Why not just add a capacitor across the output of the charger that is large enough to smooth the ripple? That is after all how power supplies give clean DC power.
If you want to regulate output using an arduino controller consider using a buck boost circuit. James electonics had a nice PDF on the design info at one time so look for that. Using a buck boost with FET devices for the switch and flyback will reduce power input considerably. PWM is very lossey in comparison. If you are considering using this project for multiple inputs you need to leave PWM to less demanding projects.
Most high tec batteries show signs of heat build up when no more chemical reaction takes place. The best chargers use this temp change to signal a full charge. An arduino can keep track of temp changes and make the design simple for a smart charger that would excell over any other design you may come up with.
You can also get the atmel AVR programmer from digikey and get unprogrammed IC's and build a board without any outside components. This move could make the controller a 5 dollar project plus power devices.
What I just discribed was my MPPT solar project. It can be done.
Kevin
Mark Stafford here agreeing with OakBox about the "very, very slow pulses" to finish charge lead-acid batteries. Define very very slow: 1.5 seconds between voltage spikes. This lets the nascent hydrogen or oxygen bubble dissipate instead of grow.
I can't document this 1.5 second number; I only recall it being between one and two seconds from previous info hunting. You could experiment with a fully charged FLA, then vary the pulse frequency to avoid bubbles over time. Or observe the behavior of smart chargers prior to float charging: they pulse about every 1.5 seconds.
Or just buy super cheap 10A LA or Li chargers from batteryspace.com. If they are not conformal coated already, you could disassemble and spray/paint/dip the boards for better moisture resistance.
Mark Stafford
--- In electricboats@yahoogroups.com, oak <oak_box@...> wrote:
>
> How are you stepping UP your voltage?
>
> Also - be careful about the PWM. Â If it were very, very slow pulses, then that might be ok. Â But if you're using fast pulses (in the sense of pwm), you could have problems. Â The following couple of paragraphs were in the BatteryUniversity.com web site under Lead Acid Charging:
>
> Ripple voltage imposed on the voltage of large stationary batteries also causes a problem. The voltage peak constitutes an overcharge, causing hydrogen evolution, while the valleys induce a brief discharge that creates a starved state that results in electrolyte depletion. Manufacturers typically limit the ripple to five percent, or 5A for a 100Ah battery.
> Much has been said about pulse charging of lead acid batteries. There are apparent advantages in reducing sulfation; however, manufacturers and service technicians are divided on the benefits, and the results are inconclusive. If sulfation could be measured with accuracy and the pulses applied as a corrective service, then the remedy could be beneficial. Assumptions without knowing the underlying results can be harmful.
>
>
> ________________________________
> From: "dominic.amann@..." <dominic.amann@...>
> To: electricboats@yahoogroups.com
> Sent: Tuesday, October 29, 2013 2:26 PM
> Subject: [Electric Boats] RE: Battery Charging Basics - homebrew chargers
>
>
>
> Â
> Hi John,
>
> I am embarking on a similar project. I expect to use a PWM source signal to charge the batteries. Ultimately, my goal is to have a super-charger, capable of having several inputs connected simultaneously (solar, alternator/generator, wind, "pushed" electric motor), charging multiple banks of batteries using per-bank profiles, and covering all popular types - FLA, AGM, Lithium. A smart controller such as the Arduino would be at the heart of the system - and it could produce a NMEA message string output for standardized monitoring.
>
> At the moment, I have a basic one-profile (FLA) charger configured that can control a single input between 8 and 18 v (yes, it will "step up") and charge the battery smoothly in the "best" time.
>
> I will open source the arduino code (written in "wiring") once I have properly tested it.
>
> In the end, the connectors will cost more than the MCU.
>
>
>
> ---In electricboats@yahoogroups.com, <electricboats@yahoogroups.com> wrote:
>
>
>
>
> I hope this will be both fun and educational....
>
> I'm trying to build a small, home brew battery charger. Â From what I see, it sounds like the primary difference between FLA, AGM, and Li batteries (from a CHARGING) perspective is the charging profile - notably, the voltage level of the charge.
>
> Granted, if you're trying to do a FAST charge with lots of amps, there are a lot more details to worry about. Â But given the limits of readily available parts, I'm looking at 1-2 amps charging current - so I think I can simplify things a LOT.
>
> It sounds like building a "constant current supply", attaching it to the battery (at 1-2 amps), and monitor the battery till the battery voltage comes up to the prescribed level, then disconnecting the current source.
>
> According to the Battery University (http://batteryuniversity.com/learn/article/charging_the_lead_acid_battery), it looks like I can use a constant current source for the primary charging (stage 1), then maybe connect to a regulated power supply for the topping charge (stage 2), and then use a small microcontroller (Arduino) to monitor the voltage over the next few days / months, and occasionally turn on the "float charge" voltage for a couple of hours. Â (stage 3)
>
> Does this sound reasonable?
>
> For charging Lithium batteries, is there any difference, other than paying a LOT of attention to the voltage levels for the peak charge level?Â
>
> From the BatteryUniversity.com site, it sounds like charging Li batteries to 4.1V per cell, and stopping there will extend the life of the battery bank, and provide a bit more guardband to avoid problems.
>
> Thoughts?
>
> Thanks!
> John
>
| Reply via web post | Reply to sender | Reply to group | Start a New Topic | Messages in this topic (66) |
No comments:
Post a Comment