Shop OBEX P1 Docs P2 Docs Learn Events
Battery Backup Question — Parallax Forums

Battery Backup Question

ArchiverArchiver Posts: 46,084
edited 2002-11-12 16:46 in General Discussion
We are currently finalizing an in-house security system based on the
BS2, and are left with one final obstacle to overcome...

We have a power supply rated at 12VDC which powers the unit (And a 5V
regulator for the BS2 & LCD), and we have a 12VDC Battery Pack
(Consisting of 8 AA batteries). What we are trying to do is design a
power switching circuit to switch over to battery power in the event
that main power is lost.

Anyone have any good circuits for handling this? The currently
tabled idea is to use a relay...When power is lost, it will release
and the DPDT contacts will switch over to the battery. To overcome
the switch we thought maybe a couple of capacitors.

Anyone have any ideas?

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 17:49
    D1,4001
    power---->|
    o
    circuit
    |
    D2,4001 |
    battery ---->|
    '

    circuit is powered by whichever power supply has higher voltage. D1
    may not be necessary, depending on the power supply.

    There is a voltage drop across the diodes. For lower drop, use a
    schottky diode, say 1N5817. There are circuits using transistors
    that have lower drops still. Your relay scheme should work too, but
    may be overkill.

    -- Tracy




    >We are currently finalizing an in-house security system based on the
    >BS2, and are left with one final obstacle to overcome...
    >
    >We have a power supply rated at 12VDC which powers the unit (And a 5V
    >regulator for the BS2 & LCD), and we have a 12VDC Battery Pack
    >(Consisting of 8 AA batteries). What we are trying to do is design a
    >power switching circuit to switch over to battery power in the event
    >that main power is lost.
    >
    >Anyone have any good circuits for handling this? The currently
    >tabled idea is to use a relay...When power is lost, it will release
    >and the DPDT contacts will switch over to the battery. To overcome
    >the switch we thought maybe a couple of capacitors.
    >
    >Anyone have any ideas?
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 18:24
    I am not sure what you would do with "a couple of capacitors" as you
    mentioned. But I have often used a simple diode to provide switching from one
    power source to a battery pack. This also provides an easy means of keeping
    a trickle charge flowing in the batteries (rechargeable).


    [noparse][[/noparse]Non-text portions of this message have been removed]
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 18:46
    --- Tracy Allen <tracy@e...> wrote:
    > D1,4001
    > power---->|
    o
    circuit
    > |
    > D2,4001 |
    > battery ---->|
    '

    Thanks Tracy...I've seen this idea in use before
    (without the second diode). Our Power Supply is a
    simple regulated 12VDC supply using an LM7812. This
    does look like a better option.



    =====
    Chris Savage
    Knight Designs
    324 West Main Street
    Montour Falls, NY 14865
    (607) 535-6777

    http://www.knightdesigns.com

    __________________________________________________
    Do you Yahoo!?
    U2 on LAUNCH - Exclusive greatest hits videos
    http://launch.yahoo.com/u2
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 18:50
    --- azeasi@a... wrote:
    > I am not sure what you would do with "a couple of
    > capacitors" as you
    > mentioned. But I have often used a simple diode to
    > provide switching from one
    > power source to a battery pack. This also provides
    > an easy means of keeping
    > a trickle charge flowing in the batteries
    > (rechargeable).

    Well, the Capacitors would've been used as filtering
    caps on the output of the relay so that while the
    relay was switching from one side to the other the
    caps would provide voltage.

    On your reply, are you saying that if I use a Diode
    that there will be voltage flowing back into my
    battery pack? Because the battery pack we have is NOT
    rechargeable, so this would not be ideal in this
    situation...Last thing I need is a battery leak.



    =====
    Chris Savage
    Knight Designs
    324 West Main Street
    Montour Falls, NY 14865
    (607) 535-6777

    http://www.knightdesigns.com

    __________________________________________________
    Do you Yahoo!?
    U2 on LAUNCH - Exclusive greatest hits videos
    http://launch.yahoo.com/u2
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 18:54
    --- In basicstamps@y..., Tracy Allen <tracy@e...> wrote:
    > There is a voltage drop across the diodes. For lower drop, use a
    > schottky diode, say 1N5817. There are circuits using transistors
    > that have lower drops still. Your relay scheme should work too,
    but
    > may be overkill.

    One more question Tracy...Regarding Azeasi's reply, would this cause
    any leakage into my battery pack? The battery pack is made up of 8
    Duracell batteries...Non-Rechargeable...I myself have never tried
    this design before.

    Chris Savage
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-11 21:04
    One other consideration, if your power supply doesn't mind having the 12
    volts across it when it is without commercial power, you don't have to worry
    about isolation. Let the system run off battery all the time and the power
    supply can charge it as required.

    The schottkey diode solution is good as well and there are versions with
    only .25 volts dropped across them. As the power drops from the power
    supply, the diode on the battery will begin to conduct and you will never
    even know power dropped (except you're sitting in a dark room <grin>)

    Tim



    Original Message
    From: ulibasic [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=A-n5S6fqmA9PBoptDkFcJPObueQdwAlKrkC0zhZv_fXP1reRLkZyuvIdPo79s47QPD3S0Dj2BhFsuHoHw6eNpzT5XyQJb6hZzuET]ulibasic@r...[/url
    Sent: Monday, November 11, 2002 1:46 PM
    To: basicstamps@yahoogroups.com
    Subject: AW: [noparse][[/noparse]basicstamps] Battery Backup Question


    Hey Tracy,

    do you think this will be fast enough to work without resetting when the
    power is interrupted abruptly? Could this even be some kind of
    brownout-prevention? Are there diodes that suit this purpose better (in
    other words: faster diodes)?

    Thanks, Uli



    Urspr
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 01:13
    Yes, there is leakage back into the battery from the diode.

    A 1N4001 diode back biased will have a leakage current at room
    temperature of about 10 nanoamps. That increases to maybe 100
    nanoamps at the highest ambient temperatures you would encounter.

    I have always heard that alkaline cells are safe for "leakage
    charging" at that level, that is, the leakage is less than the
    self-discharge rate of the battery.

    To put this in perspective, if you are worried, then you could
    install a 22 megaohm resistor in parallel with your reserve battery
    pack. The resistor would draw a current constantly, 12 volts/22
    megaohms = 55 nanoamps. Don't worry too much about the extra load.
    Your 8 AA cells are rated at over one amp-hour, and it would take
    almost 2 million hours, over 200 years, for the batteries to
    discharge through the 22 megaohms. (They would self-discharge long
    before that!!!). The resistor would sink the excess charging
    current. But my point really is, you don't even need the resistor.

    Schottky diodes are _muc_h more leaky in reverse, maybe 10 microamps
    at room temperature. That might argue against a Schottky.

    The best diode, the lowest leakage, would actually be the
    base-collector junction of a transistor. The collector base junction
    of NPN transistor will have a leakage of orders of magnitude lower
    than the 1N4001. For example, I just pulled a 2N3904 NPN transistor
    out of the drawer, and measured a reverse leakage at room
    temperature, 10 volts reverse bias, of about 90 picoamps. I also
    measured a big 2N3055 in a TO220 case, and that was low too, only 270
    picoamps, 0.27 nanoamp.

    D1
    power
    >|---;
    |
    C E |
    battery
    \ />----o--o----load
    |
    |B | NPN transistor
    `
    ' base connected to emitter
    Vcbo > 20 volts
    Ic ample for the load


    I do not know exactly what level of reverse current would tolerable
    for an alkaline battery. You could always add the parallel resistor
    to absorb the current. Jim Higgens on this list is more of a battery
    expert, and I would defer to his suggestions, if he happens to check
    in on this one.

    The need for diode D1 in your circuit will depend on the level of
    leakage back into the power supply when the power is turned off.
    Don't assume that it will be zero!

    -- best regards
    Tracy Allen
    electronically monitored ecosystems
    http://www.emesystems.com
    mailto:tracy@e...




    >--- In basicstamps@y..., Tracy Allen <tracy@e...> wrote:
    >> There is a voltage drop across the diodes. For lower drop, use a
    >> schottky diode, say 1N5817. There are circuits using transistors
    >> that have lower drops still. Your relay scheme should work too,
    >but
    > > may be overkill.
    >
    >One more question Tracy...Regarding Azeasi's reply, would this cause
    >any leakage into my battery pack? The battery pack is made up of 8
    >Duracell batteries...Non-Rechargeable...I myself have never tried
    >this design before.
    >
    >Chris Savage
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 01:16
    After I posted this, I read the rest of the thread <oops>

    One question bugging me, if this is a critical security system, why not use
    a sealed lead-acid or gel-cell rather than the alkalines? Too much
    overkill? How much power is wasted by the regulator dropping the 12 volts
    to 5 which becomes important when running off battery? I'm not trying to
    infer which is better, just curious.

    Tim



    Original Message
    From: Tim Tapio [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=yb8JqPwUVgvTgnt05rYIBrCRAgvv37LJv6zAiPH0pp51BxVyD5kBoZubIP73nMbeTC51-BEF]tim@t...[/url
    Sent: Monday, November 11, 2002 4:05 PM
    To: basicstamps@yahoogroups.com
    Subject: RE: [noparse][[/noparse]basicstamps] Battery Backup Question


    One other consideration, if your power supply doesn't mind having the 12
    volts across it when it is without commercial power, you don't have to worry
    about isolation. Let the system run off battery all the time and the power
    supply can charge it as required.

    The schottkey diode solution is good as well and there are versions with
    only .25 volts dropped across them. As the power drops from the power
    supply, the diode on the battery will begin to conduct and you will never
    even know power dropped (except you're sitting in a dark room <grin>)

    Tim



    Original Message
    From: ulibasic [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=lmB3dS45dQoCkr1TEze7Yv2PRzdTM-KfG-E4bmq56qr4S91e6Z02QQgeBOpVJhxrSd40jCskqcNziygBAXwRorpjeDBy09guF7vj]ulibasic@r...[/url
    Sent: Monday, November 11, 2002 1:46 PM
    To: basicstamps@yahoogroups.com
    Subject: AW: [noparse][[/noparse]basicstamps] Battery Backup Question


    Hey Tracy,

    do you think this will be fast enough to work without resetting when the
    power is interrupted abruptly? Could this even be some kind of
    brownout-prevention? Are there diodes that suit this purpose better (in
    other words: faster diodes)?

    Thanks, Uli



    Urspr
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 01:45
    Correction to the circuit. I showed the transistor backwards. Should be:

    D1
    power
    >|---;
    |
    E C |
    battery ----o--< /
    o----load
    |
    | |B NPN transistor
    `
    ' base connected to emitter
    Vcbo > 20 volts
    Ic ample for the load
    equals D2 --->|----
    very low leakage

    NOT....

    > D1
    > power
    >|---;
    > |
    > C E |
    > battery
    \ />----o--o----load
    >
    |
    > |B | NPN transistor
    > `
    ' base connected to emitter
    !!BACKWARDS ---|<--- !!
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 02:20
    --- Tracy Allen <tracy@e...> wrote:
    > Yes, there is leakage back into the battery from the
    > diode.
    > to absorb the current. Jim Higgens on this list is
    > more of a battery
    > expert, and I would defer to his suggestions, if he
    > happens to check in on this one.

    Okay, well, I am going to save your message, as you
    went to great lengths in this, and I am going to try
    several variations as you have mentioned to see what I
    get. Perhaps I should consider using Rechargeable
    batteries...I never really thought that much about it
    when the whole battery backup idea was mentioned.

    Interestingly enough, while testing the "door-chime"
    section of our system today we were subjected to a
    double-shot brownout, which caused me to lose a temp
    BS2 program I was in the middle of putting
    in...Although it did help me test the power-fail
    response of the siren and door-chime outputs. =)


    =====
    Chris Savage
    Knight Designs
    324 West Main Street
    Montour Falls, NY 14865
    (607) 535-6777

    http://www.knightdesigns.com

    __________________________________________________
    Do you Yahoo!?
    U2 on LAUNCH - Exclusive greatest hits videos
    http://launch.yahoo.com/u2
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 02:25
    --- Tim Tapio <tim@t...> wrote:
    > One question bugging me, if this is a critical
    > security system, why not use
    > a sealed lead-acid or gel-cell rather than the
    > alkalines? Too much overkill?

    Well, we were trying to stick with what we have in
    stock in the shop, since this is more of an
    afterthought. We have surplus LCD displays, Sensors,
    Buzzers, Sirens, etc. We even have a digital keypad
    for Radio Shack Alarm Systems which we haven't yet
    deciphered it's operation (serial interface). We
    could use anything, although another consideration is
    space.

    > How much power is wasted by the regulator
    > dropping the 12 volts
    > to 5 which becomes important when running off
    > battery? I'm not trying to
    > infer which is better, just curious.

    The regulator is the one on the BS2 OEM board, which
    is what's powering the 5V devices in this design.
    Actually the only other device at 5V is the LCD
    display.



    =====
    Chris Savage
    Knight Designs
    324 West Main Street
    Montour Falls, NY 14865
    (607) 535-6777

    http://www.knightdesigns.com

    __________________________________________________
    Do you Yahoo!?
    U2 on LAUNCH - Exclusive greatest hits videos
    http://launch.yahoo.com/u2
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 15:52
    At 20:13 11/11/02, Tracy Allen wrote:
    >Yes, there is leakage back into the battery from the diode.
    >
    >A 1N4001 diode back biased will have a leakage current at room
    >temperature of about 10 nanoamps. That increases to maybe 100
    >nanoamps at the highest ambient temperatures you would encounter.
    >
    >I have always heard that alkaline cells are safe for "leakage
    >charging" at that level, that is, the leakage is less than the
    >self-discharge rate of the battery.
    >
    >To put this in perspective, if you are worried, then you could
    >install a 22 megaohm resistor in parallel with your reserve battery
    >pack. The resistor would draw a current constantly, 12 volts/22
    >megaohms = 55 nanoamps. Don't worry too much about the extra load.
    >Your 8 AA cells are rated at over one amp-hour, and it would take
    >almost 2 million hours, over 200 years, for the batteries to
    >discharge through the 22 megaohms. (They would self-discharge long
    >before that!!!). The resistor would sink the excess charging
    >current. But my point really is, you don't even need the resistor.
    >
    >Schottky diodes are _muc_h more leaky in reverse, maybe 10 microamps
    >at room temperature. That might argue against a Schottky.
    >
    >The best diode, the lowest leakage, would actually be the
    >base-collector junction of a transistor. The collector base junction
    >of NPN transistor will have a leakage of orders of magnitude lower
    >than the 1N4001. For example, I just pulled a 2N3904 NPN transistor
    >out of the drawer, and measured a reverse leakage at room
    >temperature, 10 volts reverse bias, of about 90 picoamps. I also
    >measured a big 2N3055 in a TO220 case, and that was low too, only 270
    >picoamps, 0.27 nanoamp.
    >
    > D1
    > power
    >|---;
    > |
    > C E |
    > battery
    \ />----o--o----load
    >
    |
    > |B | NPN transistor
    > `
    ' base connected to emitter
    > Vcbo > 20 volts
    > Ic ample for the load
    >
    >
    >I do not know exactly what level of reverse current would tolerable
    >for an alkaline battery. You could always add the parallel resistor
    >to absorb the current. Jim Higgins on this list is more of a battery
    >expert, and I would defer to his suggestions, if he happens to check
    >in on this one.
    >
    >The need for diode D1 in your circuit will depend on the level of
    >leakage back into the power supply when the power is turned off.
    >Don't assume that it will be zero!

    I've been reading and I see very good advice being given.

    I think I saw something questioning the reverse recovery time of a 1N4001.
    Remembering, it will switch as soon as the voltage of the supply drops
    below the voltage of the battery - about 12 volts - and that both will
    still be well above the voltage the regulator needs to maintain regulation,
    I don't think a 2 microsecond recovery time is a problem. That's a lot of
    head room in my book.


    Jim H
  • ArchiverArchiver Posts: 46,084
    edited 2002-11-12 16:46
    I am sorry for the mis communication. The diode arangement suggested by Tracy
    of EME should solve your problem!


    [noparse][[/noparse]Non-text portions of this message have been removed]
Sign In or Register to comment.