Just a thaught from exercise equipment I have seen
though it may be "redefining the problem"
Could you put a super/ultra capacitor or some such inplace of the battery,
then before you use it you "run" the machine a little.
The example I am thinking of is a bike at a local gym
you get on it pedal once or twice and the display comes up,
if you stop for a while it will eventually turn off.

You arent supplying power continuously but then at 250mah draw a 3ah
SLA wont last long anyway.
Its also pretty light and small which is typically an advantage.
(Unless the exersice is getting the thing home and out of the box ;->)
It might cost a little more though probbly not when done in quantity. 

should be good for 100k uses, and you don't have to worry about some luser
trying to drink the battery acid or dropping it on their cat or some such.


put a low voltage sensor in your code and save your state to eeprom before 
switching off, person "pedals" again and it comes back where they left off.

There are some very funky supercaps around the place hundreds of farads in 
a C-cell package (though you pay for it ~$10 in qty last I checked) Upside 
is you can start a car off a single string of sub-c cells and a bunch of
boostcaps.


> -----Original Message-----
> From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu]On Behalf
> Of Russell McMahon
> Sent: Friday, December 16, 2005 00:18
> To: PIC List
> Subject: [EE] Battery choices
> 
> 
> I have a volume application that requires a battery.
> Best choice seems to be "sealed lead acid" (SLA) for reasons listed
> below. Suggested alternatives welcome.
> 
> NB: Solutions MUST NOT redefine the problem, suggest alternatives
> which do not meet all criteria etc :-). Often enough when I ask
> similar questions many answers address requirements which I do not
> have. There are reasons for most of the requirements . I will
> partially explain some of them if it seems useful.
> 
> Application:
> 
> Exercise equipment operated by 'user'. When operating, power is
> provided by user actions via an 'alternator'. When at rest or when
> user speed drops too low (startup, programming, pausing etc) power is
> supplied by a battery. Battery is recharged by user action when
> possible. Current drain is up to 250 mA. Supply internal rail is
> typically 5V but may use the dread LM7805 type regulators so input
> needs to be say 8v min (dropout, headroom, wiring loss, connector
> loss, ...). Higher OK but too much higher may stress 7805 thermally as
> manufacturer probably anticipated a 6 cell battery pack (nominal 9v).
> Equipment is in two parts. Part A draws up to 250 mA as above and its
> design cannot be altered. Part B draws under 20 mA, design is flexible
> and includes power conversion and control functions.
> 
> Requirement:
> 
> Rechargeable battery.
> Good shelf life under self discharge.
> Lowest practical cost.
> OK recharge rate (faster the better).
> Reasonable capacity - say 1 to 2 AH OK.
> Battery life needs not be optimum but should be reasonable. ie some
> violation of best practice float etc voltages and charge currents is
> OK as long as battery life of say 3+ years is achievable.  Much of
> this equipment gets very little use but some is used extensively. A
> mains supply is a possible option in some cases but cannot be relied
> on. Mass and volume energy densities are not important.
> 
> Alternator output is rectified to a smoothed DC rail and may reach 120
> VDC plus in some cases. For reasons not covered here this is to be
> converted to battery charging and equipment supply voltage by a linear
> regulator so dissipation is up to about 30 Watt worst case for short
> periods. This can be handled OK.
> 
> ____________
> 
> Solution so far:
> 
> 12v SLA. Typically 1.2 or 3 AH.
> 
> 6v SLA with a converter to lift output to about 8v for "part A" supply 
> is a possibility but cost an issue.
> 
> 12V SLA requires a voltage rail from alternator of 15v+ to get enough
> headroom to charge SLA.
> Dissipation in equipment part A at 250 mA is therefore about 
> (15-5)*0.25 =
> 2.5 watt. A series resistor can be added to drop some of this.
> SLA charging circuit is crude but acceptable (and cheap).
> Self discharge of 10% month - may be less. Varies rather with 
> construction and chemistry.
> 
> NiCd, NimH have excessive self discharge rates (typically 20% & 30% 
> month)
> 
> Lithium Ion. Cost?
> Capacity degradation with time regardless of usage.
> Goodish self discharge.
> 
> Lithium Ion Polymer - better self discharge than Nickel chemistries. 
> Wide voltage swing charged to discharged.
> Price ?
> 
> Other?
> 
> 
> 
>         Russell McMahon
> 
>  
> 
> -- 
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist