Searching \ for '[EE]: Swiching current regulator for intelligent N' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/power/regulators.htm?key=current
Search entire site for: 'Swiching current regulator for intelligent N'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Swiching current regulator for intelligent N'
2001\01\18@213611 by Peter L.

flavicon
face
I want to throw together a fast charger that charges up to 8 D-size cells
_independently_. I understand there are plenty of charger chips available
(such as the MAX715 if I recall correctly) but they seem to be targeted
towards the recharging of battery packs. My instict tells me it's not a good
thing to connect cells at different levels of depletion into a series chain
and charging them together.

To fast charge a 4000mAh D-size cell in 5 hrs would require a current draw
of 1A. to Take advantage of the 'super-fast-charge' facility of my Sanyo
Cadnica cells would require a whooping current draw of 6A for 1-1.5 hrs. Now
let's multiply that by 8...

Say I were to be content with a 2A charge current per cell, that would still
require a 2x8=16A power supply.

Here are my options:

1.  Transformer+ a 2A linear current source for each cell - 16A @ 3V-5V
transformers are
   not easy to come by. Furthermore, such a combination would be awfully
inefficient, not to
   mention the need for _wheels_.

2.  PC power supply + linear current regs - Any PC power supply should be
able to put
   out 16A with ease. But the problem of inefficiency excess heat
dissipation still remains.

3.  Transformer + switching current regulators for each cell - on the
surface this doesn't
   seem to be much of an improvement over option 1. Then I realized
otherwise. If I used a
   transformer of a modest output current (eg. 3A) and a much higher
secondary voltage
   than required eg. 20V , I can actually distribute it's output power
amongst the 8 cells to
   be charged. Each cell would be connected to a buck converter. The crux
of this concept is
   that the PWM converters won't be freerunning. Rather, they are
synchronised and stepped,
   so that only one branch is drawing off the transformer at one time.
Consequently the
   maximum duty cycle for each branch cannot exceed 12.5%. In practice a
feedback loop would
   keep the duty cycle marginally below that, as such a level as required
to maintain a
   constant 2A into the each battery.

4.  PC power supply +switching curr reg.-Same idea as #3. But why bother
with a hefty PC power
   supply when a handy 24V-3A transformer would do the job? On the other
hand, there's enough
   juice for a 8 freerunning PMW's. So, no need for delicate power sharing.

5.  Direct step down from 240V mains to 8x2A outputs -I won't even go into
this. If I had the
   right education, and were being hired $100/hr, this would be the
ultimate
   solution. Unfortuately I don't have the knowhow to afford such elegance.

How would you go about this if you were in my position?


Thanks

Pete

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\01\18@215928 by Barry Gershenfeld

picon face
>My instinct tells me it's not a good
>thing to connect cells at different levels of depletion into a series chain
>and charging them together.
  ...
>3.  Transformer + switching current regulators for each cell -
> If I used a transformer of a modest output current (eg. 3A)
> and a much higher secondary voltage than required eg. 20V ,
> I can actually distribute its output power amongst the 8
> cells to be charged.

Idea number 3 is the kind of thing that crossed my mind as well,
because otherwise you're trying to shave everything down to
the 1.5 volt range and then run really big currents, when you
had enough wattage when you started at the higher voltage.

But...maybe now you see why they like to charge in series.  I would
suggest rethinking your premise.   You can charge at 1/10 capacity
without damage indefinitely.  And if you can tell what condition
the cells are in you can go much higher.

How about taking the trouble to _measure_ each cell independently?
You could do all kinds of tricks with differential amplifiers and
it _still_ wouldn't be as painful as trying to route 6 or 48 amps
around, because now you're measuring voltages.  I think if you
ran a tap (a wire) to each cell you could measure the voltage
at each with respect to ground, and then just use arithmetic
to get the voltage of each particular cell, and skip all that
differential hardware.

Just watch the voltage of each cell and reduce the charge if
any are too low (discharged or shorted) or too high (charged
or open).   If any one cell is forcing you to keep your charge
rate too low you probably want to change it out for a good
one anyway.

I'll leave it to someone else to suggest putting temperature
sensors on each cell...

I think the fact that you're willing to monitor each cell somehow
gives you an advantage the other guys didn't want to bother with.
So you ought to be able to make a pretty good charger.  I've
seen laptop batteries with lotsa wires and I think this is
what they were up to.

Barry

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\01\18@230742 by Bob Ammerman

picon face
How about building a module that passes a current (up to say 2A) by feeding
some of it thru a battery and the rest thru a FET or BJT in parallel with
the battery.

Each module would actually contain the bypass FET/BJT in parallel to (a
small resistor in series with the battery). You would read the voltage on
both sides of the resistor to compute the current thru the cell. You would
read the voltages from two consecutive cells to compute the voltage across a
cell.

Put 8 such modules in series and drive it off a single variable voltage
linear or switch mode power supply.

This will allow you to determine the voltage of each cell, and the current
through each cell (by computing the voltage across the appropriate series
resistor).

You can even bypass modules that contain no cell by just cranking up the
FET/BJT a little more.

Basically you adjust the parallel FET/BJTs to control each cell's charge
rate, and use the main power supply voltage to set the appropriate voltage
across the entire chain.

Sounds like a perfect job for a PIC.

Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\01\19@093534 by Mark Skeels

picon face
Actually, this is exactly what our TurboMatcher4 does for 4 C cells. they
are all charged in series and measured and peak detected independently. a
mosfet then bypasses each as they peak and reach full charge.

Mark

{Original Message removed}

More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...