Connect with us

Brown out protection.

Discussion in 'Electronic Basics' started by Control Freq, Oct 3, 2007.

Scroll to continue with content
  1. Control Freq

    Control Freq Guest

    Hi,
    Recently, I experienced a Cisco firewall device break during an
    electricity power failure. The power went off, came back up briefly,
    then suffered a brown out then went off again.

    Inspection of the device showed that the power supply transformer
    device had broke, not the firewall iteself. We got a new firewall
    device (and transformer) to get our systems up and running again.

    Anyway, I was wondering how a brown out could have caused this
    problem. The device is connected to a surge supressed mains outlet!

    So, what is the usual way of protecting sensitive equipment to this
    sort of failure?

    All of our computers are connected to UPS, and they don't (yet) suffer
    from problems with power failures, but all of our network hubs/
    switches are not on UPS, only surge supressors.

    Can someone fill in the blanks in my knowledge.

    Thanks in advance.
     
  2. -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    A brownout is the opposite of a surge - too *low* voltage for safe use.
    Any sensitive, CMOS circuitry inside a complex piece of kit (such as a
    cisco device) can be damaged by a brownout just as much as a surge.
    Then get another UPS! They're cheaper than Cisco firewalls ;]

    - --
    Brendan Gillatt
    brendan {at} brendangillatt {dot} co {dot} uk
    http://www.brendangillatt.co.uk
    PGP Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xBACD7433
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.3 (MingW32)

    iD8DBQFHA+6CkA9dCbrNdDMRAuxBAJoCHtfdUiiHazL7l7ksyUizjYQBhwCgxZDZ
    nKiLR9aPlYNiDnAD9HalCeU=
    =D9Uh
    -----END PGP SIGNATURE-----
     
  3. Meat Plow

    Meat Plow Guest

    Put all your networking gear on a UPS.
     
  4. Rich Grise

    Rich Grise Guest

    As others have said, put the stuff on a UPS - the same one if it
    has the capacity, and another if it doesn't.

    As to why it broke, when the undervoltage happened, the power
    supply was struggling to regulate its output, and with less
    voltage available, it needed to draw more current from the mains,
    which exceeded the input tranny's ratings.

    Hope This Helps!
    Rich
     
  5. Phil Allison

    Phil Allison Guest

    "Control Freq"

    ** The brown out part is what sometimes does the damage.


    ** The Cisco DC supply is a switching type (ie no iron transformer) with
    a wide operating range of 100 to 240 volts AC. For that to stop working or
    be harmed, the normally 240 volt AC supply must have been below 100 volts.

    Such failures are still unusual and only a post mortem on the particular
    supply will reveal the cause.



    ** When a sustained brown out occurs - turn everything off.

    Specially any fridges or air conditioners as it can burn out the motors.




    ........ Phil
     

  6. The power supply is likely to be a small 'Switched Mode Power
    Supply', or SMPS, like those used in your computers. As the supply
    voltage drops, the 'on time increases, and draws more current trough the
    switching transistors. If the design is marginal, it can cause one or
    both transistors to short out, or blow an internal fuse. You need to
    run them from the nearest UPS, or put them on one of their own.


    --
    Service to my country? Been there, Done that, and I've got my DD214 to
    prove it.
    Member of DAV #85.

    Michael A. Terrell
    Central Florida
     
  7. w_tom

    w_tom Guest

    Brownouts do not cause electronics damage. That was an industry
    standard even long before the IBM PC. Low voltage must not harm
    electronics. Many even forget what happens when powering off
    electronics. Electronics suffer a brownout - then eventually a
    blackout - and without damage. Many will say otherwise based only
    upon assumptions, observations, or popular urban myth. For example,
    they saw the brownout but failed to see a reason for that brownout -
    such as a preceding overvoltage. Many of your replies are based on
    speculations. They assume; therefore it is a fact?

    A UPS is to maintain power during blackouts and extreme brownouts.
    Electronics must remain operational for a short time as that UPS
    disconnects from AC mains and connects to the battery. Why is that
    switchover time significant? Because the destructive surge could
    repeat 100 times before that switchover occurs - the switchover takes
    that long and surges are that short. A UPS connects electronics
    directly to AC mains when not in battery backup mode. A UPS cannot
    disconnect from a surge fast enough. The typical UPS does not even
    claim to protect from typically destructive surges. It simply switches
    to battery when AC voltage gets too low.

    Usual way of protecting that Cisco is the same and simple solution
    found in the telephone CO (central office). Any destructive surge is
    earthed before it can enter the building. Do same with one properly
    earthed 'whole house' protector. Such devices are not sold under
    brand names of lesser companies such as Belkin, Tripplite, or APC.
    Effective protectors are sold under responsible brand names such as
    Siemens, Intermatic, Leviton, Square D, Cutler-Hammer, and GE.

    Review spec numbers for that UPS. Chances are its joules are even
    inferior to a paltry circuit found in power strips. That is near zero
    protection. Those who recommend a UPS for surge protection often do
    not even know how the UPS works and did not even first consult spec
    sheets. They know only from half truths promoted on color glossy
    sales brochures.

    All electronics contain internal protection. Computers tend to have
    protection numbers well in excess of other appliances. A surge that
    might harm some appliances may also be irrelevant due to protection
    inside a computer.

    Protection inside any electronics can be overwhelmed. So we install
    one 'whole house' protector to earth the rare and so destructive
    surge. If properly earthed, that rare surge will not overwhelm
    protection inside maybe 100 household appliances.

    How many electronics fortunately were not damaged this time? Did
    you lose furnace controls, dishwasher, smoke detectors, bathroom
    GFCIs, or dimmer switches? Some of those electronics devices are more
    critical to your safety. All have internal protection. Internal
    protection that may be overwhelmed should the rare and destructive
    surge not be properly earthed. We earth any incoming surge so that
    even internal protection in every appliance (includeing that Cisco
    power brick) is not damaged.

    Brownouts only harm electronics where someone saw a brownout,
    assumed no surge preceded it, and proclaimed their assumption as
    fact. This 'cherry picked' reasoning also proved Saddam had WMDs.
    Even 35 years ago, industry standards required electronics not be
    damaged by brownouts. Chances are the brownout was preceded by a
    surge. A surge sufficient to damage only some items and not
    sufficient to overwhelm protection in others.

    Phil Allison makes an important point. Such failures are unusual
    and only a post mortem can provide further definitive facts.
    Potentially destructive surges average maybe once every seven years -
    and can vary significantly even without same town.

    All electronics must either work just fine even when lights are
    dimmed to 40% intensity, or shutdown without damage if that voltage
    drops farther. It is an industry standard so old ... and yet how many
    posted in denial of well known industry standards? Any power supply
    damaged by a brownout was defective when designed. Brownouts do not
    damage electronics.
     
  8. Guest

    I suggest you go get a Variac and try some experiments on your own
    computers and TVs. Since you say it causes no problems, you have
    nothing to lose. I, however, would not bet on your success.

    GG
     
  9. neon

    neon

    1,325
    0
    Oct 21, 2006
    Surge protection only protecs from overvoltage and spikes on the line . UPS on the other hand protects from power loss of significant time. Depends on the design how fast it detects and how fast it switch over. For hospitals aiports there is a load cycle spec designed for it. Usualy batteries supply immidiate power source until a diesel or generator from another source can be switched on
     
  10. A truely uninterruptable supply always drives the load from the inverter,
    with the battery floating on the inverter input. There is no switching
    time. Standby supplies are for people who can't afford a real UPS.
     
  11. Phil Allison

    Phil Allison Guest

    <
    w_tom

    ** I'll take a bet on there being a particular AC supply voltage that
    does, if held long enough, cause damage or fuse blowing to one of the items.

    You will needed to fit a amp meter in line to find the "bad" voltage.



    ......... Phil
     

  12. +-------------------+ .:\:\:/:/:.
    | PLEASE DO NOT F :.:\:\:/:/:.:
    | FEED THE TROLLS | :=.' - - '.=:
    | | '=(\ 9 9 /)='
    | Thank you, | ( (_) )
    | Management | /`-vvv-'\
    +-------------------+ / \
    | | @@@ / /|,,,,,|\ \
    | | @@@ /_// /^\ \\_\
    @[email protected]@[email protected] | | |/ WW( ( ) )WW
    \||||/ | | \| __\,,\ /,,/__
    \||/ | | | jgs (______Y______)
    /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
    ==============================================================
    --
    Service to my country? Been there, Done that, and I've got my DD214 to
    prove it.
    Member of DAV #85.

    Michael A. Terrell
    Central Florida
     
  13. w_tom

    w_tom Guest

    I was doing that decades ago which is only one in a long list of
    reasons why we know brownouts do not damage electronic appliances.
    But again, what is a power off? A brownout followed by a blackout.
    Putting a Variac on electronic designs and lowering voltage is
    standard for all designs because industry standards DEMAND no damage
    on brownouts.

    What industry standards? You don't ask? Why not? Your replies
    never provide numbers or industry standards. So how do you know
    anything? Notice what the chart states in the brownout region: No
    Damage Area. The standard applies internationally. One copy is
    page 3 at:
    http://www.itic.org/archives/iticurv.pdf

    How do appliances fail when this CBEMA standard (and many other
    standards) have existed for over 30 years? Urban myths are widespread
    - and are posted without facts or numbers.

    Those who knew by doing used that variac and therefore knew long ago
    that brownouts must not cause electronics damage. One example is
    "Motheboard Problem? Post Problem?" in alt.certification.a-plus on 7
    Sept 2001 by Tom MacIntyre
    Notice how he was routinely damaging electronics with that variac?
    Oh. No damage. How curious.

    Routine design means determining the actual brownout voltage that my
    design will cut out at (obviously without any damage). All 120 volt
    electronics must operate at 105 volts and work for many seconds at 96
    volts. But computers must even startup and work at 90 VAC - another
    industry standard. Everything I design ran just fine until voltage
    dropped to 85 volts. Did we damage electronics doing these variac
    tests? Never. Obviously not. When that voltage drops below 85
    volts, electronics just turn off. What is required of all designs?
    No damage ever from brownouts. Electronics must work just fine or -
    as even industry standards demand - cut off in the "No Damage Area".

    Please take your own advice - get a variac and start learning by
    doing. Then learn what designer must conform to - industry standards
    as even on Page 3. Brownouts only cause electronic damage when urban
    myths are promoted by those who never did this work and cannot post
    numbers.

    Industry standards are old and blunt: brownouts must never cause
    hardware damage - see that chart on page 3. Intel standards for
    computers demand even better. How many reasons here demonstrated
    brownouts don't cause damage? Too many - with numbers, industry
    citations, and justified contempt for those who would know without
    first learning such facts. Brownouts must not cause electronics
    damage - which is even common sense. Otherwise an appliance would
    self destruct when powered off.
     
  14. w_tom

    w_tom Guest

    Which means most every UPS is not a true UPS. Those UPSes without
    switching are $500 retail and higher. UPSes being discussed here
    connect electronics directly to AC mains when not in battery backup
    mode. Then the power to electronics is a clean sine wave.

    Some in the UPS industry are quite good at promoting these myths.
    Many only assume a UPS always powers from battery because propaganda
    gets promoted as fact. The typical UPS connects electronics
    directly to AC mains AND exposes electronics to some of the 'dirtiest'
    electricity when operating from battery. Why is UPS electricity so
    'dirty'? That inverter is as cheap as possible so that UPS need not
    cost closer to $500.

    So a relay can disconnect from a power surge? Of course not. Notice
    the switchover time - about 10 milliseconds. Just another reason why
    UPSes list (in numeric specs) near zero protection and another reason
    why the UPS does nothing useful for Control Freq's problem. The UPS
    provides temporary power during extreme brownouts and blackouts that
    would otherwise power off that appliance.
     
  15. Control Freq

    Control Freq Guest

    Well, that was, erm, quite an answer. Thanks.

    I wonder what Victor Frankenstein would have made of it all !


    Regards
     
  16. Rich Grise

    Rich Grise Guest

    On Thu, 04 Oct 2007 19:11:40 -0700, w_tom top-posted:
    [top-post repaired]
    Sorry, but that seems to have been what took place.

    And don't top-post.

    Thanks,
    Rich
     
  17. Jamie

    Jamie Guest

    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    Not true, Sorry..

    Still today, PSU's are still made where if a brown out occurs, the osc
    circuit will deadlock (stop) and things like the power transistors
    will sit there and heat up and most likely be damaged before the unit
    will shut down from temperature over load, if it even has that
    protection in it.

    It's what you call cheap and dirty;
    Out the door by 4:30;
     
  18. w_tom

    w_tom Guest

    When the oscillator stops working, then no current flows; controller
    shuts down. Do you really think those PWM chip designers are so
    stupid as to design as Jamie has just posted? Of course not. Doing
    it to meet 30+ year standards is routine. The brownout region was long
    ago called the "No Damage Area". Either a power supply must provide
    DC output voltages; or power supply shuts off. Nothing new about this
    reality.

    What happens in a PC when a brownout is so major? Power supply even
    signals the computer of a shutdown. Does that signal announce a
    hardware failure? Of course not. A most extreme brownout simply
    creates a power off. Supply even signals the computer of a shutdown.

    If a power supply controller was designed as Jamie posted, then
    normal power off could create damage. Industry standards demand that
    brownouts never cause electronics damage. Just one of numerous
    functions standard inside power supplies.

    How curious. Testing new designs in a brownout is normal design
    practice. Strange how damage never results. According to Jamie,
    damage should be routine. If a brownout ever caused damage, then
    instead a major design failure is solved. Electronics perform as those
    old and blunt industry standards demand. Those who actually design
    stuff would know brownouts must not cause electronics damage - ever.

    So how extreme is a brownout? Well Intel specs demand that a fully
    loaded computer even start up normally when voltage is so low that
    incandescent bulbs are at well less than 40% intensity - a brownout.
    Either it must work just fine or shutdown. What Jamie must and fails
    to dispute are industry standards. Why does he post speculation AND
    ignore industry standards? Even 30 years ago, brownouts must not
    cause hardware damage. Suddenly we are making inferior designs
    because Jamie says so? Yes, but only where speculation must also be
    fact.

    Brownouts can cause electric motor failures. Many heavy appliances
    (ie air conditioner compressor) now include circuits to power off
    motors should a brownout be excessive. Curious how even electric
    motors are made resistant to brownout damage. Why are appliances that
    need not meet those industry standards doing so anyway?

    So how often are electronics damaged by bad breath? No standards
    define bad breath induced failures. Therefore bad breath damage is
    possible? Of course. Myths can claim anything when justified by no
    numbers and no facts. However even old standards are quite blunt
    about brownouts - the "No Damage Area". Bad breath damage? Anybody
    can make claims only on speculaton.
     
  19. Jamie

    Jamie Guest

    Yes, there are crap electronics out there.
    Cheap and dirty..
    I've replaced a good many smps with better brands do element that
    problem.
    Being in the industrial field, i've repaired a good
    many smps that took out the bipolar and power fets (depending on the
    voltage at stake) just because of an electrical brown out that cause the
    PS to stop outputting and over heat the transistors if the smps wasn't
    repowered up in time to get the driving OSC started.
    That's crap. You haven't been around enough have you?

    There are many bad designs out there that depend on the OSC circuit to
    operate. when this circuit fails to start or stalls. driving bias can
    be at it's highest point where normally the OSC circuit be working in a
    state of pulling this bias down.

    It's a common problem with bad and cheap designs. Get over it.

    Most of the supplies we use now have a watch dog timer in it now
    to element burn out of components. They work very good how ever,
    you won't find to many of those in the cheap throw away classes.

    Take my word for it. Bad designs still exist on the market!
     
  20. Phil Allison

    Phil Allison Guest

    "Jamie"

    ** SMPS are not the only products that can misbehave when the AC supply
    voltage is very low.

    I have come across transistor amplifiers (a Peavey bass guitar amp) that
    will blow their AC fuse if slowly brought up on a Variac. This happens at
    about 50% of rated AC voltage.

    A Carver C-500 stereo power amp, I tested on the bench just recently,
    cannot control its output voltage until the AC supply is at least 25% of
    normal. When the AC supply is between 20 - 24 % of rated, both channels
    output a DC voltage ( plus hum) of 20 volts into the load.

    The Carver amp itself is not harmed by this, but any connected speakers (
    read expensive hi-fi woofers ) could be expected to go up in smoke.

    The same amp makes only a tiny " plop " noise when switched on the normal
    way and a small "click" after being switched off. The DC offset at the
    output was normally under +/-10 mV.

    BTW:

    Industry design standards that cover the entire electronics manufacturing
    industry, world wide, DO NOT EXIST.

    The Chinese have simply never heard of them......



    ....... Phil
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-