In article <Kfktj.9$wK4.1@trndny01>, ehsjr@bellatlantic.net says...
> James Beck wrote:
> > In article <ddac6f54-dd64-45ac-bd54-
> > 3f43655f468d@i7g2000prf.googlegroups.com>, redbelly98@yahoo.com says...
> >
> >>On Feb 14, 10:29 am, James Beck <j...@reallykillersystems.com> wrote:
> >>
> >>>In article <4b78e703-68e3-4822-b318-a1cd9fb8ff04
> >>>@e6g2000prf.googlegroups.com>, redbell...@yahoo.com says...
> >>>
> >>>
> >>>
> >>>
> >>>>On Feb 13, 9:09 am, James Beck <j...@reallykillersystems.com> wrote:
> >>>>
> >>>>>In article <6194982a-5fc6-406e-b08d-dd15013ba985
> >>>>>@y5g2000hsf.googlegroups.com>, redbell...@yahoo.com says...
> >>>
> >>>>>>>Why is your simple charger so complicated? Why not use a 12 volt DC
> >>>>>>>wall transformer and 13 ohm resistor (5 watt)? You get 540mA when the
> >>>>>>>battery is low at 5 volts, and about 400mA as the battery voltage
> >>>>>>>rises to 7 volts.
> >>>
> >>>>>>>-Bill
> >>>
> >>>>>>If you forget to turn off a charger like that, it will seriously
> >>>>>>overcharge a 6V battery! On the other hand Ed's charger will not do
> >>>>>>that
> >>>
> >>>>>It won't?
> >>>>>It should continue to charge the battery up to the wall wart voltage,
> >>>>>which is still too high if you leave it plugged in too long.
> >>>>>I would, and do, just use a CV source that is set to the float voltage
> >>>>>of the battery. Pick a regulator that has over temp and current
> >>>>>limiting and let it float.
> >>>
> >>>>> Jim
> >>>
> >>>>What about all the voltage drops between the wall wart and battery:
> >>>
> >>>>1 to 1.5V drop-out voltage of regulator
> >>>>1.2-1.3V between regulator "out" and "adj" pins (across 2.5R resistor)
> >>>>0.6-0.7V diode drop across 1N400x
> >>>
> >>>>Mark
> >>>
> >>>Build it and try.
> >>>As the current drops I'll bet you don't get the drops you think,
> >>>especially across the resistor used for the current sense. You are also
> >>>assuming that the 9V wall wart is regulated. A cheap unregulated wall
> >>>wart that is a "nominal" 9V under X% of load will usually be quite a bit
> >>>higher than you expect as the load drops. A CV float charger could be
> >>>used with any wall wart that is the V drop of the regulator or higher.
> >>
> >>You make good points, I had been thinking simplistically about the
> >>voltage drops at 500 mA.
> >>
> >>Mark
> >>
> >
> > Mind you, you can have problems with a simple float charger too.
> > If you rely on the regulator's current and thermal limit to hold the
> > current under a certain level it will get HOT. I also had a National
> > part that the info in the datasheet didn't match real life and we blew
> > out a few wall warts before I caught it.
> >
> > Jim
>
> I'm amazed at this thread. The "very simple charger"
> I diagrammed *stinks* if _used_ as a float charger, which
> is what you were de facto discussing when you talked about
> forgetting and leaving the batteries on it too long.
> If used that way, it *might* prevent damage to the
> batteries by the cumulative voltage drop which you and
> Mark have discussed, but it is the *wrong* tool for
> that job, and the wrong usage of the tool.
>
> I am glad to see that you are taking the discussion to a
> float charger. I'm responding below because you mentioned
> relying on the chip's thermasl & current limits and that
> some wall warts blew.
>
> It would be poor practice to design a float charger that
> relied on the regulator's current and thermal limit to
> hold the current under a certain level, assuming by
> "the regulator's" you mean the IC chip. Those things - the
> current and thermal limit - only indirectly hold the
> current under a certain level. They are design maximums
> for the chip, not for whatever load the chip is feeding.

When the datasheet and app note show the part being used in this very
way, I would assume the manufacturer has rated the part as such.
Wouldn't you?

>
> The circuit design needs to keep the current under the
> maximum rating of the chip under worst case conditions.
> The designer specifies a heat sink and/or a design that
> keeps the chip temperature below the maximum spec. He/she
> needs to ensure that any other limitations (eg Vin-Vout
> rating) for the chip are adhered to.
The IC had a built in current limit and, like I mentioned above,
National had several design examples that used over current protection
as an integral part of the design. The problem was that the limit was
wrong on the datasheet and in the app note. 1 - 1.2A was suppose to be
min to max range, it turned out to be 1.9 - 2.2A or some such.

>
> A float charger may not need current limiting for normal
> conditions, but it does for worst case: a shorted battery.
> That's where a float charger without current limiting fails.
> Under normal conditions, the battery will limit the current
> drawn as the battery voltage increases, and additional
> limiting may not be required. Still, you need to consider
> the whole circuit. You mentioned that you had some blown wall
> warts. If the batteries require more than the wall wart can
> deliver, that may be a specification rather than circuit
> problem. (ie use a bigger wall wart) If they blew because the
> circuit relied on the chip to shut down when it got too hot,
> that's a design issue. You indicated an error in the National
> datasheet was the cause - do you still have the details? It
> could be helpful to know which part and what spec was wrong.
>
> Ed
>

The part was a National LM2941CT and the wall warts blew because the
current limit of the IC was wrong on the datasheet. I didn't want to go
to a bigger wall wart, we didn't want to charge at a greater rate
anyway. I'm sure (hope) they have corrected the datasheets by now, but
the funny thing was one of the applications engineers sent me a "fix"
for this problem that entailed a small sense resistor and a transistor
to implement a half assed current regulator and they had that wrong too.
It was set up to pull the feedback pin lower as the current increased,
that would increase the voltage out as the current demand increased
causing the system to slam hard against the + rail. I think they were
relying on the internal current limiting to do its' thing at the level
the datasheet had listed erroneously. I had already fixed the problem
myself. The point was I didn't want to add complexity to the design,
that was why I chose the part in the first place, but I ended up adding
parts anyway.

As a side note, they have changed the datasheet, but there is still one
hold over. In the application notes they still show a "1A Coil Driver"
or some such, that is suppose to limit the current to 1A using the
internal current regulation of the part. Well, it won't do it.

Jim