Shop OBEX P1 Docs P2 Docs Learn Events
Selecting resistor — Parallax Forums

Selecting resistor

ThePenguinMasterThePenguinMaster Posts: 89
edited 2007-07-16 18:47 in General Discussion
ok so heres ythe problem...

im wrking on a power supply that outputs both 5 and 12 volts. im working with 18 volts to start. 18 volts and 2.5 amps.
what i have is two power regulators and two caps, the two regulators are a 12 volt regulator and a 5 volt regulator. i have 18 going into the 12 volt regulator and the 12 comming out checks out. so im running 12 volts into the 5 volt regulator and the 5 volt regulator maintains a steady 5 volts so everyting is working well..

however, after a few minutes, both regulators begin to over heat. the power supply is to drive a hard drive, a basic ata hard disk. after it over heats the power goes on and off. theyre both heat sinked and im still having the problem.
i tryed to place a resistor between the 12 volt regulator and the 5 volt regulator to decrease the voltage going into it. when i do this, it stops overheating and there is a steady 6.5 volts going into the 5 volt regulator but my hard drive shuts off ant wont turn on, as if maybe the voltage is correct but the amprage is off?

so if adding a resistor shuts off the drive, and adding resistance lowers the amprage and shuts off the drive.. how do i balance this? am i selecting the wrong resistor?

any help would be much appreciated. i know there is a basic conccept im missing and i would appreciate if someone could clue me in.

Comments

  • bennettdanbennettdan Posts: 614
    edited 2007-07-09 18:39
    You might trying to step down the raw 18v with a voltage divider with large wattage resistors... How much amperage does the harddrive use on the 5v side? If it is low enough you might be able to run your 5v regulator off of your raw 18v supply voltage.
  • stamptrolstamptrol Posts: 1,731
    edited 2007-07-09 19:11
    The clue is to check the input to the 5 volt regulator during the startup of the hard-drive. It is most likely being drawn much lower than the 6.5 volts you measured at light load. Traditionally, resistors would be the last resort in getting voltages from a supply to be stable. At a minimum, you might be able to beef things up with an external pass transistor. Just Google "transistor regulated power supply"

    Also, unless the 12 volt regulator is capable of significant current, it will overheat once you get the 5 volt regulator fixed up.

    Other than the useful experience of designing a power supply, I'd use a surplus computer power supply or purchase a new switchmode supply.

    Cheers,

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tom Sisk

    http://www.siskconsult.com
    ·
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-09 19:17
    well lets see.. i think the 5v draws 500mA i think..
    so is it possiple im drawing more than the regulators can handle, and its not because theyre working with an input voltage that is too high? is that an issue(having a vin too high for the regulator?)?
  • metron9metron9 Posts: 1,100
    edited 2007-07-09 20:36
    It might help if you told us what regulators you are using.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Think Inside the box first and if that doesn't work..
    Re-arrange what's inside the box then...
    Think outside the BOX!
  • FranklinFranklin Posts: 4,747
    edited 2007-07-09 21:43
    P=E*I so if you drop 7 volts at 500ma you are creating 3.5 watts of heat

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - Stephen
  • phil kennyphil kenny Posts: 233
    edited 2007-07-09 22:49
    The +12 volt supply has a high initial load due to the spindle
    motor trying to accelerate the drive to running speed. This can
    be 1 to 2 amps for 5 to 10 seconds. If you haven't mounted the
    regulators on adequate heat sinks, they can easily overheat.

    Some regulators employ an over temperature sensor to
    prevent destruction. When the over temperature protection
    kicks in, the +12 v regulator shuts down. Since you are
    feeding the + 5v regulator from the + 12v regulator, it
    shuts down also.

    Solution: Use heat sinks large enough to accommodate the
    expected thermal load. I wouldn't try to use resistors.

    phil
  • LilDiLilDi Posts: 229
    edited 2007-07-10 04:36
    You should input both the 12 volt and the 5 volt regulators from the 18 volts. That will seperate the 12 volt and 5 volt loads. Most voltage regulator chip are not shunt regulators, so the 3.5 watts dissipation is not right.
    These regulators only draw as must current as the load. Phil is right. The drive spindle motor may be drawing a lots of amps until it settles at its run speed. Heat sinks help, but I suspect the voltage reulator is not rated at 1 amp or better. If the regulator over heats, it will indeed shut down. Try the LM7812 and the LM7805 one amp regulators. With a good heat sink, these regulator should not overheat at 1 amp.
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-10 13:45
    ok i believe im using the LM7805 regulator for the 5 volt. the hard drive is rated at 12v 500mA and 5V 300mA
    so it is possiple that it is drawing more than that?
    i pulled the heat sink out of an old broken computer power supply and it is about 4inches wide and 3 inches tall with a fanned top. im using heatsink compound and the whole heat sink gets way too hot while the hard disk is in idle mode.
    you said most regulators are rated at 1 amp? i have 2.5 amps comming in. and lets sy im using 800ma that means im burning off 1.7ma?
    i guess it is a possibility i could have a short somewhere. that could do it.
    how can i check how many amps im pulling. i used a multi meter, however its a cheap walmart meter. i put the posistve terminal from the meter onto a hot wire and ground out my negitive. i read 5v running stable and when i switch it to DcMa it reads only .01mA which dosnt seem right. this is while im idleing.
    is it possible that when the hard disk is inactive i have too much spare current and its burning up my regulators? or is it more likely im just shorting out?

    i looked into how they use the zener diodes for voltage regulation. i think its an interesting concept. i might try to incorprate that concept.

    also i will try to run both regulators off of the 18v 2.5 mA supply, that will be safe right?

    and i have thought about using a pre made power supplu.. trust me.. its tempting. but i want to figure out some more intermediate uses for analog components. i guess its a learning experience lol.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2007-07-10 14:16
    You are likely burning off 1.7Amps and that is your problem.
    The LM7812 would run considerably cooler with a mere 15 volt supply. 18 volts may be too high.

    If the regulators are top quality, they have thermal shut down that protects them.

    As point out before the 12volts drives the motor of the hard disk and is most of your power consumption. The 5 volts is quite minimal.
    Nonetheless, going from 12volts to 5volts is again a big heat dump.

    It might be easier to get a dual powered switching supply that provides +12 and +5 as that was what they were designed for.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········
    ···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
  • metron9metron9 Posts: 1,100
    edited 2007-07-10 15:34
    Did you use bypass caps? (see attached picture)

    Can you connect just the LM7812 to the 18V with the bypass caps and no load without it getting hot? Try the same with the LM7805.

    Then try a simple LED using a 1K resistor for both voltages. The regulators should not even get warm with these tests. If they do perhaps you dont have them connected correctly.


    If you test the current you need to put the positive from the 7812 to the positive lead of the meter and the negative lead of the meter to the load. That way the current flows through the meter, in its positive lead and out its negative lead. Your meter has to be rated for the amps that will flow however. Typically it will be fused and the fuses are inside the meter. Some have separate connectors for low current measuring (typically up to 200 milliamp s) and high current measuring (up to 10A).

    Another way to measure the current is to use a large wattage resistor in series with the circuit and measure the voltage across the resistor and calculate the current using OHM's Law.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Think Inside the box first and if that doesn't work..
    Re-arrange what's inside the box then...
    Think outside the BOX!
    808 x 374 - 25K
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-10 18:01
    i am using a single cap on the output side of the regulator. i didnt pout one for the input because im using a steady 18 volts. i have a power supply from a laptop so it should be relitivally stable.
    ok so lets say that im burning off too many watts. oi can potentially use a external pass transistor and zener diode to control the wattage?

    i know sthease are some basic questions but there concepts that i need to learn. i appreciat your help guys..

    oh and on a side note.. when the hard disk is not running the regulators do not overheat.. so i dont think its a short. prehaps i will be unable to use a voltage regulator because it draws more amperage than the regulator can handle.
  • LilDiLilDi Posts: 229
    edited 2007-07-10 18:20
    Connecting the input of both regulaor to the 18 volt supply is safe and the preferred method. At the risk of repeating myself, these voltage regulator only draw as much power as the load. If no load is present, no current is drawn from the 18 volt supply and the regulator will be completely cool. Even if your supply voltage is 35 volts instaed of 18 volts, if there is no load, no current is being used and no wattage used. If the load is 800 ma·at the output of the 12 volt regulator, then the current draw from the 18 volt supply is 530 ma. With a 35 volt supply the current draw from the 35 volt supply is 275 ma.

    I hope that clears thing up for you.
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-10 19:35
    ok yes that dose help a lot.. that sort of sums up a lot of different things ive been hearing. so i can see why its bad to pull the 5v off of the 12 because im pulling way to much through that regulator. i will also check for shorting and check my amperage. i appreciate the help because not i think i know where to start. and i will stay away from resistors for current limitng on power supplys.
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2007-07-10 20:37
    LILDI,

    better check your math....

    Regardless if your supply is 18V or 35V and you are drawing 800mA through the output of a 12V regulator, you are also drawing 800mA from the 18V or 35V supply.


    Suppose that your supply is 35V and you are drawing 800mA from the output of the 12V regulator. That means that there is a difference of 23V across the regulator that has to go somewhere. It does this in the form of heat. Using the formula P=I*V ... P = 800mA * 23V = 18.4 Watts of HEAT.

    The same could be said for an 18V supply drawing 800mA from the 12V regulator. In this case there is a difference of 6V. Using the same formula P=I*V ... P = 800mA * 6V = 4.8 Watts of HEAT.

    Hope this helps.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • LilDiLilDi Posts: 229
    edited 2007-07-10 21:59
    You are absolutely right Beau. I will try to refrain from talking about thing I know absolutely nothing about in the future.

    My apologies!!
  • metron9metron9 Posts: 1,100
    edited 2007-07-11 03:10
    The thought crossed my mind if you could use 2 regulators in parallel. I googled and found this. It can be done. See bottom of page 3

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Think Inside the box first and if that doesn't work..
    Re-arrange what's inside the box then...
    Think outside the BOX!
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-11 13:55
    i see what you mean about two regulators in parallel. i set mine up simular to the picture you sent me. hmm.. now this is interesting.. i wrote down a lot of what you guys said about what i could possibally do to fix this problem and also the comment about figuring out the watts is also an importiant concept. i beleive the regualtor has a max dissapation of about 15 watts. now another thing, i noticed on the adapter it has a watt rating of 30 watts. dose this mean 30 watts of heat are generated by the current flowing from the supply at max ratings? what i mean, is that if the adapter is running at maximum 30 watts of heat will be generated?
    i also tryed checking how many amps the drive is drawing on startup. what i did is i have two wires going from the adapter to the project board. i disconnected the positive wire comming from the adapter where it connects to the project board.
    i then connected the positive wire comming from the adapter to my positive probe on my meter, then connected the negitive probe to where the positive charge from the adapter would enter the project board.

    i turned on the board and the amps jumped up and then went back down. now i have a cheap meter and its hard to understand what im reading here, there are two settings, mA and A. wether i set it to amps or mA it still reads 6.1
    just the symbol behind it changes from m to ma which makes no sence lol. so should i assume my project board is consuming 6 amps? i think i need a better meter..

    i know that there are places i can buy a supply, and i do google a lot of stuff.. google is my friend, but i ask a lot of basic questions, and finding some of thease basic concepts on google can be a little challenging, also its not as good for trouble shooting, so thanks for dealing with me asking my noobie questions lol


    EDIT:

    oh! ok! i see it now! the parallel power regulators. at first glamnce i thought it was 2 different voltage regulators, like a 5 and a 12, but its a way to distribute the current being drawn across two regulators of the same voltage.. ok i get it. thats interesting.. ill have to try out that idea and see if i can distribute the load.

    Post Edited (ThePenguinMaster) : 7/11/2007 2:00:46 PM GMT
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2007-07-11 14:42
    ThePenguinMaster,

    Keep in mind that D1, D2, etc for the schematic that metron9 supplied ALSO need to handle the current demand of 1A.

    In my opinion, there are better ways to do this. If you look at the datasheet for the regulator.
    on Figure 12 and 13 there are already established methods from the manufacturer on how to increase the output current drive.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • metron9metron9 Posts: 1,100
    edited 2007-07-11 14:59
    Another option is just use a higher current regulator. Here is one that is only 90 cents qty 1 that will deliver 3 amps.

    The previous post was could it be done, I don't think using dual regulators should be done unless you really know what you are doing.

    The 90 cent part here and others that range up to the 8 dollar range all have different features. This one has a 4% voltage tol so it is cheap and perhaps not recommended for a hard drive, I din't know.

    I added another pdf file on a 5A regulator in a to-220 package that has only 1.2V dropout. Comes in fixed and adjustable versions. Under 3 bucks

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Think Inside the box first and if that doesn't work..
    Re-arrange what's inside the box then...
    Think outside the BOX!

    Post Edited (metron9) : 7/11/2007 3:28:19 PM GMT
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-11 15:40
    i didnt really think of that as an option at first because im dealing with radioshack stock for the time being. i can probibally order some 3 amp regulators from newark and see what happens. i just didnt think id be drawing that many amps. im using a cheapo wd caviar hard disk.. heres the specs:


    Electrical Specifications
    Current Requirements
    12 VDC
    Read/Write 450 mA
    Idle 470 mA
    Standby 23 mA
    Sleep 23 mA

    5 VDC
    Read/Write 800 mA
    Idle 750 mA
    Standby 330 mA
    Sleep 200 mA

    Power Dissipation
    Read/Write 12.80 Watts
    Idle 9.50 Watts
    Standby 1.90 Watts
    Sleep 1.30 Watts

    i hooked the 5 and the 12 volt regulators directally to the 18 volt supply, instead of having the 5 volt mooch off of the 12 volt regulator.
  • metron9metron9 Posts: 1,100
    edited 2007-07-11 16:28
    Well you see right there, the 12V regulator had to provide 800ma to the 5V regulator alone with not enough headroom to supply the 450ma for the 12V requirement.
    At 800ma on the 5V regulator it will get hot, too hot to touch so it needs a good heat sink. I like to double the maximum supply so your 12V at 1A max is fine. The 5V 1A is near the top but should still work.
    The peak current however may be higher when starting up though. if you use a big capacitor make sure you read the datasheet on the diode needed to protect the regulator, there is a schematic and warning.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Think Inside the box first and if that doesn't work..
    Re-arrange what's inside the box then...
    Think outside the BOX!
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-11 17:50
    i tryed two designs. one was to have each draw from the 18 volts wich should be well within spec because it can take up to 35 volts. and the 12 also runs straight off of the 18 volts.

    another way i tryed was to have the 5v feed off of the output of the 12 volts.

    they both provided thew same result. When i get home today im going to draw a schematic and take some pics. its something i need to do anyways because i like to document my projects. then ill post them on here. things are rapidally comming into focus on how thease things work. apperentally there is a problem but its not what i thought it was at first.. being the whole adding a resistor thing.. now i realise that wouldnt help lol.
  • ThePenguinMasterThePenguinMaster Posts: 89
    edited 2007-07-16 18:47
    ok well sorry for double posting here but ive resolved the problem! it turns out that it was a bad power regulator. and well.. muy grounds wernt hooked up veary well which could have contriburted to it. i rearranged the board and fit things togather much better. i also added capacitors before the regulators and ran my 5 volt off of an 8 volt regulator i ripped out of rsome ild electronics.. and hthe 87 volt regulator is running right offf of the 18 volts. i also put another heat sink on the regulators.. so two regulators are gorouped togather oin one sink (it has two spots.) and the 12 volt is on its own heat sink. also i wired in a cool little cooling fan that snaps right ove rthe radioshack breadboard. i pulled it off·of an old pentium cartrige style processos.. it works perfectally.. i·ran the device with my·propeller running a basic tv output program while driving the hard drive.. the power was stable and the heatsinks warmed.. but stabalized at a cool temp.·the fan makes a huge·difference!

    thanks for your help guys.. you helped me understand a lot of different concepts, which can be applied to other projects.·
Sign In or Register to comment.