far as I know if I get a 2.6.32.9 kernel tree, and get ti-st driver to compile under it, I don't see why it shouldn't insert as a module just fine. Also, I'll look REAL closely to see if I can discern how it gets audio out, so I might have an hcitool command or two to add if that pans out.

hwertz said:
Regarding merging the bluetooth and FM drivers, TI's solution appears to be what they call a shared transport line discipline driver. The way this should work, each driver has what appears to it to be dedicated access to it's respective core, and the line discipline driver takes care of any queueing or delaying of commands & such that has to happen to keep them from stepping on each others toes.
Oh, and "IP" is "Intellectual Property" in this case. So the core, generally you'd say the WL127x has a wifi core, bluetooth core, etc. and 128x has a gps core as well. An IP core uses a description language (it used to be VHDL) to describe the layout of the core, so for instance if a company wants to build wifi onto their own chip, they can buy use of the IP core from TI instead of having to buy a phyiscal chip and interface to it.
I've got a debian install wedged onto my Droid 2 Global, I'm going to look into the "ti-st" V4L2 FM drivers, and see if I can get a module that will insert. The kernel can't be replaced on D2G yet, but as far as I know if I get a 2.6.32.9 kernel tree, and get ti-st driver to compile under it, I don't see why it shouldn't insert as a module just fine. Also, I'll look REAL closely to see if I can discern how it gets audio out, so I might have an hcitool command or two to add if that pans out.
Click to expand...
Click to collapse
Thanks for the info hwertz .
OK I understand "IP" now; never heard it used in that way to designate blocks on a chip.
Yes I read something about the line discipline. Some block diagrams: http://omappedia.org/wiki/Wilink_ST
If you know of any HCI commands dealing with audio routing, please post or pm whatever info you can share.
So I'd guess your opinion is that the v4l2 api is the best route to support FM radios on Android ? That was among my first thoughts until I saw that there is currently virtually no such support on any Android ROM I've heard of. HCI seems to work fine, but it requires chip specific commands of course.
Clearly though, at least Nokia and TI both are working on efforts to bring V4L2 apis for the TI chip in the embedded linux or Android environments.
I'm not sure how audio routing would be configured on Android when using v4l2 apis. The PC environment requires moving digital data from source to destination. But SOC devices often can move digital or analog data directly and without software support.

FM Transmitting Radius
So I hope this question is not too basic for this forum, but I'm new to it and wonder how large the FM transmitting radius of such a chip might be. I basically just need to get an idea, but of course I'd also be thankful if you can refer me to all sorts of literature, specs, overviews, etc.
Thanks!

Lipton1 said:
So I hope this question is not too basic for this forum, but I'm new to it and wonder how large the FM transmitting radius of such a chip might be. I basically just need to get an idea, but of course I'd also be thankful if you can refer me to all sorts of literature, specs, overviews, etc.
Thanks!
Click to expand...
Click to collapse
This thread is over 2 years old now. The original purpose was to try and find others interested in sharing the undocumented secrets of TI's FM chip.
But after my info dump, nobody showed up to share with me, and likely nobody will, so I will close it after this post.
Any further discussion about FM can move to the Spirit FM thread in my sig.
These chips put out tiny amounts of power in transmit mode, maybe 10-30 milliwatts or so. The only phones I have that transmit have to have their headset cable antenna wrapped around the receiver antenna to get anything resembling decent quality.
So I'd call the transmit radius a few centimetres at most. Little wonder then, perhaps, why so few Android devices support transmit.

[DEV][TUTORIAL] How to poll GPIO values from Haret

In case you need to know how to poll GPIO values from photon, here is a small how-to:
First of all, read THIS:
http://htc-linux.org/wiki/index.php?title=HaRET_Documentation#Polling_GPIOs
To disable automatic launch of Android when you execute HarET, you must rename your file "startup.txt" to "default.txt"
The issue is, when you poll for GPIOs with Haret, you got a numeric value that does not match with GPIOs in Linux (board-photon.h)
To translate "HarET GPIO" to "Linux GPIO", here is a trick:
Haret gives you this line:
009.179 GPIOS 92f00808(117)=300100
Click to expand...
Click to collapse
Take the number between parenthesis: 117
Divide it by 32:
117/32 = 3, rest = 21
3 is the GPIO bank:
bank0 = 0 to 15
bank1 = 16 to 42
bank2 = 43 to 67
bank3 = 68 to 94
bank4 = 95 to 106
bank5 = 107 to 121
in our example, bank3 starts at value 68, so the Linux GPIO is:
68 + 21(rest of division) = 89
note: if the bank number is higher than 5, it means it is the following bank, or the one after. in this case, remove 6 or 12 to the bank number so you have a value between 0 and 5, and do the normal computation.

I must to say THANK YOU!

hmm I not understand (rest) I got 3,65625 (bank 3, rest 65626)
r0bin said:
117/32 = 3, rest = 21
Click to expand...
Click to collapse
Also next situation:
057.215 GPIOS 92f0083c(313)=61c8102
9,78125 (bank 9, rest 78125)

munjeni said:
hmm I not understand (rest) I got 3,65625 (bank 3, rest 65626)
Also next situation:
057.215 GPIOS 92f0083c(313)=61c8102
9,78125 (bank 9, rest 78125)
Click to expand...
Click to collapse
lol
on my sample, another way to say it:
117/32 = (3 * 32) + 21
or:
117/32 = 3.65625, you take the 3, rest 0.65625 * 32 = 21
in your case:
303/32 = 9.78125, you take the 9, rest 0.78125 * 32 = 25
9 is too big, remove 6 => it's bank3
bank3 = 68+25 = 93
in board-photon.h, it is defined like:
LIBERTY_GPIO_LS_EN 93
in the code (board-photon.c) it seems it's the GPIO for power

thanks, understand now \o/

Hi,
I got slightly different format for GPIOS.
Please refer to here
http://pastebin.com/xr4p3cd8
How to I convert them?
Thanks in advance

Enable GPU 5th Step?

I was looking around in the gpu driver files and I noticed that there was a hidden 5th 533MHz OC in there. I tried to enable it by copying in the "asv_3d_volt_9_prime" table in for the "non-prime" table. I made sure to remove the -1 in the non prime's table so it wouldn't think the extra voltages for 533MHz was out of bounds. I made sure to comment out these crippling lines:
Code:
/*if ((samsung_rev() < EXYNOS4412_REV_2_0) && (maliDvfsStatus.currentStep == 3))
{
level=get_mali_dvfs_status();
}*/ //a crippling check for "prime" note 2?
I also (after looking at siyah's kernel) copied in the entries for the step4 sysfs variables into mali_kernel_linux.c but my cell phone seems to be bent on not giving me this 5th step.
Sorry again for asking a question that's probably obvious to elite developers. Oh and thank you for answering my question on adding frequencies to the cpu. I now have my own (not copy and pasted :cyclops: ) lines in there to give me 2OC frequency and 1 underclock frequency.

Where's that code snippet from? Anyway it's not related to the limiting if the 5th step.
Code:
static mali_bool mali_dvfs_table_update(void)
{
unsigned int i;
unsigned int step_num = MALI_DVFS_STEPS;
//if (samsung_rev() < EXYNOS4412_REV_2_0)
//step_num = MALI_DVFS_STEPS - 1;
Just comment out those two lines in the DVFS decision making and it should scale up to 5 steps.

AndreiLux said:
Where's that code snippet from? Anyway it's not related to the limiting if the 5th step.
Code:
static mali_bool mali_dvfs_table_update(void)
{
unsigned int i;
unsigned int step_num = MALI_DVFS_STEPS;
//if (samsung_rev() < EXYNOS4412_REV_2_0)
//step_num = MALI_DVFS_STEPS - 1;
Just comment out those two lines in the DVFS decision making and it should scale up to 5 steps.
Click to expand...
Click to collapse
I tried that already. It didn't work. The code snippet I pasted comes from the same file.

Big Bang steering wheel interface support page

Support Page for the Big Bang Steering Wheel Interface
The purpose of this page is to provide support and answers to questions for the Big Bang Steering Wheel interface until our Wiki and support site at UnderGroundElectronics.net is finished.
We are discussing the Big Bang Steering Wheel interface project on Kickstarter.
A Quick Description: The Big Bang is a interface between the car and your tablet. Our goal is to enable you to install your tablet in your car with OEM like integration. We had to start somewhere and we are planning other items for the future but this critical first step we felt was the most nessasary thing for a every day user experiance.
We know that there are a lot of people and threads on installing tablets in cars. We've seen various methods to achieve this solution but no one has made a item for this specific purpose. Even if your car does not have steering wheel controls, this device allows for the waking and sleeping of the tablet while still allowing it to charge constantly.
This device is fairly simple in operation, however the automobiles they go in are not so simple. There are lot of things that can cause problems and we hope to address them here (until we can get our site up, and maybe even after since XDA is a go to place for trouble shooting problems with our tablets and phones).
Lets Start with a basic description of operation
Most autos that have steering wheel controls use a single wire behind the head unit. This wire is typically one of three types: Resistive Negative, Resistive Positive, and Digital. A few Autos use independent wires, one example is Harley Davison. Each button is independent and thus must use resisters to combine the buttons to a single wire.
Currently the Big Bang is made to work on Negative and positive resistive Autos. Data type is expected to be developed by the time we ship.
Finding the wire in your auto:
At the steering column is the best way. But BE WARNED, there is danger probing around with the electrical system of any car that has airbags. You should disconnect the battery and let the car sit at least 30 minutes before doing anything with the wiring in a car with air bags. (Yes we realize that this is very cautious stand but we don't want people getting hurt. If you are un-experienced with Automotive electrical or unsure of what you are doing you probably should consult you local Car Stereo Shop. At the very least use sites like "the 12volt .com" or ask for assistance here. AND NEVER EVER USE A TEST LIGHT, Voltmeters only here.
We can't stress this enough. The data systems and air bag systems in modern cars are very sensitive and can be damaged or set off with just static electricity. We can not be held responsible for what you are doing, we are simply trying to guide you. In my case I have years of car audio experience. I still run into things in new cars that I've never seen before and I've seen plenty of cars damaged by people cutting the wire that was the right color or "injecting voltage" to test something.
Moving forward:
Negative systems:
Typically a Negative system is one that the control wire will go negative from a resting position that is positive. In the case of most Honda's it rests at 5V and goes negative with each button. If you have a stock navigation system (In the case of the Honda. Others may be different) some buttons may be dedicated to the Nav Brain. That unit needs to be removed with the head unit to function properly and has special "reverse wiring" to provide all buttons functions.
Positive systems
Positive systems work just like negative systems but instead rest at or near ground (zero voltage). The voltage at the wire will go positive when a button is pressed.
Programming the Big Bang
Our site will cover most of this but at present the Big Bang is programmed via a PC. We do hope to one day be able to program it from the tablet, allowing one to tweak the controls after install and to improve installation experience. We will provide more detailed instructions closer to launch day but the short of it is you will download a program, modify it to fit your car/application then run it to program the Big Bang.
Connecting the Big Bang
We recommend using the Big Bang with a USB Hub that is powered. We have found some 12V to 5V dc to dc Converters we will link below as well as the USB hubs we recommend. The Big Bang will connect to the USB hub and the USB hub will plug into the tablet or phone. The big bang will have two wires that need to be connected. The Big Bang is powered from the USB power, so it and the tablet are always receiving power. It will have two wires to be connected. The white wire is the control wire and needs to be attached to the steering wheel control wire. The Red one is the ACC[essory] or IGN[ition] wire and needs to be tapped to it as a reference for the cars voltage. This wire should be fused as there is always a potential for shorting. It may be redundant but we still recommend fusing this wire independent of others, or using the same fuse used for the USB 12v to 5v converter.
Problems we've seen in our testing
First and worst problem I've experienced is cheap USB OTG cables. Some are very finicky about the position the sit in to make contact. It could be that they are old and worn out too. We plan to explain a "best methods" guide but best I can say is get decent quality USB OTG adapters for this use.
The second issue I've seen is random reboots. My n7 experiences these only when I'm using it in the car. I'm running Paranoid Android, and I don't believe it is the issue. I believe it is due to the faulty USB cable described above.
Other then that, we've experienced few issues but we are still in somewhat early testing phase.
I'll update this post with Q and A's as they come in. When asking questions about your installation it would be helpful for you to be as thorough as possible. Let us know if you are an experienced car audio person or not. Also make model and year of car may be important.
Until now this has been our idea for our car, we now open it to you. We can't promises anything but if you have suggestions, tell us. We may include the feature. We are working on other products as well to enable people to install their tablet and have it operate as normal as possible.
Tested Devices:
We've tested it on these devices so far but you can help. Get a USB OTG cable for your tablet, then plug in a USB keyboard or mouse and see if it functions. post your results and we'll add it to the list.
Confirmed working:
N10
N7
Galaxy Note2
Galaxy S3
Know to NOT work:
N4

Post #2
So we have run into a problem with the USB OnTheGo mode and charging. The Big Bang requires the tablet to be in USB OTG to receive control inputs. While in USB OTG mode the tablet is not charging. We have found there is another mode, Rid_a mode. By enabling this Rid_a mode the tablet will do both USB OTG and charge. We have it working on some Samsung's, we now need to figure out the nexus line.
It is possible that this is something coded into the kernel so we are looking into that. Right now the two default options, if we can't work it out, are:
A) A source built kernel with this one thing enabled.
B) a much more complicated setup that would auto switch between host and client mode from the ignition control. This we feel is less desirable based off the fact that if one is driving long distances the tablet could die out.

Post #3
Ok, I know I’m a bad person, I have left you hanging for ages. Well, it was not without feeling bad, and it was not on purpose, I’ve just been busy trying to pay bills.
So I suppose I’ll start with an overview. So our prototype is a DigiSpark. You can get one for like $9.99 or less. They are neat little devices. If your cool you can follow this little how to to use the Atiny85 http://youtu.be/30rPt802n1k, Which is where we were going with our own custom board. You might need to understand a little about voltage, voltage dividers and such. I will try to help, but I can’t climb in your car and measure the voltage or resistance for you so you’ll need to know how to use a DVOM and some pretty basic electronics.
Ok so probably the most common steering wheel interface type is a negative type, but there is other types. So for the standard type, I’ll try to diagram This is pure example.
Code:
(-gnd)----|---/\/\/\/\/\-----(s1)--------|
| 12k R1 |
|---/\/\/\/\/\-----(s2)--------|
| 8k R2 |
|---/\/\/\/\/\----(s3)---------|
| 5k R3 |
|---/\/\/\/\/\----(s4)---------|
| 2.4k R4 |
|---/\/\/\/\/\----(s5)---------|
| 1k R5 |
|---/\/\/\/\/\----(s6)---------|
450 R6 |
|---------/\/\/\/\/\---- (+12v (vcc) )
| 1k Rb
|
\ /
\/
MPU pin ?
So basically the way this works is on the steering wheel side the wire is grounded. In my honda there is actually both wire sides, the positive and negative or common resistor/common Switch sides. Since I pulled all of the stock Nav Electronics out, I had to ground the one side and put the 12v through the other. You can find a lot of this info out from PIE, PAC and peripheral websites (They make steering wheel interfaces and they should have some info in the installation manual PDF on what your car will be).
This may not be new to you, but for others: The Math is
R1+Rb
--------- * V (probably 12v, 14.4v running)
R1*Rb
So for each switch we’ll have a voltage, or more specifically a voltage range. I had a bit of a solution for the range part but for most applications having a range will be acceptable. In my case the range was +(-) 0.5v.
This should give you approximately something like S1= 11v to 12v, S2= 8v to 9v, etc
So the problem is we can only use a 0v to 5v range. So either we must adjust Rb or use another solution, such as an OP Amp to reduce the voltages. However it should be enough to adjust the resistance of Rb.
So we are going to use the A/D input pin of the ATtiny85 (digispark). We’ll convert that to a digital number and then compare it to some reference values. We’ll just consider the MSB Most significant bits and ignore the last few. This gives us some range to tolerate voltage fluctuations due to the cars electrical system.
That basically explains the car interface. We can do all kinds of neat stuff. We have a few useable pins so we can add pins for ACC, or door triggers, etc.
Now we need to consider the USB interface. The Digispark makes this easy. It is built on a board that has a USB interface built in. If you are using the ATtiny85 on a DIY you’ll need to make up a USB cable for this part. We’re going to make the steering wheel control look like a USB keyboard or Audio control to the tablet.
For the USB part I found a neat description on XDA in a reply on a post that inspired the basic workings. Understanding the USB ASCii code is important here. Android uses many standard ones. It’s a complicated code as it is meant to be able to grow and handle all kinds of different possible things.
The Main uses I found were in the Standard keyboard codes and in the Audio control codes.
We use the Digispark hex program that comes in the digispark start up guide, keyboard.h . We pass the code calls to keyboard.h and it does the rest, except we need to add our specific functions to keyboard.h.
We’ll modify keyboard.h and write our sketch and program the Arduino/Digispark/ATtiny85. It will loop through reading the A/D and comparing to the pre-determined reference values. When it matches we’ll output a USB code to do something on the tablet.
This is where the problem came in. Of all things I didn’t expect to have a issue charging while acting as a usb host to, effectively, a keyboard. As of USB standard 1.1 using a certain resistance between the two data pins will put it in different modes. Unfortunately what I found was on some Samsung devices this value put it in a different mode, and yet a mode that gave us USB host, didn’t allow for charging. On the Nexus devices I couldn’t find a way at all, without requiring a different kernel to be installed. This left two main choices. A) charge when the car is off only. Which is unacceptable in my opinion. A long drive would kill even a Nexus7 and then you’re stuck with nothing. B) Require an aftermarket kernel/ROM installed. Well we realized guiding one to remove the deck and do some major work to put a tablet in the dash ment they were serious, it didn’t seem fair to expect them to also root/hack the OS on said device. Sure I suppose we could have made it such but it didn’t fit our goal. There is a third option I just now thought of. One could add a circuit that triggered the Host mode change only when the buttons are pressed. This can be as simple as causing a closed/open connection between data+ and data- . This could cause some missed keystrokes but if planned right, it might be a viable solution.
Sketch Code:
Code:
#include "DigiKeyboard.h"
/*
Basic Steering wheel to Tablet (or any computer) controls. Posted here are the basics from a honda ridgeline '07. The steering wheel wire is Green with red stripe, pin 3
The Brown wire, pin 11, Needs to be connected to ground (the buttons are resisters accross those two). This is only Vol+,Vol-, Mode, Ch+, Ch-. I will search for Back and Talk later.
The Resistance Values read from it was: Rest=9.9k, Mode=3.73k, Ch+= 1.69k, Ch-= 780, Vol+=356.8, Vol-=101.4. Using a 1k pull up to the 5V USB lead seems like a good value and will yeild
a large variance in voltage output. The outputs should be Rest=4.5454v, Mode=3.94v, Ch+=3.146v, Ch-=2.19v, Vol+=1.314v, Vol-=.4603v. These can be converted to a 10bit value and are about
Rest=931.4, Mode=807.4, Ch+=644.6, Ch-=448.8, Vol+=269.26, Vol-=94.3. This variance is more then 100 dec thus to allow a wide tolarance we will drop the last four bits and
test only the four MSB. I.E. if analogRead(0)==11010000 then Ch+. The LSB will need to be striped away due to them not being exactly the same otherwise.
Grey wire, at steering coulumb is the Talk and Back button. It works with The Same brown wire as with the volume. 9.99k at rest, Talk= 2.238K, Back= .64k. Used with a 1K resister, They are
Rest=4.5454v, Talk=3.455v, Back=1.95v.
## DISCOVERY, analogRead is a DEC value and is difficult to convert. was going to read value then roll value right and then compare. This would have given a range:
// audioState = audioState >> 7;
// navState = navState >> 7;
##Delay output can act as a Remote turn on for Amplifire.
*/
/* Rest= 931 = 1110100011 = 0x3A3
// Mode=807 = 1100100111 = 0x327
// Ch+=644 = 1010000100 = 0x284
// Ch-=449 = 0111000001 = 0x1C1
// Vol+=269 = 0100001101 = 0x10D
// Vol-=94 = 0001011110 = 0x5E
//
// Rest= 931 = 1110100011 = 0x3A3
// Talk= 708 = 1011000100 = 0x2C4
// Back= 399.6= 0110010000 = 0x190
*/
//const int audioCtrl = 5; //pin 5, analog Read 0
//const int navCtrl = 2; //pin 2, analog Read 1
const int ledPin = 1;
int audioState = 1023; //B1110100011; //Variable to store button Value
int navState = 1023; //B1110100011; //Variable to store button Value
int twiceSel = 0; //Variable for checking for double press of vol- and of back
int storeState = 0; //Store the current state for double press check.
void setup() {
pinMode(ledPin, OUTPUT); // Turn on onboard LED when button is pressed
// pinMode(audioCtrl, INPUT); //P0 is volume up
// digitalWrite(volUp, HIGH); //set the pull up resister
// pinMode(navCtrl, INPUT); //P2 is Volume down
// digitalWrite(volDown, HIGH); //set the pull up resister
}
void loop() {
DigiKeyboard.update();
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
audioState = analogRead(1); //read value at audio control pin2, write it to audioState for further processing
navState = analogRead(0); //Read the value of navagation Control pin5, write it to navagationState for further processing.
// Rest= 931 = 1110100011 = 0x3A3 = 111
// Mode=807 = 1100100111 = 0x327 = 110 >675 <850
// Ch+=644 = 1010000100 = 0x284 = 101 >550 <675
// Ch-=449 = 0111000001 = 0x1C1 = 011 >300 < 500
// Vol+=269 = 0100001101 = 0x10D = 010 >120 < 300
// Vol-=94 = 0001011110 = 0x5E = 000 >50 <110
//
// Rest= 931 = 1110100011 = 0x3A3 = 111
// Talk= 708 = 1011000100 = 0x2C4 = 101 >650 <750
// Back= 399.6= 0110010000 = 0x190 = 011 >350 <450
// TODO:
//Need debounce, or slowdown, or delay. Currently it cycles too fast.
// Add double press Vol- triggers mute
// Add double press Back triggers Home
//Fix the "/" causing search to display "//
// Use Dpad instead of arrow (probably a function of the Keypad arrows)
if (audioState >675 && audioState < 850 ) { //Mode inbetween 675 and 850
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_ENTER); //tell computer Select
digitalWrite(ledPin,LOW);
delay (200); //200ms Delay
}
else if (audioState >550 && audioState < 675) { //Ch+ >550 < 675
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_ARROW_RIGHT); //tell computer Right Arrow
digitalWrite(ledPin,LOW);
delay (200); //200ms Delay
}
else if (audioState >300 && audioState < 500) { //Ch- >300 < 500
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_ARROW_LEFT); //tell computer Left Arrow
digitalWrite(ledPin,LOW);
delay (200); //200ms Delay
}
else if (audioState >120 && audioState < 300) { //Vol+ >120 <300
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_VOL_UP); //tell computer Volume Up
digitalWrite(ledPin,LOW);
delay (200); //200ms Delay
}
else if (audioState >50 && audioState < 110) { //Vol- >50 < 110
delay (200);
for(int x =0; x < 5; x++) {
audioState = analogRead(1); //read button again
if(audioState >50 && audioState < 110){
digitalWrite(ledPin, HIGH);
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_VOL_DOWN); //tell computer Select
digitalWrite(ledPin,LOW);
//if statement test for 5cycle if button pressed again, if not changes vol- if so mutes
}
else if (navState >675 && navState < 850) { //Talk >650<750
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_SEARCH); //tell computer Select
digitalWrite(ledPin,LOW);
delay (200); //200ms Delay
}
else if (navState >350 && navState < 450) { //Back >350 <450
//if statement test for 5cycle if button pressed again, if not back, if so home
// storeState = 0; //Clear storeage
// twiceSel = 0;
// storeState = navState; // back was pressed, we need to store it's current state.
// navState = 0;
delay (200);
for (int x=0; x < 5; x++){ // Test 5 times
navState = analogRead(0); //Read the value of navagation Control pin5, write it to navagationState for further processing
if (navState >350 && navState < 450) { //Back >350 <450
// twiceSel = 1; //pressed twice
//Something different
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_HOME); //tell computer Select
digitalWrite(ledPin,LOW);
}
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_ESC); //tell computer Select
digitalWrite(ledPin,LOW);
}
}
}
/*
OLD code left for Refrence
buttonState = digitalRead(volUp);
if (buttonState == LOW) {
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_VOL_UP); //tell computer to turn the volume down
}
else{
digitalWrite(ledPin, LOW);
}
buttonState = digitalRead(volDown);
if (buttonState == LOW) {
digitalWrite(ledPin, HIGH);
DigiKeyboard.sendKeyStroke(0); //this is generally not necessary but with some older systems it seems to prevent missing the first character after a delay
DigiKeyboard.sendKeyStroke(KEY_VOL_DOWN); //tell computer to turn the volume down
}
else{
digitalWrite(ledPin, LOW);
}
*/
Keyboard.h Code:
Code:
/*
* Based on Obdev's AVRUSB code and under the same license.
*
* TODO: Make a proper file header. :)
* Modified for Digispark by Digistump
*/
#ifndef __DigiKeyboard_h__
#define __DigiKeyboard_h__
#include <avr/pgmspace.h>
#include <avr/interrupt.h>
#include <avr/delay.h>
#include <string.h>
#include "usbdrv.h"
// TODO: Work around Arduino 12 issues better.
//#include <WConstants.h>
//#undef int()
typedef uint8_t byte;
#define BUFFER_SIZE 2 // Minimum of 2: 1 for modifiers + 1 for keystroke
static uchar idleRate; // in 4 ms units
/* We use a simplifed keyboard report descriptor which does not support the
* boot protocol. We don't allow setting status LEDs and but we do allow
* simultaneous key presses.
* The report descriptor has been created with usb.org's "HID Descriptor Tool"
* which can be downloaded from http://www.usb.org/developers/hidpage/.
* Redundant entries (such as LOGICAL_MINIMUM and USAGE_PAGE) have been omitted
* for the second INPUT item.
*/
PROGMEM char usbHidReportDescriptor[39] = { //USB_CFG_HID_REPORT_DESCRIPTOR_LENGTH] = { /* USB report descriptor */
0x05, 0x01, // 0x05, 0x01 USAGE_PAGE (Generic Desktop)
0x09, 0x01, // 0x09, 0x06 USAGE (Keyboard)
0xa1, 0x01, // 0xa1, 0x01 COLLECTION (Application)
0x25, 0x01,
0x15, 0x00,
0x75, 0x01,
0x95, 0x05, // REPORT_COUNT (5)
0x09, 0xb5, // USAGE (Scan Next Track)
0x09, 0xb6, // USAGE (Scan Previous Track)
0x09, 0xb7, // USAGE (Stop)
0x09, 0xcd, // USAGE (Play/Pause)
0x09, 0xe2, // USAGE (Mute)
0x81, 0x06, // INPUT (Data,Var,Rel) - relative inputs
// -------------------- volume up/down bits
0x95, 0x02, // REPORT_COUNT (2)
0x09, 0xe9, // USAGE (Volume Up)
0x09, 0xea, // USAGE (Volume Down)
0x81, 0x02, // INPUT (Data,Var,Abs) - absolute inputs
// -------------------- padding bit
0x95, 0x01, // REPORT_COUNT (1)
0x81, 0x01, // INPUT (Cnst,Ary,Abs)
//0x05, 0x07, // 0x05, 0x07 USAGE_PAGE (Keyboard) change to 0x0c for consumer device.
//0x19, 0xe0, // 0x19, 0xe0 USAGE_MINIMUM (Keyboard LeftControl)
//0x29, 0xe7, // 0x29, 0xe7 USAGE_MAXIMUM (Keyboard Right GUI)
//0x15, 0x00, // 0x15, 0x00 LOGICAL_MINIMUM (0)
//0x25, 0x01, // 0x25, 0x01 LOGICAL_MAXIMUM (1)
//0x75, 0x01, // 0x75, 0x01 REPORT_SIZE (1)
//0x95, 0x08, // 0x95, 0x08 REPORT_COUNT (8)
//0x81, 0x02, // 0x81, 0x02 INPUT (Data,Var,Abs)
//0x95, 0x01, // 0x95, 0x01 REPORT_COUNT (simultaneous keystrokes)
//0x75, 0x08, // 0x75, 0x08 REPORT_SIZE (8)
//0x25, 0xE7, // 0x25, 0xE7 LOGICAL_MAXIMUM (130) 0x65 originally
//0x19, 0x00, // 0x19, 0x00 USAGE_MINIMUM (Reserved (no event indicated))
//0x29, 0xE7, // 0x29, 0xE7 USAGE_MAXIMUM (Keyboard Application) 0x65 originally
//0x81, 0x00, // 0x81, 0x00 INPUT (Data,Ary,Abs)
0xc0 // 0xc0 END_COLLECTION
};
/* Keyboard usage values, see usb.org's HID-usage-tables document, chapter
* 10 Keyboard/Keypad Page for more codes.
*/
/*
#define MOD_CONTROL_LEFT (1<<0)
#define MOD_SHIFT_LEFT (1<<1)
#define MOD_ALT_LEFT (1<<2)
#define MOD_GUI_LEFT (1<<3)
#define MOD_CONTROL_RIGHT (1<<4)
#define MOD_SHIFT_RIGHT (1<<5)
#define MOD_ALT_RIGHT (1<<6)
#define MOD_GUI_RIGHT (1<<7)
//#define KEY_ARROW_UP 0x52 //Added 2-5-2013 JWA
//#define KEY_ARROW_DOWN 0x51 //Added 2-5-2013 JWA
*/
//Tested with Logitech keyboard on N7. Right Meta key (MOD_GUI_RIGHT) plus F1 = home. Right Meta plus F3= search. Home = Internet home, Search = Windows search Application specific according to Logitech page.
#define KEY_HOME 157 //Added 2-7-2013 JWA according to android doc 3, According to other doc 165
#define KEY_MUTE 0xE9 //Added 2-7-2013 JWA keypad 3 and Page down according to android doc 91/127
#define KEY_SEARCH 127 //Added 2-6-2013 JWA according to android doc 84
#define KEY_ESC 41 //Added 2-6-2013 JWA KEboard ID ESC = 41
// /system/usr/keylayout/geniric.kl
// 172=Home
// 217= Search - 127
// 213== Music
// 140== Calculator
// 158 = Back - 1
// 102=power Confirmed hid usage ID
// 101=menu confirmed
#define KEY_ARROW_LEFT 0x50 // Android dock d-pad left = 105
#define KEY_ARROW_RIGHT 0x4F //Added 2-5-2013 JWA Andorid dock D-pad right=106
#define KEY_ENTER 40 //mode long press could do recent apps key 187
#define KEY_VOL_UP 128 //Added 2-5-2013 JWA 0x80 Andoid dock volume up=115 vdown= 114
#define KEY_VOL_DOWN 129//Added 2-5-2013 JWA 0x81
class DigiKeyboardDevice {
public:
DigiKeyboardDevice () {
TIMSK &= !(1<TOIE0);
cli();
usbDeviceDisconnect();
_delay_ms(250);
usbDeviceConnect();
usbInit();
sei();
// TODO: Remove the next two lines once we fix
// missing first keystroke bug properly.
memset(reportBuffer, 0, sizeof(reportBuffer));
usbSetInterrupt(reportBuffer, sizeof(reportBuffer));
}
void update() {
usbPoll();
}
void sendKeyStroke(byte keyStroke) {
sendKeyStroke(keyStroke, 0);
}
void sendKeyStroke(byte keyStroke, byte modifiers) {
while (!usbInterruptIsReady()) {
// Note: We wait until we can send keystroke
// so we know the previous keystroke was
// sent.
usbPoll();
_delay_ms(5);
}
memset(reportBuffer, 0, sizeof(reportBuffer));
reportBuffer[0] = modifiers;
reportBuffer[1] = keyStroke;
usbSetInterrupt(reportBuffer, sizeof(reportBuffer));
while (!usbInterruptIsReady()) {
// Note: We wait until we can send keystroke
// so we know the previous keystroke was
// sent.
usbPoll();
_delay_ms(5);
}
// This stops endlessly repeating keystrokes:
memset(reportBuffer, 0, sizeof(reportBuffer));
usbSetInterrupt(reportBuffer, sizeof(reportBuffer));
}
//private: TODO: Make friend?
uchar reportBuffer[2]; // buffer for HID reports [ 1 modifier byte + (len-1) key strokes]
};
DigiKeyboardDevice DigiKeyboard = DigiKeyboardDevice();
#ifdef __cplusplus
extern "C"{
#endif
// USB_PUBLIC uchar usbFunctionSetup
uchar usbFunctionSetup(uchar data[8])
{
usbRequest_t *rq = (usbRequest_t *)((void *)data);
usbMsgPtr = DigiKeyboard.reportBuffer; //
if((rq->bmRequestType & USBRQ_TYPE_MASK) == USBRQ_TYPE_CLASS){
/* class request type */
if(rq->bRequest == USBRQ_HID_GET_REPORT){
/* wValue: ReportType (highbyte), ReportID (lowbyte) */
/* we only have one report type, so don't look at wValue */
// TODO: Ensure it's okay not to return anything here?
return 0;
}else if(rq->bRequest == USBRQ_HID_GET_IDLE){
// usbMsgPtr = &idleRate;
// return 1;
return 0;
}else if(rq->bRequest == USBRQ_HID_SET_IDLE){
idleRate = rq->wValue.bytes[1];
}
}else{
/* no vendor specific requests implemented */
}
return 0;
}
#ifdef __cplusplus
} // extern "C"
#endif
#endif // __DigiKeyboard_h__
I had planned and some idea of how to implement things like double press vs long press vs single short press all without lots of delay. It had to do with using changes in the AD values to trigger events and use timers to allow other things to happen in the meantime. Also things such as two buttons pressed at the same time could potentially trigger an event.
Unfortunately I’ve abandoned this whole thing. Firstly I have a background in Car audio, so unlike many of you, I am used to a gaping hole in my dash. My stereo is in a state of flux while I come up with exactly what I want to do. Right now I enjoy just popping my Note2 on the dash and using it for my nav/music needs it allows me to keep my tablet in the house. I’m not saying the idea isn’t still great, nor that I don’t love it. If I was in Car Audio still, I would be using Nexus7’s in place of Alpine car stereos all day long.
Links:
http://www.Digistump.com buy your digispark here, Also the source of much of the code, software and info for making things work.
http://www.arduino.cc/ you’ll need the software, bare minimum. The Digispark page may walk you through installing it.
http://www.atmel.com/Images/Atmel-2...ller-ATtiny25-ATtiny45-ATtiny85_Datasheet.pdf The chip used in this project is based on the ATtiny85. One can use an Arduino if they choose.
https://github.com/digistump/Digisp...er/libraries/DigisparkKeyboard/DigiKeyboard.h Bluebie’s Keyboard.h library.
https://github.com/Bluebie/micronucleus-t85 Bluebie’s bootloader which provides the USB funcitonality
http://www.obdev.at/products/vusb/index.html More info on using an AVR as a USB device
http://www.microchip.com/forums/m618147.aspx Use of a PIC microchip as a USB device
http://www.microchip.com/forums/m440956-print.aspx HID Audio Controler Usage table
http://www.freebsddiary.org/APC/usb_hid_usages.php USB Useage tables.
http://www.usb.org/developers/devclass_docs/Hut1_11.pdf More USB useage tables
http://forum.xda-developers.com/showthread.php?p=36930621 Tunner app for USB FM radio
https://droidperception.wordpress.com/2013/01/13/nexus-7-webcam-attached-update/ Info on using Extrernal USB Webcams
http://forum.xda-developers.com/showthread.php?t=2113259 A great thread on what can be done with a tablet in the car. He beat me to the install by just a month or two. He also was the reason I realized the limitation on the Charging while acting as a host.
Much of the code above is not my work. I made changes to it where I needed to but, specifically in the keybaord.h almost all of it is not mine. Credit goes to Bluebie and/or Digispark for the original work. If they would like me to remove the code, Please ask, I will. However since it's posted on Github I believe it's safe to post here.

I'm working on a way to adapted steering wheel controls on my mustang. But the issue is I have a base with no controls. I want swap out a premuim steering wheel to get this, but I have to have fords flash the acm for it to work. The issue is I still wont have controls for the tablet since mt car does not have bluetooth. So it will become costly. My solution is to use arduino or some other way to communicate with the factory HU. So im looking forward to this as a viable option.

Just an FYI, CM10-based kernel allows USB OTG and charging at the same time.
I've been on this hunt for a couple months now, albeit not hardcore. I have a rooted Galaxy Tab 2 7.0 that I'll be installing into an Infiniti G35. I would be more than happy to test stuff out for you if you want another guinea pig.

Pics to come
I'll post some pics from my installation so you can get an idea of what I did. Hopefully that and the code, etc will help you figure out how to do yours.

So hopefully Posting this will help some people get their own going. Here are the photos from the unit in my car.
My car has two sets of controls, first is the volume and channel buttons, second is the talk and navigate buttons. Also because I removed the complete Nav system the whole circuit was open. This gave me more options on how to do it, but I wen't with the more conventional. The Orange wire is going to ground. The Red and Yellow are each set of controls. The 1K resistors is a pull up resistor sourced from Vin on the digispark.

Hey joeavery2
thanks a lot for the hard work
If I am understanding the connections correctly, the tablet will draw the charge power from the USB Hub correct? If this is the case, is the tablet limited to 500mA charge current?

Hi guys i would ask your help. I have an Alfa Romeo Giulietta, Installed in the dash a nexus 7 2013 and I want use the steering wheel buttons. The car use the protocol is 15765-4 29bit and I have on Bluetooth obd reader. Can you please help me out? Thanks in advance.
Inviato dal mio FRD-L09 utilizzando Tapatalk