[Noisebridge-discuss] An introduction, and a plea for help with electrical systems

[jim]

no one seems to mind, so comments interspersed below continued... 

On Tue, 2008-12-23 at 16:52 -0800, Christie Dudley wrote:
>         comments interspersed below.
> 
> So I'll do the same! :)
> 
>  
>         On Tue, 2008-12-23 at 15:04 -0800, Christie Dudley wrote:
>         > batteries to run them.  While the Luxeon packaging is pretty
>         > attractive for fixed installations, it's not something I
>         recommend an
>         > artist drop that kind of money on.
>         
>         
>         what is the purpose of the luxeon devices?
>         what is meant by "display light"?
> 
> Luxeons are just high-end "name brand" LEDs.  "The brightest
> available"

   Thanks. they had been called "flat panel" and "display", 
so it wasn't clear what was meant. 


>         one of the possible downsides
>         of buying from scrapheaps and other such (e.g. ebay)
>         is that you can't get enough of a particular thing
>         to make a design cost-effective: 
> 
> You obviously haven't shopped for LEDs on Ebay!  I've found exactly
> the same part from exactly the same supplier every time I've gone to
> look for more.  It's not a "scrap heap" at all, so much as a channel
> for getting stuff out of China.  I'd encourage you to check it out
> before disparaging it!

   i didn't mean to disparage. "scrap heaps" such as the 
now-gone mike quinn's in san leandro are great places 
in my mind. and you're right, i haven't used ebay at 
all, figured it was onesy-twoesy stuff from individuals. 


>           also, what's the reflectivity of the concrete
>         (or whatever material used for the columns)? is
>         it possible to add pigments and/or reflective
>         sediments to improve the reflectivity of the
>         columns?
> 
> Oooh!  Neat idea! I know someone who added a pearl pigment to a
> concrete finish to give it a fancy glow.  It takes some skill to get
> right, though.
> 
> 
>           great idea! even though this runs counter to the
>         don't-buy-onesy-twoesy ideas (above), it works
>         because getting lots of 12VDC batteries is easy and
>         it doesn't matter their capacity, just that their
>         form-factor is within the space limits.
> 
> Yeah.  I imagine you could use lantern batteries as well.  If
> different columns have different power needs, then we can probably
> work around that.

   any battery that can be recharged, it seems to me, 
will work, maybe with a little (circuitry) help. 


>         > Although I'd be a bit concerned about fitting them inside a
>         5" column.
>         > They really should sit flat.  You could potentially use
>         smaller
>         > batteries if you got the power requirements down, but that's
>         really
>         > dependent on how much illumination you'll need, and how it's
>         > distributed.  Lighting design probably should be something
>         that
>         > involves sitting down with drawings, etc. etc.
>         
>         
>           more good points. batteries should be upright.
>         there are sealed batteries, but although they
>         don't (normally) allow seepage, their internals
>         are essentially the same as unsealed batteries:
>         the chemical solution is in a cell that should
>         be normally oriented (upright).
>           again, you can stack multiple small batteries
>         within the columns' internal cavities, and the
>         batteries can be different sizes as long as
>         they're all the same voltage.
> 
> eeek!  That concerns me.  I wouldn't "mix and match" battery types.  I
> guess you're a lot more confident with dealing with problems of uneven
> drainage and issues as one wears down and others in the array are
> still going.  I'm assuming you've dealt with this a lot in the past?
> I'd be interested in talking with you about the circuitry involved.
> (Power diodes leading into the array, etc. etc.?)  

   not a lot, but some. all batteries must present the 
same voltage, best they all have the same chemistry, 
but (again with help) not necessary. 
   the trick is to provide each battery with not too 
much voltage/current during charging. the simple safe 
way to manage is to ensure that the charging current 
is small ("trickle charging"). 
   my (old) design used op amps (i love the 307) and 
transistors (some reasonable NPN) and some resistors 
around to set voltages and currents. that was long 
ago, but it was simple and never failed and can be 
re-thought and made to work for whatever rechargeable 
batteries come along. 


> At this point I'm assuming these are going to be charged elsewhere.  A
> puny solar panel could be used for each column, but how much could you
> really get out of that?  Would it be worth the cost?  Convenience
> suggests the fewer number of batteries per column, the easier it is to
> swap them out.  (And the higher amp-hr/lb., the less to lug about.)  

   even a puny solar panel might do the trick. depends 
on the battery capacity and power requirements. for 
example, a big battery with lots of charge will not 
get much more charge during the day (small, chintsy 
solar panel) but that little charge may be enough to 
raise the battery's charge to make it through the next 
day. at some point the battery swappers will have to 
come along.... 


>         also, if batteries are used, you should know
>         the temperature extremes, both hot and cold,
>         in which they'll be working. that will have a
>         bearing on power draw and recharging.
> 
> Yeah.  Here's where experience counts for a lot on playa.  I really
> wouldn't go with anything that some other project in the past hasn't
> used. The playa is a harsh environment, and this matters. 

   battery manufacturers provide temperature stats. it's 
a matter of getting the times and temperatures for 
typical days and then applying some arithmetic to the 
expected power and recharging times. not a biggy, but 
prudent to do, possibly can rule in or out some 
particular batteries. 


>           seems the most important data point to nail
>         down is the luminosity: how lit do the columns
>         need to be to be seen from how far away? from
>         that it seems you can figure power requirements.
> 
> Yep.  Lighting design must be done before the power can be calculated.
> 
> Christie
> 
> -- 
> You can't learn what you think you already know.