No announcement yet.

Question for Jeff...

  • Filter
  • Time
  • Show
Clear All
new posts

  • Question for Jeff...


    I am gonna try to build your ATX shut down controller (I have been using another electromechanical controller 'till, now that recently started being unreliable).
    The thing is that I don't need the part of your circuit with the relay that disconnects the PSU from the battery with some delay, 'cause I use STR. I could just build it and not use it, but I would prefer not to, for size reasons.
    So the question is: Will the rest of the circuit function properly if I build only the part in the square?

    Can I just leave Pins 1 to 6 of the 74HC14 IC free (open) or I have to modify the circuit somehow?


    [ 02-14-2002: Message edited by: Dimitris1976 ]

  • #2
    i think it should work ... dont connect the pin 10 to pin 3 and you will have to connect the pin 7 to ground cause that the gound for the ic ..

    if you want i can make one and check i have all the parts and a project board ..!



    • #3

      thanks for the offer, but I already bought the parts and I 'll build and test it tommorow myself. It won't take more than 15 mins on a perforated board.
      I found out you have been selling this controller for US$20. Was it just the board or the whole kit?


      • #4
        Originally posted by Dimitris1976:
        Will the rest of the circuit function properly if I build only the part in the square?

        Can I just leave Pins 1 to 6 of the 74HC14 IC free (open) or I have to modify the circuit somehow?

        [ 02-14-2002: Message edited by: Dimitris1976 ]</STRONG>

        It will function just fine. It is a good idea to tie the unused inputs high or low to prevent unwanted current draw. CMOS devices only draw current during transistions, and if the input are left open it wastes power.

        In Moahdib'd drawing, tie pins 1 3 and 5 to ground and leave 2 4 and 6 open. It will work just fine!

        If you want to understand the circuit, I find my drawing easier to read, and moadibs drawing easier to implement. I have a quick and dirty timing diagram of what is going on. The I's and O's on the lines are the input and outputs of the Gates.
        MPEGBOX - Plexiglass Computer


        • #5

          you are a treasure!
          Thanx, again.


          • #6
            hey 1976 they are ready to use ... types ..


            • #7
              Up to now, I believed that the power header on the ATX motherboards has two pins... one of them being the Power_on and the other a Ground. If you short the momentarily these two together the computer turns on/off.
              Jeffs, controller does exactly this - shorts the Power-on (+) pin to the Ground.
              It seems that it's not the case in all the motherboards!
              I built Jeff's circuit and then I realized that my motherboard (Supermicro 370SED) does not turn on/off like that.
              The only way to do it is to short these two pins together (not one of them to the ground), 'cause none of them is ground.
              So, I'll have to use an extra small relay and a diode...

              Advice: Be patient and test the circuit on a perf. board first in case you have to make any mods. I didn't do that and now I have to make a new PCB right from the start.

              [ 02-16-2002: Message edited by: Dimitris1976 ]


              • #8

                I built the modified circuit the way you told me to and it doesn't work.
                One other thing I noticed and it might help with the trouble shooting is that 78L05 transistor gets very hot-very quick as soon as I apply power to the circuit. It's in a plastic TO-92 package (the one without heatsink) 'cause I thought there won't be any significant current draw in this application..
                The resistance between the OUT pin of 78L05 and ground is 218Ohms with the IC on it's socket and the circuit is open when the IC is removed. Pins 1,3,5 and 7 of the 74HC14 IC are grounded as you told me to. The rest of the pins (8 to 14) are connected exactly as in Moahdib's schematic. Is there any chance that the circuit is shorted through the IC?
                I know it's hard to trouble shoot a circuit when you don't have it in your hands, but this is the last thing that has to work before I put the computer back in the car and I have tested and traced the tracks on the PCB and double checked for mistakes again and again without any success. So I am afraid that the modified circuit might have a mistake on the first place.
                Please, help!



                • #9
                  Check the resistance between power and ground of the HC14.

                  It is very easy to fry an HC14 I have fried many of them. It is my guess that your HC14 is fried. I would try another one. It will probably be a pain in the *** to remove it, but if you have a socket, that is the way to go. These are Cmos parts and if you give them a 12 volt input, it will fry the device. I did it just by breifly touching a alligator clip to the input and I couldn't figure out what happened for the longest time.

                  Also, the other type of atx power pin, you say that neither of them is ground. With the computer off, use a DMM, see if either of them is +5 or +3.3 If your meter beeps, or measure ohms. It is possible that it uses the supply as the ref instead of ground.

                  MPEGBOX - Plexiglass Computer


                  • #10

                    what about the resistance between power (14) and ground (7)pins of the IC? How much it should be on an intact IC?
                    I'll try a new IC... It IS a pain in the *** to unsolder ICs. Thank god I have used a socket.
                    As for my strange motherboard's power pins, I have connected a small relay and a diode on the transistor, the same way the big 30A relay is connected to the second transistor on the lower-right part of the schematic (the part I didn't built). Is it OK like that?


                    • #11
                      Hmm I'm not sure. More than 100 ohms or so.

                      Yes a relay like you mentioned is ok. But chances are you could do it with a transistor. It doesn't really matter. If pulling the line to ground didn't work, maybe pulling the line to vcc would work. I guess it really depends.
                      MPEGBOX - Plexiglass Computer


                      • #12
                        The IC was good. I have built a lot much more compicated circuits from schematics in the past, but can't make this one work.
                        I give up!
                        I am not in the mood to start over and make a new (third) PCB for this project. For the time I'll repair and use my old controller that did the exact same thing in another way. I might try again in the future when I have some spare time.
                        Thanx anyway...