FS#12521 - iPod Classic/6G: cpu freq scaling

Attached to Project: Rockbox
Opened by Cástor Muñoz (prof_wolfff) - Sunday, 08 January 2012, 21:20 GMT
Last edited by MichaelGiacomelli (saratoga) - Wednesday, 11 April 2012, 01:59 GMT
Task Type Patches
Category Operating System/Drivers
Status Closed
Assigned To No-one
Operating System Another
Severity Low
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No


This patch implements HAVE_ADJUSTABLE_CPU_FREQ capability, CPU and bus frequency is divided by 2 and core voltage is decreased from 1.2 to 1.1V when unboosted.

Tthe previous ipod6g_holdswitch_poll patch must be applied before, otherwise the system is very unstable.

Closed by  MichaelGiacomelli (saratoga)
Wednesday, 11 April 2012, 01:59 GMT
Reason for closing:  Accepted
Comment by Nils Wallménius (nls) - Monday, 09 January 2012, 19:47 GMT
Tested this together with  FS#12520  and it seems to be working fine on my 120GB classic
Comment by Cástor Muñoz (prof_wolfff) - Thursday, 02 February 2012, 00:22 GMT
I am testing a more aggressive configuration: CPU: 54MHz, bus: 108MHz, Vcore: 1.050 V., it seems stable but need to test it a bit more, i will try to post it as soon as possible.
Comment by Albert Barca (axismundi) - Monday, 13 February 2012, 22:57 GMT
Works very well for me too, I've been using it for weeks on two Classics (80GB and 120GB). In addition to #12520, I have applied also #12523 and I get noticeably longer battery life while the player is very stable. Good job!
Comment by Cástor Muñoz (prof_wolfff) - Monday, 20 February 2012, 20:05 GMT
Thanks! v1 patch seems very stable, the new v2 patch i am posting sets CPU=54MHz, buses=54/54MHz (not sure, it could be 108/54) and Vcore=1.050V, it increases battery life about 15-20% (vs cpu-freq-v1 patch).

Battery life could be similar to OF when using this patch in conjunction with lcd_sleep-v2 patch (#12523), one tester reported battery bench lasted for 36h45m on a new 160Gb iPod, OF reached about 38 hours under similar conditions: one continuous run, autorepeat folder (70min. mp3@320), headphones inserted, volumen=-35dB, no EQ/effects, holdswitch on. I wonder if CPU voltage could be decreased one or two more steps...
Comment by Paul Collins (pjdc) - Wednesday, 29 February 2012, 06:50 GMT
I've been running v2 of this patch (plus charger v1 and holdswitch-poll v1) for a day or so on my skinny 160GB, and the results are excellent so far.