Intermittent Battery Float Charger

Discussion in 'The Projects Forum' started by iONic, Nov 22, 2008.

  1. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    The following diagram is an Idea of an Intermittent Battery Float Charger for multiple batteries. I was considering the CD4541 for the timing portion but am at a loss for the circuitry for switching which battery is to be charged.

    I could use a hand with that part of the circuit.
     
  2. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    Something that counts 0 - 4 will work. Use the output to pull individual relays that pass the float charging current.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Why not use a 4017 to control the gates of N-ch power MOSFETs?

    If you use logic level MOSFETs such as IRLR7807Z, IRLZ24, etc. you could use a CMOS 555 timer to clock the 4017.
     
  4. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    Can I make that long of time delay's with just a 555 and 4017??
    I'm looking for 4hrs of charge time every 24hrs and each day a different battery, 1 - 7 and a charge current of not more than 300mA.
     
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Well you COULD, but I was thinking more of cycling between batteries every 1/2 hour or so. Why just float-charge one a day, when you can rotate around between all of them just as easily?
     
  6. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68

    True! Never thought of it, I guess. Till not I've been manually changing from one battery to another a couple times a day, the final battery being left on overnight till morning.

    In other words I make a 30min timer with a CMOS 555 triggering a 4017 for up to 10 batteries from it's ten outputs? Each 4017 output would have it's own logic level MOSFET controlling a different battery?

    Maybe something like this:
     
    Last edited: Nov 24, 2008
  7. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    Do you thing I need diodes on the source side of each MOSFET?
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Yes. 1/2 hour is an arbitrarily chosen time period. The 555 won't run the way you have it wired.
    Each MOSFET controlling the ground path of a seperate battery. That was the idea. ;)

    No, because diodes would create a voltage drop, which would significantly reduce the float voltage.

    If you want to protect against connecting a battery backwards or shorts, you could use fuses.

    If you were only float-charging a few batteries, you could use the lowest outputs from the 4017, and connect the first unused 4017 output to the RESET input.

    Rather than have a fixed time for each battery, it could be made adaptive. That is, when a new battery is selected, the charging voltage is allowed to stabilize for a relatively short period of time, say 10 seconds. The voltage is then tested using a window comparator. If the voltage is between 12v and 13.2v (or whatever your float voltage is), the charging continues until the battery voltage exceeds your float voltage or a period of time, then proceeds to the next battery.

    If the voltage is below 12v, the battery is either heavily discharged and needs more charging current than the float charger can provide, or has a shorted cell and needs replacement.
     
    Last edited: Nov 27, 2008
  9. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    The batteries are 6V Lead Acid. The power source I am intending on using is fully equipped to charge or trickle charge the 6V batteries, thus I wouldn't need any additional circuitry.
     
    Last edited: Nov 28, 2008
  10. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK.

    If you use 2 meg resistors for R1 & R2, and a 330uF cap for C1, you'll get about 23 minutes per battery. For the cap, use one that's rated for 35v or more (for low leakage), and pre-condition it; basically the same as rejuvenating it.

    Basically, you apply a DC voltage across the cap with a 10k resistor in series, increasing the voltage slowly over a period of time, monitoring the current through the resistor (and thus through the capacitor) by measuring the voltage across it.

    Unless the voltage across the resistor measures less than 9mV with 4v supplied to the circuit, the cap will not charge high enough to permit the 555 to toggle states. Of course, it will take quite a while for the cap to stop charging via a 10k resistor at higher voltage levels; this is the reconditioning process. If you can slowly take a cap up to it's rated voltage by using the current limiting resistor and allow it to stabilize over time, it's leakage current will be far smaller than if you simply placed it into service without the conditioning.
     
  11. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    SGT,
    You are a great help and asset to this forum! The Cap information is new to me and seems like an important piece of valuable information that aa great deal of people could benefit when building more accurate timers with the 555.

    iONic
     
  12. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    Is the 1K resistor and LED a way to indicate which battery is being charged?

    Or would this restrict the MOSFET from turning on?
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    That won't work. 4000 series CMOS ICs have very low current source/sink capability; perhaps 1-2mA @15v. You're running at much less than that.

    You could drive the gate of a 2N7000 MOSFET in parallel with the power MOSFET's gate. Rds(on) for a 2N7000 with Vgs=5 is around 5 Ohms.
     
  14. iONic

    Thread Starter AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    Maybe something like the following??
     
  15. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Yes, that would work.

    It's preferable to use MOSFETs since there is virtually no current draw once the gates have been charged/discharged.
     
  16. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    My two cents: 4017 will drive a LED with 680Ω @ 12V, at 6V may limit LED to 10 mA. I use 12V on my 4017.
     
  17. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    I'm beginning to like the taste of crow. I rechecked resistor, it's 6.8k not 680.and the LED's ,5mm, are verry readable on 9V.,that's 1mA. National shows 8.8mA source @ 15V.
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Bernard,
    Ionic is working with rather narrow margins to begin with. While a 4017 could easily drive one or two MOSFET gates at the extremely low switching rate being discussed, trying to drive an LED in addition would not be a good idea, as it would severely impact the charge/discharge rates of the remaining MOSFET gate(s), quite possibly causing their early demise.
     
  19. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    I started to plot 4017 source capability some years ago but it died. Here is todays result.
     
  20. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Keep in mind that you're stressing your 4017 quite a bit beyond it's design limitations.

    At Vdd=5v, Motorola/ONsemi gives a typical source current of -0.88mA with Vout of 4.6v, and typical sink current of 0.88mA at 0.4v.

    Less than 1mA source/sink capability.

    Mind you, they can source or sink more current - but not at those voltage levels.

    Since MOSFETs require certain voltage levels at their gates to fully turn on or shut off, anything that might slow those transitions down would not be a good idea.

    I see you didn't try any tests with Vdd=5. See how much current you can source before you get below 4.5v, or sink before you get above 0.5v.
     
Loading...