Rockbox

Tasklist

FS#7464 - Small weaknesses in FM-radio on iAudio X5

Attached to Project: Rockbox
Opened by Davy Kager (DavyKager) - Saturday, 21 July 2007, 18:26 GMT
Task Type Bugs
Category FM Tuner
Status Unconfirmed
Assigned To No-one
Operating System iAudio X5
Severity Low
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 0%
Votes 0
Private No

Details

I'm using the newest build of Rockbox (2007-07-21) with an iAudio X5 20 GB and I have my tuner set to the European region. I encounter two little flaws in the FM-radio. I think they partially have to do with the player's hardware, but maybe Rockbox can improve the quality a bit.

1. The player is really sensitive to movement. The signal strength differs with every meter that you take the player with you. On top of that, most stations sound like they're not correctly tuned, but I can't get them any better. Maybe Rockbox could be more accurate in frequency selection? (Like 101.100, 101.125, 101.150.) I can get these stations quite well on my home radio. I don't expect the same quality from such a small device, but some finetuning would be great. The only problem is that it would take a while to scroll through the frequencies, but once set it'd be fine.
2. (I assume this is very hardware related.) The radio signal may change between stereo and mono. I have it set to stereo, but sometimes it just picks mono. I guess this is because of the hardware, since it is place related. I also didn't find any potential bugs in the source code.

What are your thoughts on this?

Oh, and keep up this good work! I really enjoy Rockbox!
This task depends upon

Comment by Michael Sevakis (MikeS) - Saturday, 21 July 2007, 21:17 GMT
Correct tuning for the TEA5767 chip isn't quite as simple as it is in the current rockbox driver. The correct high/lo injection must be picked. There's really nothing to do with fine tuning since that is done with the internal PLL. I do have code that improves things enormously for scanning and tuning in US/Canada but the scanning method for that is pretty much useless in the European band. I do have scan data collected on how to make the detection work correctly there. The driver in SVN is better suited to European scanning.

I know you didn't ask about scanning in particular but WTH, I'll just give the lowdown on all of it. Radio drivers have been made region aware since my first attempt at addressing this so it seems a better time to look into it again soon.
Comment by Davy Kager (DavyKager) - Saturday, 21 July 2007, 21:39 GMT
Okay... I saw some region code in tuner.h, and the European region was almost the same as the US/Canada version. I'm absolutely not into this though.

So is there just no way to go to a frequency of 101.1375, or is it just useless to do? I didn't do a test with the original firmware, but now I really wonder if they had a "magic solution" for the chip. :)

By the way, I don't use the scanning feature much. For me it's just preseetting.
Well, and the rest of Rockbox just rocks!
Comment by Davy Kager (DavyKager) - Sunday, 22 July 2007, 10:53 GMT
Extra info: I tried the American region setting and I actually got some reception. Not as good as in European mode, but I could live with it. So an improved American region might work here as well.
Comment by Rani Hod (RaeNye) - Sunday, 05 August 2007, 13:12 GMT
>So is there just no way to go to a frequency of 101.1375, or is it just useless to do?
IIRC the TEA5767 (or actually the compatible D&A202 present in X5) allows for this resolution, but why do you want it?
(Does this FM station use a frequency which is not a multiple of 0.1, or even 0.05, MHz?)

>I have it set to stereo, but sometimes it just picks mono
Normal behaviour for FM tuner when reception is poor.
Comment by Davy Kager (DavyKager) - Sunday, 05 August 2007, 13:27 GMT
> IIRC the TEA5767 (or actually the compatible D&A202 present in X5) allows for this resolution, but why do you want it?
(Does this FM station use a frequency which is not a multiple of 0.1, or even 0.05, MHz?)

I thought it might improve the reception (just a suggestion). Otherwise an improved driver for the tuner will be greatly appreciated!
By the way, it has also something to do with the bad reception here (the Netherlands). I tried the device in Finland and it sounded better, but still not perfect. :)
Comment by Rani Hod (RaeNye) - Sunday, 05 August 2007, 13:43 GMT
I don't know much about RF, but I don't think that deliberate out-of-tuning could improve reception...
This whole high/low-side injection (whatever this means ;-) sounds like a better option.

To improve reception a little, note that
1. the headphones cord serves as the antenna;
2. using the remote degrades reception.
Comment by Michael Sevakis (MikeS) - Sunday, 05 August 2007, 14:43 GMT
I guess I should consider committing the reception aspect of my TEA work and leave the scanning part as a separate project since the in-SVN scanning works somewhat well in the US band anyway.
Comment by Rani Hod (RaeNye) - Saturday, 13 October 2007, 13:42 GMT
Michael, could you please commit or at least upload here the TEA5767 reception patch? 10x.
Comment by Davy Kager (DavyKager) - Saturday, 03 November 2007, 15:50 GMT
Yeah, a commit of this kind would be very useful! Has there actually been any improvement last months?

Loading...