1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

LEad acid PWM charger

Discussion in 'Microcontrollers' started by Mosaic, May 26, 2011.

  1. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Hi all:

    I want to build a 4 stage Auto battery charger.
    1) Bulk constant current (say 4-8A) charge to 14.4V
    2) Absorption charge at 15.6V constant voltage maybe 10+ amps
    3) Float charge at 13.2V constant V.
    4) Optional Equalisation charge at 15.6V constant V...maybe 10+ amps.

    PiC based of course.

    I've been looking at this sample
    http://schmidt-walter.eit.h-da.de/smps_e/abw_smps_e.htm

    But this charges an inductor which then discharges across a capacitor. This seems to be the accepted theory. But why can't I simply PWM drive the FET into a smoothing capacitor for basic ripple handling and then into the load. After all the load is a giant capacitor itself.

    If I use a .1 ohm, 2 watt, current sense output resistor the PIC can throttle the PWM to the FET GATE based on an ADC sensed voltage drop to derive constant current.
    Similarly the PIC ADC can sense the output to determine the PWM for a constant voltage. The PIC can then handle short circuit, no connection, and reverse battery voltage connection defense.
     
  2. OlPhart

    OlPhart Member

    Joined:
    Feb 22, 2011
    Messages:
    331
    Likes:
    8
    Location:
    SW Michigan (USA)
    Some things to consider:
    1> I'd be Real Certain before driving 15.6V @ 10+A into a 6 cell lead acid.

    2> Reversed connection is well handled by an inline diode (schottky or not). Sensing the trouble While it's happening is way too interesting and not necessary. Your system can easily make up for the diodes' voltage drop.

    3> I like the ADC reading on both sides of the sense resistor, but you have to rescale the range of interest, both for protection and best resolution. BTW, the inline diode can be the sense resistance, if properly characterized. Good Hunting... <<<)))
     
    • Like Like x 1
  3. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Actually, the absorption charge varies from 14.2 thru 15.6V depending on temperature...but the PIC will deal with that as well. Current is NOT limited here....when it falls to , say, under 2 amps for the avg auto batt of around 40+ AH rating we switch over to a float charge. The intention is to force some gassing and 'penetrate' the chemical reaction deeper into the battery plates, eliminate acid stratification and reverse minor sulfation.

    The reverse condition needs to be handled by the PIC in order to NOT enable charging a reverse connected batt.

    The actual concern is:
    "why can't I simply PWM drive the FET into a smoothing capacitor for basic ripple handling and then into the load. After all the load is a giant capacitor itself."
    Do I really need this inductor, which will have to be a ferrite (energy storage) as opposed to iron powder (choke).
     
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. RCinFLA

    RCinFLA Well-Known Member

    Joined:
    Jul 23, 2009
    Messages:
    1,065
    Likes:
    31
    Location:
    FLA, USA

    Depending on your source you will likely smoke the MOSFET (or the source supply).

    How are you getting from your charging power source (presumably AC mains) to low voltaage? Most common method is a current limited, voltage regulated flyback converter.

    On equalization, you have to make sure that with a large battery that the max current will take it to the equalizing voltage. Otherwise it might just sit at a voltage below 15 vdc and never quit. Same applies to absorb voltage limit.
     
    Last edited: May 26, 2011
    • Like Like x 1
  6. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean

    Why would the FET smoke? It's rated at 195Amps continuous, .0014 ohms ON resistance. It's switched by a MOSFET driver for a sharp switch transition. I plan to keep it's switch rate around 1- 5Khz to minimise miller effect switching delays.

    The source transformer is a custom wound 800W 'microwave' magnetron driver. I recycle it from the microwave, rewind it myself to the required Voltages I need.

    At 20VDC I need around 14.2 VAC minimum. I should see at least 20 Amps even with winding losses. My alternative supply is a hacked PC PSU giving me 16.5VDC regulated off the 12VDC 20Amp rail.

    I imagine some sort of secondary time out on the 'overcharge' as well as the 'bulk' charge phases is a good idea, to not boil a battery that is defective.
     
    Last edited: May 26, 2011
  7. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,869
    Likes:
    144
    Location:
    morristown,tn
    I made a home made charger and set the transformer on fire. It was rated at 50 amps I wanted to fast charge a car battery I was a little dumber back then
    figured the transformer would not put out more then 50 amps. so I set a timer to tell me when to stop charging, all was good till the battery got all most fully
    charged. Then the dang thing went off the scale and started pulling lots of amps. Then came the fire before I could unplug the thing.
     
  8. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    I'd like to share this link with the forum.
    http://www.electro-tech-online.com/custompdfs/2011/05/evs_17paper.pdf

    It appears that mcu controlled OVERCHARGING while measuring for a slowdown of voltage ramping on the battery doubles the life of lead acid cells.
    Another method comprising constant current followed by a thermally governed current interrupt overcharge yielded a 400% longevity benefit.!!
    The currents used are way outside manufacturer recommended charging regimens! But cause the battery to live MUCH longer.
    Perhaps reduced turnover reduces profits, thus the manufacturers aren't motivated to make their batts last longer.
     
    Last edited: May 27, 2011
  9. ericgibbs

    ericgibbs Well-Known Member Most Helpful Member

    Joined:
    Jan 4, 2007
    Messages:
    21,240
    Likes:
    645
    Location:
    Ex Yorks' Hants UK
    ONLINE
    hi Mosaic,
    Interesting article.
     
  10. OlPhart

    OlPhart Member

    Joined:
    Feb 22, 2011
    Messages:
    331
    Likes:
    8
    Location:
    SW Michigan (USA)
    Thank You Mosaic... Much more than interesting. I've Got to build one of these!
    Your last point is well taken, given a 10/00 publish date.
    I stand, not just corrected, but redirected. TNX <<<)))
     
  11. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Since I am doing some transformer rewinding ...8))....I'll share some useful info:

    http://www.tech-diy.com/rewinding_transformers.htm

    I never knew u could calc a transformers power rating by it's physical lamination measurements.

    I also did some desulphation research...so i'm building into the design a unit that can pulse 300Amps, 100uSec pulse width, into a .12 Ohm IR, 13vdc battery, after charging the battery.
     
  12. 3v0

    3v0 Coop Build Coordinator Forum Supporter

    Joined:
    Jul 14, 2006
    Messages:
    9,404
    Likes:
    227
    Location:
    OKLAHOMA USA
    Mosaic. Did you get the battery charger built? Does it work as expected?

    FYI The links in this thread are outdated.
     
  13. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Hi 3v0,
    The project has transitioned into early commercial deployment for durability and user feedback.
    I have a full blown, 13 stage, regenerator system now that can also multistage charge and 'exercise' lead acid batteries including AGM/SLA.
    Product differentiation is fully automated and temp limited battery regeneration up to 250 Ah size units, both 6 & 12V with early battery rejection, full cranking amps and Ah re-rating, and suggested battery application.
    The 9" cube system also queues two batteries to process sequentially ( can be different processes and battery sizes) and offers two float charge ports for completed battery maintenance.
    I am just completing the Global model to match a business model co-developed by the Univ. of Applied Sciences in Graz, Austria. The product is cited as an example of developing world innovation in the entrepreneurial undergrad courses there. It won an award for energy conservation (1 of 6 in the Caribbean) and was cited as the best of the lot.

    The Global model includes an ESP12e IoT module for capturing all battery regeneration science data and managing all system deployments. Downstream all this data will undergo analytics to develop better profiling of battery regeneration across the planet and making product improvements.

    I have processed about 200+ auto/marine batteries personally during the development and deployed all the regenerated units back into OEM service with extremely good results. The average yield of a defective back into OEM service is 25% of all batteries processed. In recycling terms : reducing the Carbon footprint when transporting huge tonnage and then re-smelting lead metal, the potential for energy and climate conservation is apparent.

    Due to feature creep and self diagnostics (for safety) I had to abandon the original BASIC approach and do pure ASM coding in order to not have to change the processor. Of the 8Kb code space, 120 bytes remain in the PIC. The website is under construction now.
    In terms of power handling , the system delivers up to 1000 Amperes of pulse reliably with a custom method of allowing the battery recovery to determine the pulse intensity as a closed loop. 90 sq inches of circuits, 400+ components and two microprocessors.

    Thanks for asking.

    PS: I have been through 30 prototypes during dev.
     
  14. 3v0

    3v0 Coop Build Coordinator Forum Supporter

    Joined:
    Jul 14, 2006
    Messages:
    9,404
    Likes:
    227
    Location:
    OKLAHOMA USA
    Wow Mosaic. This floored me.
     
  15. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    A few months ago I processed a 10 year old Caterpillar deep cycle (12V, 100Ah) battery (OEM price = $440 USD) that was sitting uncharged in a tropical warehouse (up to 40°c) since 2006. Got it at under 1V EMF on the way to being smelted.
    It regenerated to 50Ah rating and can be put to use as a new 50Ah unit!
     
  16. 3v0

    3v0 Coop Build Coordinator Forum Supporter

    Joined:
    Jul 14, 2006
    Messages:
    9,404
    Likes:
    227
    Location:
    OKLAHOMA USA
    Nice. And here is the other boot. I have a small herd of lawn and garden tractors. My intent is to build a charger maintainer desulphator that can handle between 5 or 10 batteries. The processor end should not be a problem. Thinking of using a PIC for the battery management and maybe a ESP8266 so I can keep an eye on it.

    One the analog side I am thinking of using a bank of inexpensive MOSFET N-CH 100V 36A (IRF540ZPBF) to do the power switching. Actual power requirements should not be large as most of the time there will only be one or two batteries charging or desulphating.

    Seeing that you have spent years on this I am wondering if I am in over my head.
     
  17. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    I ended up having 13 phases in the regeneration, and it's not only desulfation that must be done.
     
  18. crutschow

    crutschow Well-Known Member Most Helpful Member

    Joined:
    Mar 14, 2008
    Messages:
    10,623
    Likes:
    479
    Location:
    L.A., USA Zulu -8
    You don't need an output capacitor because the battery looks like a huge capacitor as you noted, but you definitely need an inductor, since that's what stores the energy and transfers it to the output when the switching FET is off.
    Otherwise the FET will be driving what essentially looks like a short circuit with the current only limited by the FET and DC source resistance. And the efficiency would essentially the same as a linear regulator.
    Yes.
    You can use just about any choke as long as it doesn't saturate at the peak inductor current.
    An iron powder choke may have slightly less efficiency than a ferrite one but the difference should not be that significant.
     
  19. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Thx cruts but I am long past using kickback pulse units.
     
  20. Mosaic

    Mosaic Well-Known Member

    Joined:
    Jun 3, 2010
    Messages:
    2,617
    Likes:
    129
    Location:
    Caribbean
    Sounds like you need my 12V, 8 port float maintainer/charger kit. Features 8 zone independent current limiting (500mA to 1A) and general backup crowbar protection for any voltage regulation loss- to protect the batteries. Compatible with 85 thru 230VAC.
    Does not refuse to charge a 0V battery. But will indicate a higher than expected charge rate as occurs for any battery under 12V, including reverse connected batteries.
    I have completed early deployment with the local Bosch dealership and they're quite happy with it, especially in combination with maintaining regenerated batteries. I was planning to kit it on Tindie or Ebay....If you want a kit I can send you one for about $110, includes a laser cut acrylic housing and battery clips.
    Beats the price point of the Battery Tender or NOCO genius people by a wide margin.
     
  21. crutschow

    crutschow Well-Known Member Most Helpful Member

    Joined:
    Mar 14, 2008
    Messages:
    10,623
    Likes:
    479
    Location:
    L.A., USA Zulu -8
    Who say anything about "kickback pulse units"? :confused:
     

Share This Page