Re: (ET) Fully charged pack voltage

["David Roden" <etpost drmm net>]

On 24 Apr 2009 at 10:35, Mike Wallace wrote:

> I think you could buy a rebuilt "electronic timer" for around $50 from
> Lester ... Anybody have any thoughts on this approach to upgrading the
> OEM charger?

I don't think it's as simple as connecting the charger's output to the 
Lestronic timer.  

To see how it's used, you can check out a schematic of a typical Lester 
charger.  I have one for a model 9868 (120 volt / 12 volt) in the EVDL 
library, 

http://www.evdl.org/lib/

or direct link

http://www.evdl.org/docs/lester9868_schematic.pdf

The page is oriented at 90deg so you'll want to print it out (or twist 
your 
neck reading the screen ;-).

I haven't worked on a GE charger as I don't use mine, so I don't know 
whether it implements the hardware that the Lester "black box" wants.  For 
example, at a glance, I noticed that the Lester timer controls a DC relay 
on 
the transformer's primary, so you'd at least have to add something like 
that.

Normally you wouldn't use two charge controllers with a charger as they'd 
more or less disagree over when and how to charge the battery, so I 
wouldn't 
recomment using both a Lester timer and one of Harry Landis's controllers, 
or any other pairing for that matter.

If you want a cheap and fairly easy to build charge controller, see this 
diagram :

http://www.evdl.org/docs/c_car_charger.jpg

This is a simple adjustable cycle-dropping charger design (literally right 
out of an engineering handbook).  

I'm not a EE by any means, but I've worked on these chargers.  Most 
component values don't appear to be very critical.  L1 is a small 12 volt 
pilot lamp. On the C-car it was one of the dash indicators.  You can (and 
I 
did) replace it with a resistor (sorry, I don't recall the value, but it's 
easy to determine empirically) and an LED plus series resistor.  

CR3 is a generic SCR, chosen for the current and voltage the charger can 
produce.  

SCR is a small part and again almost anything that can handle 100-200 
volts 
should work.  I think when I replaced this on mine I used something right 
off the hook at the Radio Schlock around the corner from my house.

CR5 is a generic small silicon diode; IIRC a 1 amp 400v should be plenty.  

CR6 is a 12.6 volt zener diode.  Something close to that should work; you 
might have to tinker with the values of R4 and R6. 

R5 is a 5K pot which adjusts the voltage limit.   

You'll probably have to tweak the values of R4 and R6 to use it for a 36 
volt battery instead of a 48 volt one.

This circuit charges at a rate limited only by the charger's wiring and 
transformer, until it reaches an adjustable voltage threshold.  Then it 
begins cycling on and off.  As the battery reaches full charge, the on 
cycles get shorter and the off cycles get longer.  When the on cycles are 
a 
second or less long, and separated by a few tens of seconds, the battery 
is 
essentially fully charged.

This is the same simple algorithm used by probably millions of $50 NAPA 
battery chargers -- and, incidentally, by Rudman Regulators, for those of 
you who are road EV followers.  It's essentially a faster taper charger 
with 
no temperature compensation or any other deluxe features, but it's cheap 
and 
if properly adjusted is more or less adequate for flooded batteries.

Longtime readers here know that Harry Landis and I don't agree on what 
constitutes an appropriate charge algorithm for a battery used cyclically. 
 
His is somewhat unorthodox.  However, he has several advocates here, and 
I'm 
sure they'll be more than happy to argue in his favor.  ;-)

Hope this helps.


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/
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =