Shop OBEX P1 Docs P2 Docs Learn Events
Hot capacitor — Parallax Forums

Hot capacitor

BADHABITBADHABIT Posts: 138
edited 2009-06-06 19:32 in Propeller 1
I have a product w/a 24v power supply. In it there is a small circuit that gets regulated to 5v via a 7805.
It uses a 2 caps at the regulator a 1uF 50v(electrolytic) on the 24v side and a .1uF on the 5v side.
The 1uF is getting pretty warm - almost hot. This is the first one out of 10 or so I have put together that I noticed this with.
There is not much current flowing.

I am wondering if any of you guys can think of reasons for this happening. Is it possible for bad caps to do this?

BH skull.gif

Comments

  • mctriviamctrivia Posts: 3,772
    edited 2009-06-06 00:24
    definitely a bad cap.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5

    Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
  • hover1hover1 Posts: 1,929
    edited 2009-06-06 00:29
    Off the top of my head, without a schematic (hint..hint), is the polarity correct? Reversed polarity caps will do that.
  • kwinnkwinn Posts: 8,697
    edited 2009-06-06 01:24
    Reversed polarity, low voltage rating, or bad cap.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2009-06-06 02:00
    Reversed perhaps, voltage rating is way above what it needs to be, and bad caps usually just dry out and lose capacity, not get hot.

    But 24V-5V = 19V at say 100ma (at least) = 1.9W that the 7805 has to dissipate which means it gets very hot. Anyway, the heat has to go somewhere, even with a heatsink, could the 1uF be next to it?

    Hint: Linear regulators are not very suitable for high input/output differential supplies unless you are only drawing 10's of milliamps. Switching regulators are a solution though.

    *Peter*
  • BADHABITBADHABIT Posts: 138
    edited 2009-06-06 02:24
    I attached a pic of the layout.

    Sorry for the crappiness
    circuit.jpg
    220 x 302 - 11K
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-06 03:36
    The layout doesn't tell us which direction you actually soldered the 1uF cap into the board. That's what you need to check.

    -Phil
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2009-06-06 03:54
    You need a digital power probe for this sort of thing so to do that get your little pinky and touch the regulator. If you find the digital probe emitting a smoke signal it's probably a good clue that the regulator is dissipating too much power. Likewise the proximity of the electrolytic to the regulator can be likened to a thermal receiver and you can expect a hot reception! [noparse]:)[/noparse] Of course I could be wrong but we enjoy these guessing games where the clues are doled out painfully.

    *Peter*
  • hover1hover1 Posts: 1,929
    edited 2009-06-06 04:04
    Peter Jakacki said...
    You need a digital power probe for this sort of thing so to do that get your little pinky and touch the regulator. If you find the digital probe emitting a smoke signal it's probably a good clue that the regulator is dissipating too much power. Likewise the proximity of the electrolytic to the regulator can be likened to a thermal receiver and you can expect a hot reception! [noparse]:)[/noparse] Of course I could be wrong but we enjoy these guessing games where the clues are doled out painfully.

    *Peter*
    What a perfect application for the MLX90614 Infrared Thermometer Module (10° FOV). Save your pinky!

    I do agree about regulator in close proximity. I just thought of reverse first.

    Jim
  • mctriviamctrivia Posts: 3,772
    edited 2009-06-06 04:42
    i have been using almost all ceramic caps to long. if i remember correctly reversing electrolytic usually give off a really bad smell and a small amount of smoke.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5

    Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
  • jazzedjazzed Posts: 11,803
    edited 2009-06-06 04:47
    I've seen the old aluminum shell types literally explode like a bottle rocket when reversed and power applied.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    --Steve


    Propalyzer: Propeller PC Logic Analyzer
    http://forums.parallax.com/showthread.php?p=788230
  • Clock LoopClock Loop Posts: 2,069
    edited 2009-06-06 05:39
    Oh your talking about the "pass the hot capacitor" game. I like that game.
  • LeonLeon Posts: 7,620
    edited 2009-06-06 07:59
    If it's a tantalum, the bar indicates the +ve terminal.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
    Suzuki SV1000S motorcycle
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2009-06-06 08:36
    Tants can get so hot that they can carbobize the PCB. If it is not the obvious wrongway, your mistake or wrongly labled, around or just radiant heat from the reg, try puting a 'scope onto it. If the thing is unstable perhaps it is "ripple" current that is heating it up.

    P.S. If you get a switch mode PSU that is dead, always try the small value ones on the HV side. They are always overheated/underated and dry out, if it isn't that then it probably has deeper problems that are worth the hassle of repairs.
  • MicrocontrolledMicrocontrolled Posts: 2,461
    edited 2009-06-06 12:48
    Tantilams will explode, electrolythc will burn, but check the amps o the power supply, and then check the max amps on the voltage regulator. make sure your solder is not jointed, just as a percaution.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Toys are microcontroled.
    Robots are microcontroled.
    I am microcontrolled.
  • WurlitzerWurlitzer Posts: 237
    edited 2009-06-06 14:56
    In addition to what has been mentioned, if the cap has a lot of ripple applied to it that can create a lot of heat.

    Ripple current is one of the characteristics I see overlooked frequently. (sometimes it is even hard to find that rating for a given cap)

    The cap is trying to do its job (I am assuming it is·intended to be a filter cap) and if the applied dc voltage has significant ripple then the cap is working overtime.
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2009-06-06 17:20
    And another thing ...

    Usually the input side cap is the small value one, and the output side is the bigger one (which must have ESR within the reg's acceptance range). But then you say that all the other ones worked ok.

    24V down to 5V is quite a drop, if you switch off all the lights does the reg have a "friendly glow"?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-06 17:53
    Toby,

    I checked the 7805 datasheet. His values are okay — but just barely.

    -Phil
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2009-06-06 18:20
    True, I just hope that the current isnt too big (unless its got a heatsink the size of the Titanic bolted to it)
  • BADHABITBADHABIT Posts: 138
    edited 2009-06-06 18:38
    The current is very, very small (under normal operation). This situation was not normal operation. A component down the line was inserted backwards, which was very hard to catch, and that caused another component to jitter.
    Once the reversed component was righted everthing went back to normal.
    Thanks for all the help. It's better than a capacitor chapter in a book.
  • hover1hover1 Posts: 1,929
    edited 2009-06-06 18:45
    BADHABIT said...
    The current is very, very small (under normal operation). This situation was not normal operation. A component down the line was inserted backwards, which was very hard to catch, and that caused another component to jitter.
    Once the reversed component was righted everthing went back to normal.
    Thanks for all the help. It's better than a capacitor chapter in a book.
    Glad you got it worked out BH.

    Jim
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-06-06 19:15
    That's one thing I've found in regulation circuits. Rarely does the actual regulator have a problem(or the capacitors), but instead, it's usually something it is supplying power to. In the future, if there is a problem with the regulator or any part of the regulation circuit, look at the rest of the circuit which it's supplying power to.
  • BADHABITBADHABIT Posts: 138
    edited 2009-06-06 19:32
    Phildapill said...
    if there is a problem with the regulator or any part of the regulation circuit, look at the rest of the circuit which it's supplying power to.
    Absolutely, lesson learned.
  • MicrocontrolledMicrocontrolled Posts: 2,461
    edited 2009-06-06 19:32
    Good job! Glad you solved your problem.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Toys are microcontroled.
    Robots are microcontroled.
    I am microcontroled.
Sign In or Register to comment.