Announcement

Collapse
No announcement yet.

Building a 12 V 15 amp Regulator got some ?

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Building a 12 V 15 amp Regulator got some ?

    Ok, 1st and foremost yeah I know I should of bought a opus but it looked too big but now it looks like my solutions will be just about as big. Anyways, as I'm sure most of you have already heard a 100 times I am planning to use regulators in order to send a smooth 12 volt output to that square looking 200 watt powersupply on minibox.com. Um... I also have the ITPS which of course doesn't get close to producing enough power for my current setup estimated at at least 100 watts. So here is my idea. Unlike most other people who have talked about using a ton of 1 amp or 2 amp regulators I went straight for 2 7.5 amp ones. In addition most people have used regulators with drop out voltages of as much as 1.5 volts. The model I plan to use MIC29712 http://www.micrel.com/_PDF/mic29710.pdf which has a MAXIMUM dropout of .7 volts with an estimated .4 dropout for the load I'm going to be sending through it. So now I have two 7.5 amp regulators. Looking at the diagrams in that pdf file I came up with a fairly simple design. I would take the 12v out of the ITPS and send into pin 1 which basically turns the regulators on or off. Pin 2 is unregulated positive in. Pin 3 is ground Pin 4 is used for adjusting the output voltage and pin 5 is regulated output. In order to get a 12 volt out I put in a 100 kohm resistor from pin 5 to 4 and an adjustable pot set at around 12k ohms (but adjustable) from pin 4 to pin 3. I plan to use a breadboard since they have straight lines under the surface which connect multiple spots making it easy to connect everything. One of my questions is this. I will have 15 amps running through the board and from what I've seen most say they are only designed for around 4-5 amps which could cause some serious problems with heat or resistance. So is there any backing in this since most electronic stuff I see doesn't have problems with taking a lot more current than that in the same size wire. My next question is this, most boards that people are designing contain several capacitors while I have not included any on mine at all. I don't care about the PC rebooting on crank. But I do want the system to be able to shutdown after the alternator turns off and the voltage drops down to 12v. My final question is this. I would like to make some use of the 5-7 amps of regulated 12v outputted by the ITPS so I've thought of using a diode so I can both use to power the logic input as well as go provide some extra power to the computer, relieving the regulators of some work and heat. The reason it is required is because otherwise it would be feeding itself which would make it never turn itself off. My ? is most diodes take out some current which if I can help it I would like to get around. I really thank you for reading this. I know this is pretty long. I'll try and explain it further if need be but I don't understand why the circuit need to be so hard. Thanks for all your help

  • #2
    1) I think that it will pop if the input voltage exceeds 17v or so

    2) This is a linear device. Prepare for an 8 inch fan and about 3 tons of aluminum heat sink!!

    I would use a switching regulator, the only bad thing is that it needs an inductor

    http://www.national.com/ds/LM/LM2588.pdf wow look at this thing
    perhaps it could withstand cranking. (lm2588-12)
    Carputer currently 'ghettoed' into car!!!

    EPIA MII-12000
    Ampie Case with rigged extra USB
    Hitachi 60GB
    Holux GPS mouse with iGuidance 2.0
    Lilliput 7" touchscreen
    Netgear WG121 USB WiFi
    Carnetix P1260

    Comment


    • #3
      http://www.micrel.com/_PDF/mic29710.pdf

      That's a great find. How much does one of those puppies cost?

      They may well be a better solution that the 20 Amp fixed diode voltage drop solution I've been pursuing to match my car characteristics.

      I was thinking of using a predrilled solder ringed bread board from radio shacklike this one:
      http://www.radioshack.com/product.as...t%5Fid=276-148
      but was also worried about the high currents.
      Progress: 80% - Permanent install left.
      Motion LS800 Tablet PC and dock.
      Vista, Bu-535 GPS, RoadRunner, MPT2006.

      Comment


      • #4
        Here's the thread on my cheap solution. Even if I use these regulators, I still need to solve my switching problem.

        http://www.mp3car.com/vbulletin/showthread.php?t=36041
        Progress: 80% - Permanent install left.
        Motion LS800 Tablet PC and dock.
        Vista, Bu-535 GPS, RoadRunner, MPT2006.

        Comment


        • #5
          Triwav: Just a note, could you use paragraphs when you post? It's as bit hard to read as it is...

          I assume you mean you're using stripboard / veroboard instead of breadboard? If you were using a breadboard i'd be very scared, since i think most are only rated for 500mA... Veroboard's only meant for around 1-2amps, but you can increase this if you tin the tracks with solder, bridge multiple tracks, or solder your own copper wire over the top of the tracks. Just be careful because it makes it a lot easier to short things with a big solder blob underneath.

          By saying capacitors are you asking about just normal capacitors in a circuit? like 100uF, 470uF ones? You should use capacitors wherever transients could cause problems: If you have a microcontroller or any sort of analogue circuitry that needs a stable power supply, place a capacitors near the supply pins. The same goes for anything that might go weird on you if the voltage suddenly drops (eg. relays which might disconnect momentairily). Most of the time, no-one bothers to do the transient calculations to see what capacitor you should use because they're a pain in the arse, just try something reasonable.

          For the linear regulators that you're using, the datasheets should have a typical circuit with the capacitors that you'll need to use.. if you don't put those in, the circuit will probably work but it might start doing strange things in extreme circumstances.

          If by capacitors you mean 1F mega caps and stuff around that size, search the forums, there's a lot of discussion on those..

          (hope i haven't missed the point on the above)

          Comment


          • #6
            Mm just a note about National's switchmode ICs.. A friend built me 12V to 5V buck regulator using an LM2678, and we stuck the output on an oscilliscope with a 1A load, and the ripple voltage that we saw was quite bad.. We put a 4A load on it, and the ripple was around 100mA... which isn't suitable for anything like a PC...

            Then again, the ripple might've been due to a bad board layout...

            I don't think anyone here's tried a flyback themselves because it's quite hard and i think you have to wind a custom toroid, which is lots of fun.. >_<

            Comment


            • #7
              And don't put the itps and your own regulator in parallel, that will cause serious trouble, even if you'd use a diode.
              List of front-ends/usefull apps
              XTroniC | XTroniC Direct

              Comment


              • #8
                Choyak- if you look at the specs on these regulators they can easily take 20+ volts, however regulators release the difference in voltage in heat so it would get really hot if I constantly had it running off of 20 volts, however that is not what I'm doing. I'm running it off around 13.8, which shouldn't be too much heat. If you look at your pcs motherboard you will notice some regulators and they shouldn't even have heatsinks on them. With a heatsink and a fan running inside my computer I should be fine on heat and to be quite honest these things can withstand tempatures to 125 degrees celcius and have built in thermal shutdown so it should be fine.

                Starfox- sorry about not using paragraphs I tend to ramble on anyways, it was a breadboard I was looking at and it said 4amps max so I really don't know. What would be your suggestion to solve the problem. When I was studying up on what gauge wire to run to my amps, I learned that the main factor is distance and amps. This is a lot of amps but considering it's such a short distance, the resistance shouldn't be too great but I could be wrong.

                By capacitors I'm talking about the 100uF, 470uF ones, not the really huge ones. From what everyone is saying it looks like I'll have to but I just basically wanted to know how many I need and what exactly they do. The second part I think I have figured out since you guys say they help smooth out the ripples. I thought the regulator was supposed to take of that so if you could please expand on that.

                JC-s60- I am confused as to y I can't wire it in parallel to the itps. The only thing I can think is that the current tries to jump through itps and skip the regulators or because the itps doesn't doesn't play nice when it's having to run off only 12volts so if you could please explain why this won't work I'd appreciate it.

                Thanks to all you guys for your help.

                Comment


                • #9
                  Originally posted by starfox
                  Mm just a note about National's switchmode ICs.. A friend built me 12V to 5V buck regulator using an LM2678, and we stuck the output on an oscilliscope with a 1A load, and the ripple voltage that we saw was quite bad.. We put a 4A load on it, and the ripple was around 100mA... which isn't suitable for anything like a PC...

                  Then again, the ripple might've been due to a bad board layout...

                  I don't think anyone here's tried a flyback themselves because it's quite hard and i think you have to wind a custom toroid, which is lots of fun.. >_<
                  Mmm. Bigger filter caps on the output!
                  System status: in progress

                  Comment


                  • #10
                    Originally posted by triwav
                    anyways, it was a breadboard I was looking at and it said 4amps max so I really don't know. What would be your suggestion to solve the problem. When I was studying up on what gauge wire to run to my amps, I learned that the main factor is distance and amps. This is a lot of amps but considering it's such a short distance, the resistance shouldn't be too great but I could be wrong.
                    Sorry still a bit confused on breadboards, are you soldering these components (breadboard to me means the white boards with holes and sockets that you push resistors and ICs into)? You'll have to solder these components, it's pretty dangerous not to..

                    Copper and other metal tracks have resistance too, just a lot less than most other components. When you start pushing a lot of amps through a conductor, the resistance of the component starts to become significant, and by P = I^2R, as the current goes up, the power loss through the device due to resistance goes up quite quickly. Just hypothetically (i'm making up these figures), if your cable or track has a resistance of 0.01 ohms, and you pass 20 amps through it, you'll get a voltage drop of 0.2V across the wire and 4W power dissipation. The power dissipated comes out as heat, and the wires / tracks can get quite hot after a while. To fix this problem, you reduce the resistance of the cable or track by buying a thicker cable, or by making the track larger.

                    Although i've said all that, it probably doesn't help you too much does it... If you're not sure, just make sure you can replace your cables later if you can feel them getting warm. Likewise, if the PCB which your electronic components're sitting on gets hot, solder extra copper wires on the solder side. Just be conservative when you first build it and don't choose a really thin wire.

                    Originally posted by triwav
                    By capacitors I'm talking about the 100uF, 470uF ones, not the really huge ones. From what everyone is saying it looks like I'll have to but I just basically wanted to know how many I need and what exactly they do. The second part I think I have figured out since you guys say they help smooth out the ripples. I thought the regulator was supposed to take of that so if you could please expand on that.
                    A regulator takes a dirty voltage, and drops it to a lower voltage but tries to maintain it at that lower voltage as much as possible. As the output voltage is constant, if you want to draw more power from the device, the current must increase. The increase in current on the output must be matched with a current increase on the input. If you have an fast change in the current drain (eg. you switch on a hard drive and it spins up, there should be a high inrush current when you start the motor up (?)), the power demands on the regulator increase.

                    As the change in power demands was quite fast, the regulator needs to obtain power from the input quickly. It's not possible to instantaneously change the power going to the input to the circuit (because there's some distance between the power supply and the input of the regulator, and this has a resistance), and for a very short period of time the voltage on the input will drop (while the current is held constant). This may drop enough to make the regulator stop working, so this is why you need capacitors on the input voltage side.

                    Similarly you need capacitors on the output side, though sometimes the output regulator capacitors are part of the linear regulator design.

                    Bleh that's a lot of words.. i hope that's useful

                    Originally posted by Change
                    Mmm. Bigger filter caps on the output!
                    Yeah i should try that..

                    Comment


                    • #11
                      Originally posted by starfox
                      Mm just a note about National's switchmode ICs.. A friend built me 12V to 5V buck regulator using an LM2678, and we stuck the output on an oscilliscope with a 1A load, and the ripple voltage that we saw was quite bad.. We put a 4A load on it, and the ripple was around 100mA... which isn't suitable for anything like a PC...
                      Isn't 100mV within ATX specs?
                      Progress: 80% - Permanent install left.
                      Motion LS800 Tablet PC and dock.
                      Vista, Bu-535 GPS, RoadRunner, MPT2006.

                      Comment

                      Working...
                      X