From: SMS on
Ron Hunter wrote:

> The symptom exhibited by my last camera was a shutoff of the flash
> operation, completely, followed by a period of being able to take
> flashless pictures (20 or so), and then shutdown during writing to the
> flash card. I, luckily, lost only the single picture, but had this
> happened while writing to the FAT, I would most likely have lost the
> whole card's contents.
> My current camera warns of low battery, and then just refuses to take
> more pictures when shutoff point approaches. A better solution, in my
> book.

If you read the application note by Maxim, referenced earlier, you'll
see the problem with measuring remaining NiMH capacity by voltage.

Most of the capacity is used in an extremely narrow voltage range of
about 0.05 volts. While it's possible to measure this accurately, it's
not cheap, requiring precision, temperature compensated voltage
references. The flat voltage curve of NiMH is an advantage in many ways,
but a disadvantage in terms of a cheap way of measuring remaining capacity.

With NiMH, you can know when the batteries are almost exhausted, and
completely charged, but not know much information in-between, unless you
go to a test that is more complex than just measuring voltage under
load. With most NiMH powered cameras, you simply get a low battery
warning, then maybe a few more pictures, then it shuts down. If they did
provide a gauge it would be horribly inaccurate if it was based solely
on voltage.
From: ASAAR on
On Wed, 27 Jun 2007 07:22:03 -0700, SMS wrote:

>> Remaining capacity on a LiIon also varies by temperature and discharge
>> rate. You aren't building yourself a case here.
>
> Sure I am. The difference is in the voltage curve. The Li-Ion pack
> declines linearly with capacity over a greater range. It's much cheaper
> to measure in 0.2V increments than 0.05V increments.

This is nothing but a bogus argument. You can get battery monitor
ICs that have greater accuracy than that, that can be used for NiCd,
NiMH and Li-Ion batteries, and cost one dollar or less. How is it
that manufacturers can afford to make inexpensive PDAs (under $100)
that can measure voltages with 0.01v increments, but not in $150 to
$500 digital cameras, that you're implying can only detect voltages
to 0.2v accuracy. That's complete nonsense, and if the accuracy was
that low, even "idiot light" battery warning meters wouldn't be able
to work properly.

From: SMS on
Ron Hunter wrote:

> The symptom exhibited by my last camera was a shutoff of the flash
> operation, completely, followed by a period of being able to take
> flashless pictures (20 or so), and then shutdown during writing to the
> flash card. I, luckily, lost only the single picture, but had this
> happened while writing to the FAT, I would most likely have lost the
> whole card's contents.
> My current camera warns of low battery, and then just refuses to take
> more pictures when shutoff point approaches. A better solution, in my
> book.

There's another issue with AA batteries and gauges as well, one that
Nikon addressed in the battery grip for the D200, but which is not
present in any compact cameras that I'm aware of. The nominal voltage of
AA batteries is different for NiMH, alkaline, and lithium cells, as well
as the remaining capacity based on the voltage.

AA powered cameras can use alkaline AA batteries (nominal voltage of
1.5V), as well as lithium AA batteries (nominal voltage of 1.7V) or NiMH
batteries (nominal voltage of 1.20-1.25V). On the battery/vertical grip
on some digital SLRs (which can often use Li-Ion packs, AA alkaline, AA
NiMH, or AA Lithium) there is a switch that you set to the chemistry of
battery you are using in order to have the battery gauge match the
characteristics of the battery being used. There is just no way that a
consumer level compact camera is going to implement such a switch.

A low battery warning indicator that works passably well across the
different battery types is possible, but forget about a gauge that
provides any accurate information. It's like measuring with a
micrometer, marking with a crayon, and cutting with an axe.

While it's nice to know how much battery life you have left, I suppose
that for the casual user it's not of the utmost importance. You charge
your batteries at every opportunity, and always carry spare sets. I wish
more AA powered cameras had the ability to charge inside the camera via
a dock, like some Kodak models (though these use proprietary dual AA
NiMH packs). It's a good idea to minimize the number of times that
batteries are removed and replaced due to the limited lifetime of other
camera parts.
From: ASAAR on
On Wed, 27 Jun 2007 07:25:29 -0700, SMS wrote:

> If you read the application note by Maxim, referenced earlier, you'll
> see the problem with measuring remaining NiMH capacity by voltage.
>
> Most of the capacity is used in an extremely narrow voltage range of
> about 0.05 volts.

That's completely misleading. In the 0.7C discharge rate table
(far more representative of camera use than the 0.1C table), at 20�C
(68�F) the capacity drops from 2,707mAh @ 1.35v to 910mAh @ 1.15v.
So at 1.15 volts, 34% of the battery's capacity remains. That will
decline to 2% remaining at 1.10 volts. Put another way, with these
batteries, Canon's A610/A620 would be able to take nearly 600 shots
using the CIPA test.

At 1.15v/cell, where the voltage is getting pretty low, and which
is the voltage you're using to mark the point where most of the
capacity has been used (it's at the 1.20v level for the 0.1C table),
there would be about 200 shots remaining. At 1.10v/cell, there
would still be more than a dozen shots remaining. And remember, the
CIPA test is pretty heavy on battery usage, keeping the LCD on and
using the flash at full power for 1/2 of the shots. Detecting these
voltage levels is trivial for even *very* cheap circuits. Making
this even easier, cameras don't operate from a single NiMH cell.
The A610/A620 uses four, so the battery voltages that correspond to
200 shots and 13 shots remaining would be 4.6v and 4.4 volts, well
within the ability of very inexpensive (less than $1) battery
monitor ICs. Anyone that thinks that this causes any kind of
problem in measuring remaining NiMH battery capacity is hopelessly
confused or in the wrong line of business.

From: SMS on
Thomas T. Veldhouse wrote:

> For God's sake, you didn't even know about pulse load testing until I told you
> about it and gave you a link to that tester.

Huh? One of my chargers has a slot that does pulse-load testing. I've
known about it for quite a while.