For Guino: USB Radio FM with RDS from Silabs

**FMode** · 10-15-2006, 04:30 AM

Why do you "invest" time in the silabs code ?
I have invested some time to write a RDS Interpreter
Radio Drivers Round Table

Join with Evo

And give me your email that I can send you also my RDS Interpreter stuff.

RDSRAWMessage.Block0 = registers[0xC]
RDSRAWMessage.Block1 = registers[0xD]
RDSRAWMessage.Block2 = registers[0xE]
RDSRAWMessage.Block3 = registers[0xF]
...
RDSInterpreter.decode(RDSRAWMessage)

and you are finished...
(You can now inspect RDSInterpreter)

Originally Posted by smbgaiden

...
Block B (Block 2):
._.._.._.._.._.._.._.._.._.._.._.._.._.._.._.._.
0 1 2 3 4 5 6 7 8 9 A B C D E F
|Group Type|| || |-----PTY---|| |--Addr-|
.0..0..1..0..1.| |TP / \
/ \ A/B
B0

Etc.
------------------------------------------------------------------
Source Code:
Where:
registers[0xB] == RDS Read Chan
registers[0xC] == Block A (1)
registers[0xD] == Block B (2)
registers[0xE] == Block C (3)
registers[0xF] == Block D (4)

case RDS_RTO_2A:
// RDS Block B == 0xD
addr = (registers[0xd] & 0xf) * 4;
// RDS Read Chan == 0xB
abflag = (registers[0xb] & 0x0010) >> 4;
// RDS Block C == 0xE
update_rt(abflag, 4, addr, (BYTE*)&(registers[0xe]), errorFlags);

Notice the 5th bit of the Read Chan register is used rather than the 5th bit of the Block B register for the A/B Flag? I tried to update it to use the Block B register 0xD and the text failed to decode.
------------------------------------------------------------------

the 5th bit must be masked out with 16 and shifted 4 times to right...

Originally Posted by smbgaiden

Using the above information, is it safe to say the chip or firmware is incorrectly updating the Readchan when it means to reference Block B? Is this fixable in the firmware? I also noticed other areas that gave a strange result when I tried to interpret the RDS specification literally.

I live in USA. Might this be due to the case of the USA RDS format being different than the actual RDS spec I hold? You'd think that they'd keep consistency in the block, especially with regards to the fact that READ CHAN is a chip specific register and not mapped directly to something given off the RDS Data stream over the air.

I have a flag for RBDS signalling special PI Codes used in
This affects only the return of the ProgrammTypeEnum
(i can't find more differences in rbds1998.pdf, ...)

Certainly I can't test this

AFN (American Forces Network - one Radio Station here) is not using it

**smbgaiden** · 10-15-2006, 02:00 PM

kpjuk: I had emailed silabs last night. I went through registering on their site to check their known issues and found they didn't have any app notes or documentation for supporting the chip/reference aside from those on the plain old extranet. I did notice they produced a more modern version of the chip which has a smaller die, but no code specific to it that I could examine to see if they cured the problem.

Mr. FMode,

(note, some clipped text)

Originally Posted by FMode

Why do you "invest" time in the silabs code ?
I have invested some time to write a RDS Interpreter
Radio Drivers Round Table

Join with Evo

And give me your email that I can send you also my RDS Interpreter stuff.

the 5th bit must be masked out with 16 and shifted 4 times to right...

I have a flag for RBDS signalling special PI Codes used in
This affects only the return of the ProgrammTypeEnum
(i can't find more differences in rbds1998.pdf, ...)

Certainly I can't test this

AFN (American Forces Network - one Radio Station here) is not using it

1. My email is cactus_stew at yahoo. I appreciate your kindness to offer the RDS Interpreter data.
2. I am aware of the operation of masking it with the 16th bit and shifting right 4, but the issue was with the fact that it is using a non-RDS register. Note the registers[0xb], where registers[0xd] is defined as RDS Reg B and registers[0xb] is defined as READCHAN. I am curious if your code will parse the data on the firmware I have due to your sample code looks directly at RDS registers only and the silabs product seemingly uses undocumented data elsewhere.
3. I Will be happy to test anything related to the special PI codes and post/email you results.

Best Regards,

**smbgaiden** · 10-15-2006, 02:02 PM

Originally Posted by kpjuk

This is interesting, and worthy of a note back to scilabs. It could be a simple change on their side. I am interested in knowing what the reception is like for these things. I still have a Head unit in the car, but would love to be able to take it out and have Road Runner do the whole thing (once I get rid of my ground loop problem).

keep posting, I can't help with code unfortunately but you seem to be making good progress.

cheers
KPJUK

I am not mobile with it, so I can't comment on the reception yet. Further, my computer is set up very near to a sliding glass door, so has a good open sky access. I'll let you know my results when I get the unit placed a little obstructed and within a moving vehicle.

**FMode** · 10-18-2006, 04:47 PM

Originally Posted by smbgaiden

Mr. FMode,

(note, some clipped text)

1. My email is cactus_stew at yahoo. I appreciate your kindness to offer the RDS Interpreter data.
2. I am aware of the operation of masking it with the 16th bit and shifting right 4, but the issue was with the fact that it is using a non-RDS register. Note the registers[0xb], where registers[0xd] is defined as RDS Reg B and registers[0xb] is defined as READCHAN. I am curious if your code will parse the data on the firmware I have due to your sample code looks directly at RDS registers only and the silabs product seemingly uses undocumented data elsewhere.
3. I Will be happy to test anything related to the special PI codes and post/email you results.

Best Regards,

email on its way....

sorry you need BockB to get that Flag...
Do you have the Ne50068.pdf ?

Code:

abflag = (registers[0xd] & 0x0016) >> 4;

should work

**jrtf83** · 10-20-2006, 04:45 AM

dying to hear the progress on this really cool project. got to the point where i was going to replace my stock stereo with a good aftermarket system and then realized i don't want to pay big bux for the hqct. Keep up the good work guys!

**ramaloji** · 11-05-2006, 03:41 AM

Have And Use Tu Silabs Test

I Can Assure That With The Radiator It Works
Single To Copy The Dll In Radiator Plugins

Later To Use The Radiator In The Rr And Works Several Things But Not Everything

Although Thus I Have Radio In The Rr

It Is Difficult To Add To A Keyboard For Direct Frequency And 10 Memory Button

To Be An Interesting And Comfortable DetaiL

MAP MONKEY DETECTS DATA TMC BUT IT SAYS NOT TO HAVE THE PLUGING

I Feel My English Is Google Translation

**erie** · 11-07-2006, 01:04 AM

Well I got my unit from www.mouser.com and it was shipped and arrived on time. I was able to get it installed without any problems, it was detected by XPsp2 as an "USB Audio Device" and was ready to use.

I tried the SiLabs test program for awhile and found that it is really sensitive to the attenna wire position, but after stringing out the wire I got a pretty good signal and RDS data started to flow in, and it was even formatted correctly! ( I’m in the US and an earlier post mentioned that the US might use a diff spec.)

In the house:
I did notice two things. First, the radio is about 1 sec behind an analog radio. Second, I swear sometimes the radio "skips", you really have to pay attention but I swear it does.

In the car:
Again this radio is very sensitive to then antenna wire position, and so far is not usable for the CarPC environment. I could not get a reliable signal unless the unit was out the window and the wire need to be stretched out.

Maybe something:
I took the cover off the unit, I was guessing that there might be an other antenna point on the PCB. Running my fingers over the board I get a noticeable difference when I touch the three holes stamped C12, R18 and C17.

Can anyone take a look at the PDF's posted on SiLabs site and tell us if there is another point we can use for the antenna connection?

**rando** · 11-17-2006, 02:44 PM

Originally Posted by erie

Can anyone take a look at the PDF's posted on SiLabs site and tell us if there is another point we can use for the antenna connection?

The only antenna input is the one where the oem aerial is attached. It connects into the tuner chip via a couple of filtering passives and shunt diodes to block any nasty rogue voltages at the antenna input. I played with bypassing the filtering (not a good idea if you want your radio to last long) but didn't attain any appreciable improvement in tuning performance.

**erie** · 11-17-2006, 02:46 PM

Well that sucks. So my verdict on this USB Radio is . .Dont buy it for the car. It works good for a desktop radio but it is unable to handle the automobile enviroment due to its poor reception.

**Iceman_jkh** · 12-08-2006, 03:35 AM

Is there a latest version of this software/plugin for RoadRunner?

I have the Silabs FM-USB and it works great in my car.
Read BJ's and my own account here: USB FM RDS Solution with sofware.