Rockbox.org home
Downloads
Release release
Dev builds dev builds
Extras extras
themes themes
Documentation
Manual manual
Wiki wiki
Device Status device status
Support
Forums forums
Mailing lists mailing lists
IRC IRC
Development
Bugs bugs
Patches patches
Dev Guide dev guide
Search



Donate

Rockbox Technical Forums


Login with username, password and session length
Home Help Search Staff List Login Register
News:

Welcome to the Rockbox Technical Forums!

+  Rockbox Technical Forums
|-+  Support and General Use
| |-+  Hardware
| | |-+  Philips HDD6330/17 - Manufacturing Variations?
« previous next »
  • Print
Pages: [1]

Author Topic: Philips HDD6330/17 - Manufacturing Variations?  (Read 4490 times)

Offline soundman

  • Member
  • *
  • Posts: 10
Philips HDD6330/17 - Manufacturing Variations?
« on: November 19, 2014, 01:31:10 AM »
I'd started a thread about the HDD6330 a couple of weeks ago, mostly addressing that on mine, I have to set the disk spindown value to a high level to avoid issues with the database reporting that it is "not ready" after spindown occurs. I've still been unable to resolve the issue (without increasing spindown) despite several attempts at reformatting the drive and reinstalling RB in both Windows and Linux, and trying every option available regarding file/folder attributes and permissions.

I came across a thread from 2011 where a user reported the following issue when attempting to install RB using the RB utility, and it also occurs with mine. Here is his post:

"I tried installing RockBox on my HDD6330/17 using the utility, which seemed to go fine until I tried booting it and it hung on the GoGear screen. I had also installed the boot loader, but to no avail. I've restored it to the OF using an old HDD6330/17 I have, but I really want Rock Box on there. Has anyone else managed to install RB on said model with success? Is there any reason it might not work? Does the /17 denote a different chipset? I'm a bit of a noob here..."

There must be some different versions of the HDD6330/17. Mine also hangs at the "GoGear" screen after installing with the RB utility. The cause appears to be that when using the Rockbox utility, the Rockbox "FWImage.ebn" file that replaces the original is completely different than the correct one that is available for manual install. The "FWImage.ebn" that is placed in the System directory by the RB installer is only 5kB and contains unencrypted text (file attached). The correct RB "FWImage.ebn" is 73kB and contains mostly encrypted/binary data with a small amount of text at the end of the file.

I've used the RB utility in both Windows and Linux versions with the same result. I'm always in UMS mode when attempting this, and the result is always the same.

I have no idea whether this could be related to the "spindown" issue or not, but when I looked at "disk info" on the device, the HDD is listed as a Toshiba MK3006GAL, and the "Spinup time" is shown to be 0ms. However in the datasheet for the drive:

http://www.webteufel.com/HP/ThinClient/HP-Compaq_T5730/MK3006GALUserGuide-Rev02.pdf

The spinup time is listed as 3.5ms.

I can see how this could possibly be related to the issue I'm having with the database, and the need to never allow spindown to occur or else go into a perpetual "database not ready" loop. I'm not sure how it could cause the incorrect FWImage.ebn to be written when using the installer though.

There may not be enough people using this model to make it worthwhile for the developers to look into this any further, and that is understandable. However I'm willing to provide any information necessary that might help to "cure" the HDD6330/17's that seem to have more issues than others.

One additional thing that I've noticed regarding the database on mine; after I initialize the database and reboot, when I view the "Database Info" it says:

Initialized: Yes
DB Ready: Yes
RAM Cache: No
RAM: 0/0 B
Progress: -1% (0 Entries)
Curfile: -----
Commit Step: 0
Commit Delayed: No
Queue Length: 0

Some of these are peculiar, because the database is functional as I've increased the spindown parameter as described before. Thus, it seems odd that it would say "0 entries". Also, it goes through the 1/9, 2/9...9/9 commit steps on reboot but reports a commit step of 0.

I'm just wondering, am I the only one with a HDD6330 who is currently experiencing these issues?
* FWImage_ebn.txt (4.1 kB - downloaded 109 times.)
Logged

Offline wodz

  • Developer
  • Member
  • *
  • Posts: 390
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #1 on: November 19, 2014, 03:49:30 AM »
The FWImage.ebn file should definitely be binary, encrypted/scrambled file. This looks like either bug in RB utility or some misconfiguration of download section on our webserver.
Logged

Offline b0hoon

  • Developer
  • Member
  • *
  • Posts: 71
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #2 on: November 19, 2014, 03:54:20 AM »
There are two versions of this player: HDD6330 and HDD6320 (which i have), only the last one doesn't have a radio, that's all. Firmware file is the same for both. End of the number like /17 denotes probably a small changes in design of the mainboard, chipset is the same from the beginning.

There should be no problem with installing Rockbox through the Rockbox Utility, Rockbox Utility should replace original firmware file FWImage.ebn with our bootloader (73kb), cause this allows dualboot.

Next thing: You haven't changed anything, by changing "Spindown" parameter, because Rockbox has absolutely no control over a disk spindown on this particular target, see: http://git.rockbox.org/?p=rockbox.git;a=blob;f=firmware/target/arm/philips/power-hdd.c;h=a7f1302a7c3751f3f347c1cf8c8a55eca59a1550;hb=HEAD, function -> ide_power_enable. Disk spindown is controlled by the hardware.

From other post - i'm too lazy to reply for both :) :
Quote
Another thing I noticed.. not sure if this is supposed to happen or not, but when it's updating the database and showing the count progress, pressing the play/pause button increases the database update speed dramatically without causing an error. Really speeds things up.
Yes, it's probably related to boosting up the processor speed , when using the buttons and touchpad. It's normal.

Quote
If there is one thing I could change about how RB interfaces with the device, it would be to decrease the sensitivity/speed of scrolling through menus. It's nice that you can tap the scroll bar to select, but I always seem to overshoot the desired item by one or two items.
Yeah, i realize that touchpad support is not so groovy. I wrote a basic functionality, and focused on other things, more important for me. But there is a patch on gerrit which probably will be used in future for this Gogear: http://gerrit.rockbox.org/r/#/c/786/

If you feel that your player is acting strange , you probably have a disk half dead, it has some years already in the end. Replace it, or modify your player using CF card + adaptor.
« Last Edit: November 19, 2014, 04:16:10 AM by b0hoon »
Logged

Offline soundman

  • Member
  • *
  • Posts: 10
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #3 on: November 19, 2014, 06:39:51 PM »
Thanks for the replies. I plan to get a newer player as my 6330 is literally starting to fall apart, but I'm glad I still had it sitting around so I could try out RB. My last player was a 6th Gen iPod so I was stuck with their firmware.

I agree that the installer issue for the 6330 appears to be some sort of misconfiguration between the application and the web server, as the file that gets written as "FWImage.ebn" appears to just be the HTML code for the menu of the download page.

As far as the database issue mine has, and how increasing the spindown value resolves it, I agree that it doesn't appear (from feel and sound) to actually keep the disk active any longer than it normally is. However, anytime my player exceeds the value that is set, I invariably get "database not ready" along with panic crashes (usually "could not write to sector XXXXX (error-2)). So the parameter is affecting something if not the actual spindown.

If I had the time, I might take on the task of trying to build a version that disables all database functions completely. All I've ever really wanted is a player that uses a simple file manager as the GUI. I've never had anything but issues with corruption on both my Apple and Philips using original FW. Who looks up a song they want to hear by the author's name, date released, etc? It's easy enough just to group albums in folders by artists. I suppose it's all linked to DRM, but I'm mostly copying from CDs and converting to FLAC. Compressed video is fine and necessary for bandwidth, but I can distinguish between the best-encoded mp3 and a lossless version in three seconds. They just sound listless and "dead" somehow. Oh my.. how I digress.
Logged

Offline soundman

  • Member
  • *
  • Posts: 10
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #4 on: November 19, 2014, 06:51:13 PM »
Just a question from having read the manual for another device.. the HDD6330 doesn't have the option to store the database in RAM in the database options menu. Is this due to limited RAM in the 6330 or is it just something that never got implemented?
Logged

Offline saratoga

  • Developer
  • Member
  • *
  • Posts: 8974
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #5 on: November 19, 2014, 07:17:01 PM »
Quote from: soundman on November 19, 2014, 06:51:13 PM
Is this due to limited RAM in the 6330 or is it just something that never got implemented?

I believe that option was removed from rockbox a few months ago when the file system was reworked.
Logged

Offline soundman

  • Member
  • *
  • Posts: 10
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #6 on: November 21, 2014, 05:40:32 AM »
I think it's gone into a "grounded vindaloop" of some kind. I never should have let Butters borrow it.
Logged

Offline ulmutul

  • Developer
  • Member
  • *
  • Posts: 20
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #7 on: December 19, 2014, 08:37:10 PM »
I also have the "spindown"-issue. For me, it presents this way (I don't use the database):

(1) When I select a file (or a plugin), I get an error message, the disk begins to spin. When I re-select the file while the disk is spinning, the file loads normally. (And sometimes everything works fine at the first time.)
EDIT: If I get the error message, I get it immediately and the disk DON'T spin up. When I try a second time (within a short time), the disk spins up and work as it should.
(2) When using usb file transfer with rockbox, I sometimes get the message (windows) that the drive must be formatted. When I abort, all files show up normally and I can continue without any problems.
(3) I get a *Panic* mount: 0 message when disconnecting from usb after file transfer
(4) I (rarely) get a dc_writeback_callback() panic ("Could not write sector") (e.g. when the device is shooting down/RB tries to write config file)
(5) I never had any issues when transferring files via OF bootloader UMS mode.

This issue doesn't have a high priority for me as I can get on with it, and I'm dealing with other devices at the moment, but I'll work on this when I find some time.

I think it's a matter of the disk drive, maybe different versions or just slow or faulty individual drives.
To me it looks like rockbox should give it just a bit more time to spin up and respond before reporting an error.

BTW: back says: HDD6320/00

EDIT: wrong guess, see (1)! The error occurs BEFORE the disk spins, preventing it from doing so. As soon as the disk starts to rotate, everything just works.
« Last Edit: December 25, 2014, 05:59:53 PM by ulmutul »
Logged

Offline LeoTheLion89

  • Member
  • *
  • Posts: 34
    • Matthew's Custom Computers
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #8 on: December 24, 2014, 11:39:24 PM »
FIX TO BOOTLOADER ERROR:

I TOO HAVE THIS PROBLEM IT IS SOMETHING WITH THE ROCKBOX UTILITY. IF YOU INSTALL THE BOOTLOADER MANUALLY ON THE MANUAL INSTALL PAGE AND COPY THAT FILE TO THE PLAYER IT BOOTS FINE
Logged

Offline Peebles

  • Member
  • *
  • Posts: 2
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #9 on: December 26, 2014, 08:47:02 AM »
Hi:


I'm owner of HDD6320/17.

I also faced the problem with the installation error of the firmware bootloader via RockBox Utility. Agree on some web server problem. My workaround is using the RockboxUtility for everything but for the installation of the bootloader, and then grab the file from http://download.rockbox.org/bootloader/gogear/hdd6330/ , create a "System" directory and copy the file (FWImage.ebn) there manually.


Now, for the hard stuff:

I'm getting continuous Panic errors (could not write sector ....) with no apparent rule for replicating them. Sometimes after finishing playing an album, sometimes when the device goes to sleep, sometimes when powering off by hand...

I've re-checked the disk with linux utilities finding no bad sectors on disk.
I've tried making several partition setups for avoiding potential bad sectors:

- 30 GB FAT32 covering the whole disk.
- 4GB FAT32 only, at the begining of the disk
- 4GB FAT32 only, at the middle of the disk
- 4GB FAT32 only, at the end of the disk

I've tried different firmware builds (http://download.rockbox.org/daily/gogearhdd6330/ ).

I've tried setting different values for the Disk spindown parameter (5s, 60s, 120s, 250s) with no luck.

Any clue anyone ?

Thanks for the hard work.
Logged

Offline [Saint]

  • Rockbox Expert
  • Member
  • *
  • Posts: 1662
  • Hayden Pearce
    • Google+
Re: Philips HDD6330/17 - Manufacturing Variations?
« Reply #10 on: December 28, 2014, 05:47:54 PM »
THIS TEXT IS IN BOLD AND CAPS BECAUSE I'M IMPORTANT
Logged
Using PMs to annoy devs about bugs/patches is not a good way to have the issue looked at.

  • Print
Pages: [1]
« previous next »
+  Rockbox Technical Forums
|-+  Support and General Use
| |-+  Hardware
| | |-+  Philips HDD6330/17 - Manufacturing Variations?
 

  • SMF 2.0.17 | SMF © 2019, Simple Machines
  • Rockbox Privacy Policy
  • XHTML
  • RSS
  • WAP2

Page created in 0.121 seconds with 15 queries.