• Status Unconfirmed
  • Percent Complete
  • Task Type Patches
  • Category Battery/Charging
  • Assigned To No-one
  • Operating System iPod 5G
  • Severity Low
  • Priority Very Low
  • Reported Version Daily build (which?)
  • Due in Version Undecided
  • Due Date Undecided
  • Votes
  • Private
Attached to Project: Rockbox
Opened by stripwax - 2008-02-11

FS#8596 - Calibrate ipod 5g 64MB (60/80GB) battery discharge

Adds runtime calibrated from (numerous) battery_bench runs on ipod 5g 64mb (60gb)

ryran commented on 2008-02-25 07:04

Hey Dave. Did you make this using data from the testing of Andree's PP502x work … or just from your own benches?

My own benches after Andree's work had been committed to svn. They were correct as of the time of the original patch but I don't know if svn has got better (or worse) since then. But when get committed the calibration will need to change again.

I use the iPod 5g 60GB (The battery is the original battery(600mAh)).

My iPod's battery bench result is following:

max battery voltage is maybe 4100 - 4180 mV,
min battery voltage is maybe 3300 - 3360 mV.

These result is different Dave's result.
Is the battery of my iPod old?

Can you post your battery bench file here?

I post my iPod's battery bench result.

What battery life do you get with the original firmware (= this will be the best way to determine if your battery is faulty / on the way out)
I've combined your bench with today's bench of my own (my bench is svn here, looping mp3 playback, accessory volatage turned off) - OpenDocument format complete with chart

Some observations on your battery bench:
Starting voltage is indeed lower than mine. Was the battery fully charged (e.g. charged in the original firmware until it reported fully charged)? If not, maybe your battery is indeed losing capacity through age/usage

Your bench shows clear spikes due to the spinup/down for logging purposes. The latest battery bench does a somewhat better job of this, so if you could post an updated battery bench that would be really excellent.

Ending voltage is basically 3545mv. You can see from your log that only the last two lines report a voltage lower than 3545, and both of those are around 3370mv, and the step between 3545mv and 3365mv is about twenty seconds. This looks like the knee of your battery discharge curve. I would say 3545 is meaningful, and the readings below that are not important. I.e. if your battery is already reading 3370mv, there is probably not enough charge to save state and cleanly shutdown.

Even on that basis, your final discharge voltage is also lower than mine, not sure how to interpret that (could mean you battery is old, or mine is misreporting its voltage). 3550mv or thereabouts could be the best voltage at which to perform emergency shutdown

Thank you for the comment.

Now, the battery bench is executing.
When the result comes out, I will report again.

The result of the battery bench is sent.

use player info:

iPod video 60GB (model MA003J) 55.6GB free
music file: mp3 128kbps
Rockbox: svn.19195
iPod diagnostic mode:
     SDRAMQuickTest: 64MB SDRAM OK

1) charger connect and battery full charge on original firmware
2) charger disconnect
3) Rockbox start
4) charger connect and select menu: system > debug > view battery
5) It waits until the value of the battery doesn't go up.
6) start battery bench plugin.
7) It waits until the value of the battery doesn't go up.
8) charger disconnect
9) play mp3 file (repeat)


max battery voltage is maybe 4180 mV,
min battery voltage is maybe 3400 - 3500 mV. (maybe 3450mV)
My iPod is near iPod video for 30GB according to the result of the bench mark. 

Coming back to this. So one thing I find interesting from your output is that Rockbox suggests a time remaining of 16hrs 48mins but you only get 11hours 40mins out of your battery. Another thing I find interesting is that your max voltage of 4176 does indeed tie out with the current Rockbox 30gb battery info

Is it possible that your 60GB ipod actually has a 400mAh battery rather than 600mAh? I believe this has been observed in the past for some models, particularly refurbished/repaired models.

Just reading this patch: The operation desribed by Uchida shows no clean restart after charging in rockbox. The tested software (I am not sure about current beahviour) will draw ~2.5mA higher current compared to clean restart in this case.

Next question: Is this patch still of interest or should we close it?

I think it is still of interest. it would be great if
1. Other ipod 5g 60GB users can see how my discharge curve compares to theirs, especially w.r.t maximum runtime achievable (i.e. discharge time nearer to 16 hrs rather than 11 hrs, consistent with Ipod Runtime page)
2. Uchida is able to confirm that his battery is definitely 600 mAh and/or is definitely 400 mAh

so that we can then figure out next steps.

Assuming that the discharge curve for 400 mAh is indeed slightly different than 600 mAh, then we should be able to support both within rockbox, especially if refurbished units may have a lower standard battery than factory.

Alternatively, how hard (roughly) would it be for Rockbox to calibrate its own charge/discharge/capacity curves? I'm assuming that is how the OF does this, and it's not uncommon for most devices to have their own battery calibration.

How very, very, strange indeed. So, a recently completed battery bench, on my same ipod video, gives results that look much more like Uchida's. (A few runs between r25262 and r25372, with the highest overall voltage seen in the runs being 4136mV, and the lowest 3415mV; with overall runtime 17hrs:40m or thereabouts (I wasn't able to capture a continuous successful run from full charge to full discharge though, unfortunately, due to the bug in the recent pp IDE power change that made my player hang/data abort that I only discovered later)

Andree, any ideas why my discharge curve _voltages_ appear to have changed dramatically for me since last year or so? Could there be some (combination of) settings that actually affect the reported voltage readings themselves, perhaps? or will we just put it down to 'battery wear' and be none the wiser, do you think?

Dave, I have really no idea what is / was going with your battery…


Available keyboard shortcuts


Task Details

Task Editing