From: Randal Whittle (rwhittle_at_usa.net)
Date: Thu Jun 04 1998 - 20:01:37 EDT
Okay, today I had some time to see how long the battery lasts.
WHIP ME, BEAT ME...MAKE ME WRITE BAD CHECKS!
My own little "Battery Torture Test" yielded the expected result--somewhat
less than what IBM claims (the claim is 3 hours). However, it should be
understood that I was *really* working this thing. I gave it no rest
whatsoever and operated the CD drive an *awful* lot. Here's how it played
out:
For the first hour, I did nothing particularly odd--just a little editing
and transferring a bunch of files via LapLink over an Ethernet (3Com PC
card) connection.
After about 1 hour of that, I started installing software--a *lot* of
software--via the CD drive. Just to give you an idea of how *much*
software, here's the list:
- Palm III Desktop Software
- Norton Utilities 3.0
- Norton Crashguard Deluxe 3.0
- Winfax 8.0
- Norton Internet Fast Find
- Adobe Type Manager Deluxe (4.0)
- Extensis Phototools 1.1
- Extensis Intellihance 3.0
- Adobe Acrobat 3.0 (full version, not just the free "Reader")
- Quickview Plus 4.5 (2 times--had to re-install because I forgot to
install Acrobat first so it could integrate with it)
- I put in the Norton Antivirus CD, but the battery was getting low, so I
did not install it. I *did*, however, view a video from the CD, which took
about 2 or 3 minutes of non-stop CD spinning and use of the speakers.
After that, the battery gauge showed 5% remaining, so I didn't install NAV
because I didn't want it to collapse in the middle of the installation.
A few salient points: My battery settings are "Custom", which fairly
closely approximates a middle ground between "High Performance" and being a
complete miser. However, I don't think that matters really, because the
system was being *actively* used non-stop for the entire time. I was
always either typing, transferring a file, or actively using the CD drive
to install the software. Given those circumstances, battery life settings
are pretty much irrelevant.
Oh yes, one more point: Screen brightness was all the way up (using the
slider bar below the screen), but the ThinkPad configuration setting was
set to lower the brightness somewhat when on battery power as opposed to A/C.
Also: My machine is the P-II 233 MHz, so I don't know if the other CPU
flavors of TP 600 would be significantly different. I also have a 64 MB
RAM module installed, bringing total system memory to 96 MB. Of course,
the extra memory adds some level of additional power draw (I don't know how
much), though in actual usage, it probably saves more juice than it draws
by reducing the need for disk access to virtual memory.
Total life? 2:17.
The battery alarm went off at 2:15 and then at 2:17 it automatically went
into suspend mode. The last 7 minutes were spent simply sitting in place
while turned on and my hitting the shift key every couple of minutes to
insure that the screen wouldn't blank (so I could prevent it from saving
any juice on that front). So also in my little test, the screen never
turned off.
That makes for 1 hour of "normal" use, 1:10 of *active* CD usage (I'd say
it was actually *spinning* for about half of that time), and another 7
minutes of sitting still with the screen on.
I'm fairly impressed. My 560 would do only slightly better and it
*doesn't* have a CD drive to run (Though I understand that the newer 560X's
have a longer battery life than the originals. How much longer is it, by
the way?).
For that first hour, the battery gauge did a good job of measuring things.
Each time I checked, particularly at 75% left and at 65% left,
calculations showed it was precisely on track for a 3 hour battery life.
It was only after I started using the CD that it took a dive downward.
JUICE ME UP
So how long does it take to fully recharge? IBM claims 1 1/2 hours if the
machine is off, 2 hours if you're using it. In my experience, just using
the machine while I had the A/C plugged into it again, it took 2:20 to
charge it. A little bit disappointing, but it should be remembered that I
had that battery *literally* on empty.
One nice thing about IBM's battery gauge software: It not only shows a
percentage of charge for the battery, it also shows an approximate time
(while charging) it will take to arrive at a full charge. For whatever its
worth, its nice to be able to look at the battery gauge monitor and know
that you have about 1:20 left to charge the battery. If you don't have
that kind of time before you have to go and you need a full charge, you can
make a mental calculation about whether its worthwhil to shut down /
suspend the machine so it can charge a bit faster (about 25% faster). So
if its got 1:20 to go and I know I have to leave in 1 hour with a
fully-charged battery, I know I can do that if I'm willing to stop using
the machine and let it charge for an hour while suspended. It appeared to
be quite accurate too, given the 2:20 total it took to charge to 100%, the
guage was correct in how much time was left to charge.
Its little things like that that were *never* present or included with the
NEC Versa. They rely entirely on the built-in Windows stuff without any
custom-designed software for the specific machine. One size fits all (as
long as everyone is the same size!).
BATTERY MARK TAKES A STAB
For what its worth, ZD labs has their own "BatteryMark" test they run to
determine battery life. I don't recall exactly what the published result
was for the TP 600 (in PC/Computing about 2 months ago in a review of P-II
notebooks, in which the 600 was considered the hands-down favorite), but I
*think* it exceeded 3 hours. Unfortunately, I gave the magazine away so I
can't look it up now.
Sooo...rather than go download the 15 MB BatteryMark test myself, I went
and looked it up on-line. Ziff-Davis reports a battery mark of 3:12 for
the 600 (and that's on the P-II 266 MHz version).
If you want to know more about the BatteryMark benchmark (methodology,
etc.), you can read more about it at http://www.zdbob.com .
Its *really* nice to have IBM bring in good (actually, *very* good)
battery life on a performer of this caliber and be *honest* about the
figures. When I bought the TP 750C about 4 years ago, one of the biggest
reasons I did so was the fact that it had a claimed battery life of 3.5 to
8 hours. Now I never expected 8 hours, but I expected at *least* the 3.5.
It never came close--it was lucky to pull off 2 1/2 hours. It was *very*
angering and I almost sent the machine back. I made numerous calls to IBM
about it, and finally they relented and sent me another battery without
asking for the original one back. It fared no better, but at least I got a
spare battery out of the deal. I later found out (partly from this list)
that they *knew* darn well that it didn't do 3 1/2 hours, and sending me a
second battery without asking for the first one back was their way of
placating me.
Sure I'd like a second battery, but I'd be happier with a single battery
that performs as promised. The 600's battery appears to do this.
Underpromise and overdeliver--Kudos to IBM. At 3:12, its one of the
longest-lasting notebooks out there--particularly among P-II's (though the
770 adds another 25 minutes to that).
I don't have the BatteryMark score for the NEC Versa 5080X I had, but I
can tell you that in my typical use, it barely made it to 2 1/2 hours
before it gave up the ghost. Usually it was around 2:20. If I used a PC
Card modem in it, about 2:00 even--and that was with a 233 MMX chip, not a
P-II. Pretty poor, in my estimation.
ZD estimates that on average, P-II chips deliver about 20% better
performance than their plainer Pentium MMX cousins, while notebooks with a
P-II chip get about 15 minutes less battery life. There's a small tradeoff
(though with the 600, I'm hard-pressed to see it), but one that I doubt
matters much. If you lost an hour...perhaps 1/2 and hour, that would be a
big deal. But 15 minutes less than 3 hours would represent about an 8% to
9% loss of life in the battery--more if you're running a non-ThinkPad with
even worse life. ;-) 20% gain for 9% cost in battery. Worth it? Well,
the price is a factor too. Then again there are some that will pay that
difference in price just because its the coolest thing to get. After all,
when one buys a sports car, mileage isn't the point.
"NORMAL" USE
The BatteryMark test is probably the best figure for this, particularly
since it is the closest thing to an actual standard by which to measure
objectively. It claims to measure battery life "while performing
operations typical of mainstream business applications under 'real-world'
usage conditions." That's about as good a definition as any.
What's "Normal Use"? Frankly, nobody could define that for everyone else.
For me, it means active use, but without any extraneous power draws--no
CD, no PC Cards, and no Modem usage. But it does *not* include letting it
sit with the screen turned off--that would defeat my criteria of "active
use". "Active use means mainly just editing a document or spreadsheet and
that kind of thing. That's "Normal" for me (though in truth I actually use
the modem a lot for internet access), as I don't think anyone actually
considers it "use" when they leave their computer for 10 minutes to do
something and it shuts things down to save power. Those are wonderful
features, but I'm more interested in how long I can go on a battery charge
when I'm actually *using* the machine.
So the way I tested this was to do my E-mail (and some other misc. tasks).
I downloaded an E-mail packet while on A/C, then disconnected both phone
line and A/C, started my stopwatch, and went to it. Just reading,
reponding, etc. Also did a little work on a spreadsheet and also some
other writing in Word. I think these are "normal" tasks for most
folks--certainly for me.
Again, as in my personal "Torture Test", I have the backlight on all the
way, except for the system setting that drops it just a bit when on Battery
instead of A/C. I'm also using it enough that hard disk spin-down is
uncommon.
Time? 2:45--almost on the nose exactly. I should note that this
"normal use" was probably more torturous than it should have been. There
was some significant time that I couldn't be near the machine and yet I
didn't want it to suspend or power anything down (like the screen), so I
set the power settings to never power anything down--so the HD spun
constantly, etc.
Also, I was screwing around looking at various directories and deleted
something I shouldn't have. Fortunately, I had installed Norton Utils and
was able to unerase it, but there was a significant number of crashes and
system restarts in the process. I presume that takes extra juice...
At any rate, if those things shaved 15 minutes off the 3 hours it might
otherwise would have achieved, I wouldn't be too surprised.
To conclude this 3-part series, I can only say...a BIG thumbs-up! 5
stars out of 5. Unless you need hardware DVD (which you could presumably
add with a Zoomed video card) or an S-video out port or something, I can't
imagine why anyone would buy a 770 over a 600.
For that matter, I can't imagine anyone buying another notebook unless
price is the issue (which it all too often is), or if they just *have* to
have it *really* light and *every* ounce counts (in which case maybe you
should consider getting a Palmtop or Libretto or something). But for my
money (and it has been *my* money, since the moment that 560 met an
untimely death on the kitchen floo... ;-), I think the 600 manages an
*excellent* balancing act in bringing you a machine with an *almost*
waifish form-factor without the slightest skimp in feature set.
...and there you have it! Hope this has been informative and useful.
Signing off...
-------
Randal J. Whittle whittle_at_usc.edu (213) 740-7775
Director, Electronic Commerce Program
Marshall School of Business, University of Southern California
This archive was generated by hypermail 2.1.3 : Thu Jan 23 2003 - 09:54:07 EST