Simple Current Limiting Battery Charger

[pigman]

Hey, I am working on a battery charger circuit for SLA small ~1.3Ah 12v. It is part of a greater circuit and will be triggered from a PIC program that checks the voltage of the battery and triggers a charge cycle, breaking it regularly to check on its status.

My voltage source is regulated 13.5v and this is the correct voltage I'd like to charge to, however my concern is charge current would need to be limited to ~300mA or something to prevent battery damage, but every circuit idea I find has some fancy regulating charger IC or similar, and I already have a regulated correct voltage. Just trying to find a simple Current restricted Circuit idea that I can trigger using a PIC to switch a FET and charge up the battery?

Thanks in advance!

 

[ben7]

You can also use a simple LM317 current limiter, check the web, there are tons of those simple circuits posted!

-Ben

 

[pigman]

Ben, Thanks mate!

But wouldn't this not work considering these style units require an input that is higher than its output. 2v or something drop accross a regulator like this one? Meaning my 13.5v input wouldn't be able to output (with a LM317) enough to charge the Battery to 13.5v? It sucks Because my Power supply is about 13.5v and I want the batter at 13.5v but don't want to just connect the two together without limiting the current to the advised amount for the battery at hand...

 

[MikeMl]

Your charge voltage is way too low. Printed on the side of most SLAs is "Cycle Charge at 14.7V, Float at 13.7V". It will take forever to reach even 60% charge at 13.5V.

You need a current-limited full-wave rectified output wall-wart which has an open-circuit voltage of 15 to 18V, but is current-limited to less than 0.5A (more like a constant-current supply, less like a voltage-regulated supply). A trick that works for me is to find the Wall-Wart with the right output voltage, and then connect a 24V 10W lamp in-series with the 120V input side of the wall wart (I break open the plastic case of the wall-wart, and repackage it into my own enclosure). The lamp acts as a "ballast", or crude current regulator. The non-linear filament resistance is what accomplishes that.

I use the A/D inside the PIC to monitor the battery voltage. When the battery needs charging, the charging cycle consists of turning on a high-side PFET switch which connects the wall-wart output to the battery positive pole. Since the wall-wart is current limited, the PFET is just a switch, fully on or fully off, so dissipates almost nothing. (Heatsinking not required).

The PIC watches the battery terminal voltage during charging. When the battery voltage reaches 14.7V, it keeps the charger on another 15min (regardless of how high the voltage climbs during that 15min period), and then switches off the PFET.

The PIC continues watching the battery voltage, and turns the charger back on if the battery voltage sags below 13.3V. This time, the charger turns off upon reaching 13.8V. This cycle repeats until the battery is removed. When the battery is disconnected/re-connected, the PIC goes thru the initial charge cycle again, including charging for 15min past reaching 14.7V.

If you are in a 240V country, you may have to find a different lamp to use as a ballast ahead of the wall-wart. Maybe a 120V lamp (if you can find one)? You know you have a lamp with the right filament resistance when it glows about half brightness early in the charge cycle, and then glows very dim toward the end of the charge cycle.

 

[pigman]

Mike, interesting way of doing things! I like it!

My application however wouldn't really warrant it, the device is to be located on a boat and I'm pretty limited at the moment as far as voltage goes. I suppose I could get a 24v PSU and then use the LM317 ? ? As you know from my other thread this is for the same project. This part is the charge portion of the circuit and then I'd need to do a non current limited regulated output for when the load is on the battery so the PSU can hold the voltage across the battery and the load. I chose 13.5 as it covers a range of battery temperatures and is the low end of the "float charge" range on the battery I've chosen.

Taking this into consideration, whats your recommendations?

 

[MikeMl]

13.5V will not recharge an SLA battery properly; end of story! It is almost too low to float the SLA, assuming it 100% charged to begin with.

If you are charging the SLA in a car (14.2 to 14.4V with the engine running above fast idle), you stand a better chance. If you are doing this where there is 240VAC, then get a second AC supply for charging.

 

[pigman]

Mike, just looking at it again, I could theoretically use your idea there but replace the current limiting portion with the LM317? See attached? Obviously powering from a 24vDC source.

 

[MikeMl]

I don't like that LM317 circuit. It There is **broken link removed** that uses the battery charge current to create a voltage drop across ~1 Ohm resistor to turn on the Vbe of a small NPN transistor. The collector of the NPN pulls on the LM317 Ref pin. It has a much better I/V characteristic for battery charging than the one you posted.

 

[pigman]

Like the one attached to this post?

I'm getting them from https://www.sparkfun.com/datasheets/Components/LM317.pdf

Just noticed your update. Checking it out now.

 

[pigman]

So this one here would be a winner? I guess I would have to use a 24vdc PSU and regulate it down to 16-18v or something like that? And have a separate regulator at ~14v for when it is not in charge mode and simply powering the device across the battery?

I guess now I just have to figure out what the resistors and whatnot are

 

[MikeMl]

As long as there is a PIC in the loop, you really do not need the voltage regulation aspect of the LM317; the PIC takes care of the voltage.
You need only the current-limiting, which is what the ballast lamp does in my circuit. The ballast lamp works better in the transformer primary, than in the secondary. The charger does not need filter capacitors or filtering; only a transformer and full-wave bridge.

 

[pigman]

Mike,

Can you see any way of doing it without the ballast lamp? I just can't really put a lamp inside my device?

 

[MikeMl]

Mike,

Can you see any way of doing it without the ballast lamp? I just can't really put a lamp inside my device?

Can you tolerate some heat? I'm thinking an inrush current limiter.

 

[pigman]

Mike, I guess that depends on how much heat . It is all in an enclosed IP65 enclosure so it is hose proof to a degree so I have temperature switches inside on various components to ensure nothing overheats in there. I would consider a small amount of heat ok, but too much is not good The battery is a 1.3Ah and only requires about 300mA to charge at a good rate, I just want to limit it to 300mA max during charge at whatever the final voltage is 13.5-14.2 or similar.

 

[MikeMl]

Well, that is a problem. Think of it like this: If you start with a fixed voltage, say 16V, then the current regulator will have enough compliance to get the battery up to 14.7V, and still have enough left over for some drop across the regulator (whatever it is, lamp or solid-state device). The heat problem rears its head when you walk up with a mostly discharged battery at say 11.5V. When you first start charging, the drop across the regulator element is 16-11.5 = 6.5V. At 0.3A, that is ~2W, so that is where the heat comes from. 2W of dissipation in a small closed box is manageable, but think about what would happen if you start with 24V?

The alternative would be to go with a Switch Mode Power Supply, configured as a constant-current step-down regulator. These typically have efficiencies of ~85%, so the heat inside the box would be reduced to a manageable level even starting from 24V. I guess this raises the possibility of using a step-up current regulator that starts from the original 13.5V?

I wonder if you could use one of the ~300mA BuckPucks normally used for Big LEDs as a battery charger?



The

 

[pigman]

Mike,

Your one clever dude. I just went to the local electro store and purchased a small 19vdc SMPS (I have been using 12v wound up to 13.5v, but they don't go any higher) this 19vdc unit is rated 5A and from experience barely generates any heat when low load is applied. This would mean a nominal ~2.5W of heat dissipation on the current limiting device? I may be wrong here, just trying to get my head around it all . I had thought of the step down heat dissipation issue so I went in between and got the 19v. if only there was a similar 15v one.

I had seen those BuckPuck style devices before, and I don't see why that wouldn't work? Are you suggesting that an appropriately rated buckpuck style device would suit the application without the requirement for the current limiting circuit, etc?

 

[MikeMl]

A Buck Puck is a current regulator, trim pot adjustable.

 

[pigman]

So that would theoretically do the job? a BuckPuck in the loop and feed it 14v or similar? Would they continue to force the 300mA into the battery after its charged? I guess that's not really an issue considering the PIC will be checking the battery voltage and disconnecting it when it reaches the "charged" point?

 

[MikeMl]

I haven't looked to see if Buck Puck can boost the input voltage. The name "Buck" implies they do not.

Here is a previous thread where we beat SMPS current regulators to death.

 

[pigman]

hey, I was just messing with this circuit **broken link removed**. I'm feeding it 19v and have it configured exactly as the diagram is minus the 25v 2200 cap at the beginning and obviously a SMPS infront of it instead of that linear rectified PSU.

And I can't get more than 10v out of it. 5k pot, even tried winding up the resistance on the pot with a 20k pot and nothing more than 10v. What could be causing this? Maybe no load? I just have an led on the output so I can see it on and then the multimeter to check the voltage as the text says. Any ideas?