Print

[scn]

Rather than fill up the comments (which are becoming somewhat overwhelming), I thought I'd ask here: what's your real world battery life under OS X, and under what configuration?

I was looking at around 5-6 hours with half LCD brightness, which I'd call significantly less than in Windows, especially without wifi. Everyone else seeing the same thing?

[pwm8]

Its so hard to benchmark these things, but here's my real world experience.

I unplugged yesterday evening at 8:30pm with a full charge.  Battery meter settled down at about 5 hrs 30 mins (it usually takes a few minutes to get equilibrium where its not bouncing all over the place).

I surfed, did email, and generally just messed around for about 45 minutes.  Then I wanted to really hit the machine to see how the battery held up.  I watched an episode of Jon Stewart on Hulu.com (22 mins at high resolution), and watched 2 half-hour shows on my slingbox through my LAN (streaming about 3000kbps).  Then I surfed some more and finally shut down for the night about 10:45 pm.  So, two solid hours plus of heavy use with wifi on, brightness at full, using video and sound for much of the time.

When I shut down at 10:45, my battery meter showed 2:59 remaining.  So all in all, I'd say the battery meter is pretty accurate and Im getting over 5 hours on OS X at full charge, even with heavy use.

   

[kadave]

Today i thought i would test the battery life of my 1000he running ideneb 1.3. I turned it on today at work at 9:15am and am pleased to say that it has remained on until 4:10pm with 10 minutes of estimated battery life remaining. I'm happy to learn i can get 7 hrs battery life with the brightness turned up half-way. I also played an .AVI file lasting 22 minutes 3 different times while running itunes in the background. Hope that helps.

[hceuterpe]

I bet it's that stinkin' fan that's spinning all the time that's causing the drop in battery performance..

Anyone got an idea how to control it?? I don't think simply disconnecting them is a good idea.  Considering the much lower power consumption envelope of a netbook, its cooling fan (which probably isn't using any less power than a regular fan that size)  probably would be a much bigger hit than one in a conventional notebook.  Yes?

[rcfa]

OS X isn't optimized for the Atom CPU, it also uses the GPU heavily, unlike XP, so that sucks a lot of power. For similar reasons (and being even more of a resource hog) Vista isn't even for sale on these types of machines.

Anyway, 5.5h of real work, rather than "battery testing while doing nothing but text typing on a dim screen), is pretty good. I'm not sure I'd get much more under WinXP under heavy use. In any case, it's more than I need in any given day, and I rather have the fan going than a hardware melt-down....

[scn]

Since I started this thread, 10.5.7 seems to have changed things drastically for me. I know some haven't noticed much of a difference, but I get, as far as I can tell, much closer to the same life I get in Windows 7. I'd say it's a difference of about 10% (maybe less) now compared to 25% (or more) previously. Highly imprecise numbers, I know, but still a noticeable improvement.

OS X may not be heavily optimized for the Atom, but something in 10.5.7 definitely changed...something.

[LonelyTV]

I'm running 10.5.7 Retail on my EEEPC 1000HE and in my classes (6 hours) I generally run the machine with WIFI and with about half to less than half LCD brightness and I can get about 6-7 hours on it. Now, I'm only doing basic text editing and surfing... and occasionally the screen will turn off... but when I get home I usually have about 1-2 hours left on it.

Now, I think if we can get someone who knows how to code KEXT we might can get more juice out of the battery. According to Superhai:

I don't know much about ASUS EEE, but it seem like a proprietary method which ASUS has implemented for that series of laptops. I would believe you could make a custom kext to access the ATKD ACPI device and use the CFVG/CFVS methods to control it. Anyway I am not plannng to implement such proprietary methods in this kext.

This was in response to me asking about modifying voltages correctly, VoodooPower as is only drops the processor to 997 mhz (950 mV) and 1.66 ghz (1075 mV) where Asus' "Super Hybrid Engine" on Windows drops it to 600 mhz and (on Bios 601) 1.75 ghz. So, if a new KEXT can be coded specifically for the EEE PCs we can probably get that 8-9 hour mark. VoodooPower also jumps around too much with the voltages so if it can be leveled out it might also effect battery power. Just a thought!

[rcfa]

I'm not sure about the details, I'm not a kernel hacker, but I wouldn't think that ASUS needs to go very proprietary. Much more it seems they simply do what all the overclockers do: go beyond the manufacturer specs. So it could be as simple as ignoring the limits of what is reported as the CPU range, and setting lower, respectively higher values.
I wonder if his kext could optionally provide for user supplied overrides...

I don't get that much battery life, although I likely have a similar setup, but I also notice that some apps, e.g. Skype, tend to use quite a bit CPU even when "idle". Also, do you have the factory WiFi card, or did you swap in an Apple card, like me? If so, which model? I took one out of an AppleTV, so that might not be the most power efficient card around...

[LonelyTV]

I wonder if his kext could optionally provide for user supplied overrides...

You can, In his VoodooPower Manual he details how you can add voltage minimum and maximum for the KEXT to use. However, no matter what voltages I supply the KEXT it ignores my changes and uses what it 'knows'. However, I since ASUS not only makes EEE PCs but also Motherboards, I can see them using something that is "in-house" to throttle the voltages that isn't so conventional. Though, I would assume any one using VoodooPower on an ASUS board MIGHT could do the same 'hacking' to get the KEXT to bend to their needs.

Also, do you have the factory WiFi card, or did you swap in an Apple card, like me? If so, which model? I took one out of an AppleTV, so that might not be the most power efficient card around...

I replaced the stock card with a "Apple Airport Extreme Card MA688Z/B 802.11n" card. It was, presumably, pulled from a MacBook. It has a Broadcom 94321MC chip driving it.

[rcfa]

I wonder if his kext could optionally provide for user supplied overrides...

You can, In his VoodooPower Manual he details how you can add voltage minimum and maximum for the KEXT to use. However, no matter what voltages I supply the KEXT it ignores my changes and uses what it 'knows'. However, I since ASUS not only makes EEE PCs but also Motherboards, I can see them using something that is "in-house" to throttle the voltages that isn't so conventional. Though, I would assume any one using VoodooPower on an ASUS board MIGHT could do the same 'hacking' to get the KEXT to bend to their needs.

I see. Bummer.

Also, do you have the factory WiFi card, or did you swap in an Apple card, like me? If so, which model? I took one out of an AppleTV, so that might not be the most power efficient card around...

I replaced the stock card with a "Apple Airport Extreme Card MA688Z/B 802.11n" card. It was, presumably, pulled from a MacBook. It has a Broadcom 94321MC chip driving it.

Not sure about the part number. Have the same Broadcom chip, but possibly less power conscious support electronic given that I yanked it out of a AppleTV, but then again, the AppleTV has no active cooling, so it probably uses low-power stuff to prevent overheating anyway... So we likely have the same card.
Just wanted to know if you may get better battery life due to a lower-power WiFi card. But it seems that my habit of keeping Skype open is what sucks my battery dry more quickly.

[rcfa]

I found the following here: http://www.techarp.com/showarticle.aspx?artno=420&pgno=9

Super Low Frequency Mode

This technology is only implemented in mobile Core 2 Duo processors (Merom core). In the normal low-frequency mode, the minimum clock ratio supported by Intel is 6x. If the Front Side Bus base clock is 200 MHz, then the minimum CPU clock speed would be 1.2 GHz. Intel wants the clock speed to go even lower, as power consumption is directly related to (voltage2 x frequency).

Since a new PLL design isn't a good idea, Intel reduces the Front Side Bus speed to to only 100 MHz. This forces the CPU to run at only 600 MHz with a lower core voltage than the normal low frequency mode. Desktop Conroe-core processors cannot use this mode because it requires chipset support.

I think that's what the EeePC must be using, since what VoodooPower does is use the regular voltage range that's also advertised on intel's Atom Chip spec:
http://ark.intel.com/Product.aspx?id=41411

So this must be a chip set feature, but it's likely not Asus proprietary, but specific to certain intel mobile chip sets, since obviously Asus is using here rather standard componentry.

So I guess SuperHai would have to inquire power management beyond the naked CPU and include standard chipset features, if he is inclined to get the max. out of these computers. But of course, should he just be interested in desktops, as a lot of hackintosh users just want a cheaper Mac rather than a Mac OS X machine in a segment that Apple doesn't offer anything in (netbooks), then he'll have little incentive to do so, since the article clearly states that this is only available in certain intel mobile chipsets.

[shoota]

I just turned off bluetooth in the bios since i never use it and my estimated battery life jumped up almost 2 hours!

[rcfa]

I just turned off bluetooth in the bios since i never use it and my estimated battery life jumped up almost 2 hours!

What happens if you turn off BT from the menu bar or the BT preference panel?

[shoota]

I just turned off bluetooth in the bios since i never use it and my estimated battery life jumped up almost 2 hours!

What happens if you turn off BT from the menu bar or the BT preference panel?

well my lazy butt hasn't installed the BT kext yet so i don't know. i bet it was killing my battery cuz it was always on huh?

[rcfa]

I just turned off bluetooth in the bios since i never use it and my estimated battery life jumped up almost 2 hours!

What happens if you turn off BT from the menu bar or the BT preference panel?

well my lazy butt hasn't installed the BT kext yet so i don't know. i bet it was killing my battery cuz it was always on huh?

Sure. I'm just surprised it's by that much, because BT is supposed to be a low-power device, which is why e.g. you can run a BT keyboard or mouse for weeks on a set of two AAA batteries. That's why I wonder if that for some reason triggers other stuff than just turning off the BT transmitter.

[shoota]

it's possible. now that i think of it while i was in the bios turning off BT i also turned off the card reader but i doubt that has much affect. 

i think the most likely scenario is that my battery icon estimates are off. I just don't feel like it's displaying the correct info. for example the first time i installed os x the icon would display the right percentage left but not the right time. now, on the second install, i feel like it flips between the correct time remaining and the wrong time remaining but the percentage always stays at 99% or 100%...

so i don't know what's going on, i just use it and then plug it in at night

[LonelyTV]

I've ran into a problem with 10.5.7, for what ever reason now my install is showing 2:30 hours in battery life. Mind you, it hovers around that for sometime. My percentage stays at 100% for about 30-45 minutes then it starts to decrease (quickly). Has anyone experienced this with 10.5.7 and the latest updates (aside from 10.5.?

[rcfa]

Nope, no such issue. My machine runs for at least 4-5h on battery.
Of course, currently the display is a bit screwed up, due to the 10.5.8 upgrade, and I'm still investigating the proper way to fix the battery run-time display under 10.5.8 (there was no problem under 10.5.7 for me).
But even with current display issues, it runs the standard amount of time which is at least 4-5h, and often much more.