Rockbox Development > New Ports
SanDisk Sansa c200v2, m200v4, clipv1, clipv2, clip+, and fuzev2
OldOne:
Funman,
Thanks for quick reply, I thought I would pass on my experience with recent daily builds.
Properly means to allow view, access, and playing directories and files.
With the release of 091012 I can can select Files | < microsd> | and see a listing of the directories and files and select any directory or file. With later releases, ( I haven't tried them all), Main menu | Files | <microsd> shows only the First directory( no other directory or files in the root on microsd> on the disk and if I select it, Rockbox locks and requires a hard( 5 second or more) turnoff.
I think it was somewhere around 091015 or so when I had problems and 091012 was the last working one I had.
Browsing the main disk seems to work fine on any release. I have formatted the main and microsd twice. Everything works great right now with the 091012 release.
funman:
The voltage change was committed between the 12nd and the 16th so that could be a reason.
FlynDice what do you think ?
FlynDice:
I'm thinking voltage scaling. I did remove the BUSWIDTH and BLOCKLEN commands on the 12th but besides that voltage scaling is the only significant change during that period. Since this only seems to affect the uSD and not the internal SD I really suspect it's voltage scaling.
OldOne, are you able to build your own from SVN? It sounds as if you've just been using the daily builds.
EDIT: After thinking about this for a bit I'm slightly puzzled now as the voltage should be raised to normal(1.20v) during disk accesses now that we boost on disk access.
OldOne, can you go into System/Debug/CPU frequency and scroll counterclockwise. The frequency should go from 62000000 to 248000000 and then go back and try to browse your uSD files. We're manually boosting both the frequency and voltage by doing this.
OldOne:
FlynDice
I have tried several microSd cards with the latest release and some work fine.
The Adata 8 gb card that occaisionally would panic and would not read is working fine with the new release . A Transcend 2 gb card also seems to work ok.
The PQI 8gb cards that seemed better on earlier releases don't work any better with the debug procedure you asked me to try.
The AData card failed on one of the releases about 091021 but is working now 091027.
I have the most music on the PQi cards. So the problem seems to be only on PQI cards with the current release and boosting does not seem work. They work fine on 091012.
Thanks for the reply
Edit:
After some more testing with the newest release and 091012:
Going to Debug | Disk info shows 25 mhz for microsd0 in both releases
info for microsd1 shows 25 mhz in the 091012 release but 50 mhz in the 091027 release. Some microsd cards must not like the added clock speed.
I haven't setup to build my own releases but would like to try a new build with voltage scaling but with 25 mhz speed for sd1.
So for now I am running the 091012 release. When there is an update to daily build that changes the speed, I will try it. Are the patches and bugs the best place to monitor changes in daily builds?
FlynDice:
@OldOne:
The cards are all running overclocked right now at PCLK frequency(62 MHz). The internals are v1 cards that should use a maximum of 25 MHz, but they don't really give us problems. If the uSD card is a v2 card we assume it is high-speed capable and switch it to HS timings. HS timings are good from 0 - 50 MHz. Unfortunately we cannot get either card to function at a different frequency than PCLK right now and are searching for a reason why. There is no difference in the clock speeds for the two builds you have mentioned. In both builds you can look on the View HW debug page and see the card frequencies. It needs a disk access to get a reading so just insert/eject a card and you should see 62 MHz. The speed value for the card is not a MHz value but instead a Mbit/sec value for the transfer speed on 1 data line (which is all we can use right now also although there are 4 available). On the 12th I rearranged the HS timing code to be run before the code which checks on the card data so the disk info display would show us 50 MBits/s if we indeed did get a switch to HS timings.
If you can browse the files on your card I'm thinking the problem happens sometime after the init phase and is more likely to be related to when the disk is transferring data.
I think you said you already tried chkdisk.
I've run into some problems if i've ejected the card while the OF was refreshing the database, that tends show up as an init problem though.
Formatting the card sometimes works but if your card is working with the Oct 12 build I wouldn't think it was necessary....
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version