Re: (ET) DIY Smart Battery Charger

["David Roden" <etpost drmm net>]

On 20 Jun 2015 at 16:44, David Erskine wrote:

> http://archive.siliconchip.com.au/cms/A_102420/article.html

This is pretty cool!  

It's something of a foreign language for me since I'm a hobbyist of a 
certain age (not a EE however).  I cut my teeth in the era when the 12AX7, 
CK722, and 2N107 all coexisted.  So I tend not to think in terms of 
microprocessors where simple linear circuits can do the job.  ;-) 

But I guess these days micros are cheap, and sometimes might be the most 
cost effective way to tackle a problem.  And I suppose today's engineers 
tend to think of them first, just because they're so used to using them.

This design does have some nice additional features.

It looks like you could easily adapt it to a 36 volt charger.  You'd 
probably have to change the values of the 10k and 2.7k resistors.  You 
might 
also have to use higher voltage capacitors, I haven't studied it that 
closely.

I'd add a pot to trim the finishing voltage without having to reprogram 
the 
micro.  But that's because I like having controls I can adjust.  ;-)

Another worthwhile enhancement would be to add a thermistor or other 
temperature sensing device that could be placed between two batteries in 
the 
pack, to provide a temperature compensated finishing voltage.  Could you 
just use that as another input to the processor?

I don't know whether the strategy here of "sleeping" for an hour after 5 
cycles is the right one for any particular battery.  It would depend on 
factors including the available charging current and the battery capacity. 

That said, I look at all that processing power on tap and I think, please, 
give me more flexibility!  Why implement such a simple charging algorithm? 
 
It isn't much more sophisticated than the simple voltage controlled 
approach 
I described in another recent post.  

With an onboard micro, I'd want to be able to sense not just on-charge 
voltage, but also charging current.  I'd want to be able to regulate them 
both too.  That would make possible more sophisticated charging algorithms.

Just some thoughts ...


David Roden - Akron, Ohio, USA

= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = 
Note: mail sent to the "etpost" address will not reach me.  To send 
me a private message, please use the address shown at the bottom
of this page : http://www.evdl.org/help/
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =