charging an SLA with a switch mode DC-DC converter for an alarm system

[Shadow351]

Hello all, I am working on replacing the dead alarm system in my house. I am going to build one myself but I wanted to clarify some details for charging the SLA backup battery. I saw a post somewhere (not sure where at this point in time) somebody said they were charging a SLA by simply hooking up a switch mode dc-dc converter with the current limit set at "10% of the capacity" (700mA for a 7Ah battery) and the voltage set at 14V. The battery I have is also a 12V 7Ah SLA, but on the side it says "Standby Use, Voltage Regulation 13.5-13.8V, Initial current less than 2.1A, Cycle Use Voltage Regulation 14.5-14.9V, Initial Current less than 2.1A" (datasheet says the same) Since I assume my situation would be considered standby use, If I set my dc-dc converter to 14.5V(13.7V + 0.8V drop across a diode between the regulator and battery) and current limit it to 1.5A (to extend battery life), I shouldn't have any problem right? Is more then the 10% just a suggestion or am I misinterpreting this statement/datasheet?

I plan to have the Arduino monitor the battery voltage (through a buffered voltage divider) and current (with a hall effect sensor) and disconnect the battery if the voltage drops too low while running off of it. I attached the schematic I put together for my alarm system (I apologize as it is a little messy), any other thoughts about it? 1N5400G diodes should do the trick right? Or do I need schottkey diodes? (someone suggested in another post)

Thanks
-Brad

[Ian.M]

The diode drop wont be 0.7V at low currents.  Schottky diodes have a lower Vf so make it easier to stay within the standby float voltage range as the drop will vary less with load current. However, if you start with a fully charged battery that's been rested (no load) for 24H, all you need to do is to slowly increase the output voltage until the battery terminal voltage is in the float range, give it some time to settle then do a final trim the next day under normal load.

C/10 is a conservative safe charging rate that is appropriate for any type of Lead acid battery in cyclic use.  Some batteries are specced for faster charging, yours at up to 2.1A so a C/5 rate (1.4A) or even slightly higher would be acceptable.  Unless your location suffers frequent power cuts, it will have no effect on battery life as it will rarely be discharged far enough to hit the current limit.

P.S. your battery disconnect wont (check MOSFET body diode direction), and there is no need for battery current monitoring in a standby use application, unless you add a load tester and a charging cutoff so you can take the battery offline at monthly intervals and do a load test to check it hasn't deteriorated.   

[BradC]

Do it the way alarm systems have been doing it since time began.

Power the whole thing off a 13.7V supply. Connect the battery in parallel with the power supply with a diode in the +ve leg, so if the supply dies the battery powers the unit minus whatever the diode drop is. Put a 10W 12V festoon bulb in parallel with the diode to charge the battery. Simple, cheap and effective. As a bonus when the battery starts to fail and internally shorts, the lamp will stop it cooking completely and will illuminate to show a nice glow from all the orifices in the box.

It was always easy to tell when a battery replacement was required as the boxes lit up like a christmas tree before you even opened the door.

As you won't be putting temperature compensation in your charger, aim for the lower end of the float voltage (~13.6V) and you'll see reasonable battery life.

[Shadow351]

The diode drop wont be 0.7V at low currents.  Schottky diodes have a lower Vf so make it easier to stay within the standby float voltage range as the drop will vary less with load current. However, if you start with a fully charged battery that's been rested (no load) for 24H, all you need to do is to slowly increase the output voltage until the battery terminal voltage is in the float range, give it some time to settle then do a final trim the next day under normal load.

Yea, sometimes I forget that diodes are not ideal, I will keep this procedure in mind when I set my voltage.

C/10 is a conservative safe charging rate that is appropriate for any type of Lead acid battery in cyclic use.  Some batteries are specced for faster charging, yours at up to 2.1A so a C/5 rate (1.4A) or even slightly higher would be acceptable.  Unless your location suffers frequent power cuts, it will have no effect on battery life as it will rarely be discharged far enough to hit the current limit. 

Ok, good to know. Power here is pretty stable, I've lived here for about 2 years and only had 2-3 extended power outages.

P.S. your battery disconnect wont (check MOSFET body diode direction), and there is no need for battery current monitoring in a standby use application, unless you add a load tester and a charging cutoff so you can take the battery offline at monthly intervals and do a load test to check it hasn't deteriorated.   

Yep, looks like my symbol is upside down. Normally the source is up because the voltage comes from above, the voltage coming from the bottom of the schematic is throwing me off my game. As far as the current sensor goes, It was still in there from when I was going to be controlling the charging current to the battery, but in my research for how to "Smart" charge a SLA, I found the article I mentioned above.

Do it the way alarm systems have been doing it since time began.

Where's the fun in that? I'm an engineer, it's not broke, it needs more features! 

It was always easy to tell when a battery replacement was required as the boxes lit up like a christmas tree before you even opened the door.

I do plan to have the Arduino send me a text message or email if the battery voltage drops below 10V (as in a bad cell)

As you won't be putting temperature compensation in your charger, aim for the lower end of the float voltage (~13.6V) and you'll see reasonable battery life.

Will do.

Thank you

[Gregg]

If you just regulate the current to 10% of the capacity you are likely to fry your battery.  The proper way to maximize the life of a sealed lead acid battery is to limit the voltage to 2.23 volts per cell except for times when you want to equalize the cells (like after a significant discharge).  I retired as a power engineer from a major telephone, cell phone and data company where batteries are the life savers of the critical systems and have seen many batteries that died of overcharging.  Attached is a manual from a company that makes semi-large sealed batteries; see sections 12 through 14 for a lot of information (the voltages listed are volts per cell, your 12 volt battery has 6 cells). 
For critical systems the batteries and chargers are connected to a main bus and the equipment is powered from that bus via fuses and/or circuit breakers.  That way the batteries are always on line at float voltage until the power goes out, and then the batteries take over supplying current seamlessly.  Just from a quick glance, it seems that your battery part of your system is overly complicated. 
I would suggest that you control the battery voltage with controls to the charger that raise the voltage to 14.5 once each 10 days for an hour and then drop back to 13.38 plus or minus the temperature correction factor ( a number of battery chargers for boats work on this principle).  Also if you can monitor loss of power to the charger that causes the battery to discharge, an equalize voltage will help recharge the battery more quickly.  If you monitor the current charging the battery vs current being used by the system when it is in float state; if it is more than about 100 milliamps you have a problem like a partially shorted cell.

[BradC]

I would suggest that you control the battery voltage with controls to the charger that raise the voltage to 14.5 once each 10 days for an hour and then drop back to 13.38 plus or minus the temperature correction factor ( a number of battery chargers for boats work on this principle).

This is good advice for stationary flooded lead acids that are subject to charge stratification, but won't really extend the life of a floated SLA (and in fact may accelerate plate corrosion). In addition, 2.23V is too low for an SLA cell and will lead to slow sulfation. Go by the manufacturers specs on float voltage.

[Gregg]

I would suggest that you control the battery voltage with controls to the charger that raise the voltage to 14.5 once each 10 days for an hour and then drop back to 13.38 plus or minus the temperature correction factor ( a number of battery chargers for boats work on this principle).

This is good advice for stationary flooded lead acids that are subject to charge stratification, but won't really extend the life of a floated SLA (and in fact may accelerate plate corrosion). In addition, 2.23V is too low for an SLA cell and will lead to slow sulfation. Go by the manufacturers specs on float voltage.

Maybe I didn’t make it clear and I agree that if you have the specific manufacturer’s documentation for float voltage, it is the value that should be used.  The 2.23 VPC I used as an example is indeed the manufacturer’s spec for the batteries in the attached document. 
My first point is that lead acid batteries should not be charged by regulating the current only; an upper voltage limit needs to be set.  The OP appeared to be only considering regulating the current at 10% of the manufacturer’s amp-hour rating.   Over voltage kills lead acid including sealed batteries by overheating them, sometimes to the point of breaking the case.  Excluding external conditions, a slight over voltage will shorten the life of a sealed lead acid battery faster than a similar under voltage in float condition; therefore a slightly higher equalize voltage every 10 days or so will keep sulfation at a minimum; the equalize time and voltage should be determined by the size of the battery to keep it from getting hot.

[Richard Head]

Also bear in mind that the float voltage is temperature dependent so you need to check at what temp the float voltage is specified. Most battery manufacturers state it at 20 deg C but I've also seen other temps specified. Every deg C increase in temp means you have to drop the float voltage by approx 4mV/deg/cell. ie 24mV/deg C for a 12V block. But really, this is total overkill for a 7Ah alarm battery.

[BradC]

Maybe I didn’t make it clear and I agree that if you have the specific manufacturer’s documentation for float voltage, it is the value that should be used.  The 2.23 VPC I used as an example is indeed the manufacturer’s spec for the batteries in the attached document. 
My first point is that lead acid batteries should not be charged by regulating the current only; an upper voltage limit needs to be set.  The OP appeared to be only considering regulating the current at 10% of the manufacturer’s amp-hour rating.   Over voltage kills lead acid including sealed batteries by overheating them, sometimes to the point of breaking the case.  Excluding external conditions, a slight over voltage will shorten the life of a sealed lead acid battery faster than a similar under voltage in float condition; therefore a slightly higher equalize voltage every 10 days or so will keep sulfation at a minimum; the equalize time and voltage should be determined by the size of the battery to keep it from getting hot.

Sorry my mistake. I misinterpreted what you were saying with regard to float voltage and understod it to be a blanket recommendation. Don't disagree with anything you've said. The CC/CV is the easiest methody to charge them, but low draw technology (like alarm systems) get fantastic battery life with a simple float charger using either a resistive or incandescent (variable resistive) current limit. Bateries don't die. Most of them are killed, and a great proportion are killed by over voltage float. Most chargers have no temperature compensation, so the best way to deal with it is set the float voltage as low as is manufacturer recommended for the ambient conditions. It's a balance between killing them with plate corrosion (over voltage) or sulfation (under voltage) as you stated above.

I have some nice 55AH Vision SLAs here with a 10 year design life that are now entering their 13th year of operation. A C/20 cycle test late last year showed they were all still at rated capacity. A controlled temperature environment and properly adjusted float voltages do wonders for battery longevity.

[Ian.M]

If you re-read the O.P.s original post its quite clear that he intends to use a CC-CV DC-DC converter module , and my response discussed how to get the correct CV setpoint with a series blocking diode between the converter output and the battery.

Most such DC-Dc converters have a voltage feedback potential divider to a comparator input with an internal fixed reference on the control chip, so implementing a  switchable output voltage setpoint to support three stage charging is simply a matter of switching an additional resistor in parallel to the bottom resistor of its feedback divider.  Something like a 2N7000 controlled by the MCU, source to gnd, and from its drain to the existing divider tap, a resistor ten times larger than the bottom divider resistor + a trimpot of half the extra resistor value for fine adjustment would do the job.  Set the float voltage normally with the 2N7000 off, then set the higher adsorbtion phase voltage with the 2N7000 on using the new trimpot.   Alternatively you can get really fancy and connect a filtered PWM to the divider tap via a resistor to inject or remove current from that node, allowing the MCU to  trim the output voltage so you can do temperature compensation as well.