SLA charger output circuit

[DTJ]

I've put together a SEPIC charging circuit for a 12V SLA. See below.

The end user could reverse connect the battery and damage the output capacitor so I think some form of blocking diode is needed.

Any thoughts on the sketch as a possible solution.

Maybe I could keep the feedback divider across the capacitor and simply jack the programmed voltage by 0.4V or so to compensate for the loss across the diode.


I'll probably end up with a polyswitch or fuse in series with the output too.

[Rerouter]

for 22uF, get a bidirectional, or ceramic, so polarity doesn't matter, then fit a schotkey to your resistive divider center point from ground if the fault current can exceed the pins rating

Other thoughts are make sure your switching diode is rated for at least 50V (preferably 200V), (load dump in a battery is a horrible thing)

[ZeTeX]

I've put together a SEPIC charging circuit for a 12V SLA. See below.

The end user could reverse connect the battery and damage the output capacitor so I think some form of blocking diode is needed.

Any thoughts on the sketch as a possible solution.

Maybe I could keep the feedback divider across the capacitor and simply jack the programmed voltage by 0.4V or so to compensate for the loss across the diode.


I'll probably end up with a polyswitch or fuse in series with the output too.

A big flyback diode at the output would not hurt, if you put a reverse voltage at the output most of the current will flow through the diode..

[DTJ]

Rerouter, the cap on the output is pretty much locked down in the data sheet to a very low ESR tantalum (although a 20 year old crappy aluminium electrolytic worked in the first lash up).


I'll have a think about the diode on the divider. Thanks.



ZeTeX, that's not a bad idea, coupled with a fuse on the output to the battery it would do the job. Thanks.

[Rerouter]

I will make this very clear, do not use a tantalum directly connected to a car battery, you will get fire and doom,

So ceramic then, (very very low ESR) with an electro in parallel to soak up the spikes caused by the batteries spike loads and other weird stuff they do as bridges form and dissolve,

Lets say your charging at 0.5A, over i'd say 2 meter long leads, when the battery is disconnected, or it does some of the weirdness lead acids are known for, you will have to soak up the spike the inductance of the leads has, so if you want to be cautious, make your output electo closer to 47uF, with a 22uF ceramic,

[MagicSmoker]

...
The end user could reverse connect the battery and damage the output capacitor so I think some form of blocking diode is needed.
...

D1 is also subject to destruction from a reversed battery connection, and inserting a blocking diode after C1 won't help protect either C1 or D1.

A beefy-enough diode in anti-parallel with the output will divert current from a reversed battery for long enough to protect D1 and C1 (assuming the user reacts appropriately to the big spark that will undoubtedly occur when the battery is connected wrong), but it would almost certainly cost more than those two components so that's not a terribly practical solution, nor is it guaranteed to work.

A polyswitch or other type of PTC fuse is not guaranteed to work in this application, either, because it will let through a rather large current for an uncomfortably long time before it "trips".

Thus, a regular fast-blow automotive fuse in series with the output is likely the best solution, just make sure the I²t rating of D1 exceeds that of the fuse by a good margin so that D1 doesn't bravely sacrifice itself to protect the fuse!