High battery drain in Rockbox

[saratoga]

The manual install method in the manual is offline. You can put the files on a thumb drive and run them elsewhere.

[Darthagnon]

Battery finally died, 9:50pm 23 July 2023. Off 1 charge, I listened for a few long car rides, and the odd few hours at home, the rest of the time it was off. Battery meter must be innacurate, as it didn't drop below 75%. The "3000mAh" battery lasts for around 20 days with moderate/light listening and auto-shutdown set to 5mins idle.

I noticed, when I heard a strange beeping noise - it was in lock mode and had attempted to boot on low battery by itself, so it had booted to OFW Apple Disk mode. I need to flash that modified bootloader to prevent this behaviour. Now I've put it to charge. I do wish the low battery bootloader mode wouldn't keep the screen on 24/7 telling me battery voltage is too low; this will inevitably cause screen burn; anyone know of any bootloader mods to disable this undesirable behaviour?

[Darthagnon]

The manual install method in the manual is offline. You can put the files on a thumb drive and run them elsewhere.

I'll test it with the modified bootloader; I was a bit worried due to the repeated warnings that "Windows is not supported" for this process, but I found offline Windows binaries of mk5lboot.

I have had some small success with selfhosting on localhost the zips downloaded by the Qt installer GUI and using a proxy redirect to force the GUI to download from localhost, though the URLs it searches for seem to change (e.g. random strings sometimes added) - do you have any suggestions with regards to that?

[Darthagnon]

From emails to the ipodrepairs.co.uk, it's possible the battery drain issue is probably a problem with the LTC4066 power management IC. Upon full disassembly of the iPod, there is visible corrosion around it, as well as around some areas connected directly to the 30pin port. By corrosion, I mean green/blue copper sulphate looking stuff or white crystalline. It is minimal, but the components are tiny. Maybe if I'm lucky, cleaning that off with a bit of isopropyl alcohol will make it work properly.

From my reading, it is likely a complicated fix otherwise. Other people cut traces, soldered diodes, or even suggested buying a new LTC4066 component, but replacing that is likely impossible due to it being surface mounted, with a blob of heatsink solder behind the middle of it and a bunch of small caps/resistors nearby. And I don't have a hot air station.

[saratoga]

Possibly some water got in the connector and broke a trace.  If you had hot air, I'd suggest putting flux on the corrosion and reflowing it to dissolve the corrosion.  Possibly you could do that with a regular soldering iron as well.

If you do decide to buy a replacement LTC4066, they're available new from official distributors: https://www.digikey.com/en/products/detail/analog-devices-inc/LTC4066EUF-PBF/1746224  I would not buy a second hand one from eBay sellers, you're likely to get something scrapped from an old board or worse. 

[Darthagnon]

Just an update for what I did in the end: I cleaned my iPod's motherboard around the component with isopropyl alcohol to remove the corrosion. It has now worked perfectly for me for over a year.

[iPodVT]

Just an update for what I did in the end: I cleaned my iPod's motherboard around the component with isopropyl alcohol to remove the corrosion. It has now worked perfectly for me for over a year.

This inspired me to revisit my iPod that has the same problem.  I didn't see any obvious corrosion or other problems around that IC, but I cleaned it with 99% isopropyl alcohol.  That did not fix the problem, so then I tried saratoga's suggestion of applying some flux and reflowing with hot air.  That also did not fix the problem (but nor did it create any new problems as I feared it might considering my as yet modest hot air skills).  I had high hopes, so I'm disappointed and frustrated, as that iPod works 100% in every other way.