confusion with sharing ground amongst circuits

Discussion in 'General Electronics Chat' started by patrick99e99, Aug 21, 2013.

  1. patrick99e99

    Thread Starter New Member

    Aug 19, 2013
    17
    0
    Hi everyone.

    I have a circuit that I built which contains a micro controller chip, a reed relay, and a diode.

    I have a molex right-angle connector on the board which receives +5v and GND, from there, wires under the board send that 5V and GND to the micro controller chip. The GND also goes to the reed relay's GND pin, as well as the switch (input) of the relay, and finally the GND goes to J3 of the 50v step up driver board (schematic is below).

    The micro controller chip then has an output pin which goes to the relay (to turn it ON). Then the output of the relay goes to J1 of the 50v step up driver board.

    So basically the micro controller chip sets one of its pins HIGH, and that in turn passes ground through the reed relay, and shorts out J1 and J3 with ground on the micro controller board.

    ... The problem I am having is my 50v step up driver board is not working properly when things are hooked up this way. If I run a totally separate ground wire directly to the reed relay's switch input, then it works perfectly.. But for some reason, when I share GND between the micro controller and the 50v step up board, it no longer works.

    To describe what I mean by "no longer works": The 50v driver board is basically firing a solenoid. And then the GND is shared between it and the micro controller, the solenoid "taps" like it's trying to fire.. but it can't quite do it...

    Once I run a separate ground to the reed relay (or directly to the 50v board for that matter), then it works fine, and the coil energizes.

    ... So I am trying to understand why this is happening and how I can fix it. I hate having to give my circuit two paths to ground because that feels wonky and gross (and also is inconvenient). So I am wondering if somehow it's a matter of noise? Do I just need to get a big fat resistor and put that between ground and the reed relay?

    Can anyone shed some light on this problem?
     
  2. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,548
    2,373
    I see no reason why you cannot common up the respective supply commons, I would connect them at one central point however, and also use a 10-20μf decoupling capacitor on both your 5v and 50v circuits, or up to 50-100μf on the 50v side.
    Max.
     
  3. patrick99e99

    Thread Starter New Member

    Aug 19, 2013
    17
    0
    Decoupling capacitors are the electrolytic "aluminum can" looking ones?
     
  4. patrick99e99

    Thread Starter New Member

    Aug 19, 2013
    17
    0
    And are you saying to run the ground through the capacitor?? like ground goes in one lead of the capacitor, and the other lead goes to the micro processor's ground pin?
     
  5. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    No.

    You may be having a power supply problem, not a grounding problem.

    First things first.

    Is there a diode across the relay?
    Do you have 0.1μF capacitors strategically placed across all your circuits?
    Do you have 10-47μF filter caps across the power supply rails?
    Do you have 470-1000μF reservoir caps across the power supply?
     
  6. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    I did not look at your circuit before.
    Try reducing the value of R2.
     
  7. patrick99e99

    Thread Starter New Member

    Aug 19, 2013
    17
    0
    Yes.

    No? I only built one circuit here that is integrating with the schematic that I attached. I described exactly what my circuit has on it. Why do I need this capacitor?

    I didn't build the power supply, so I have no idea.
     
  8. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,548
    2,373
    De-coupling capacitors act a shunt to noise or EMI across the power rails.
    If the board is already built it can be tricky, but you could at least put one on each power supply where it enters the board.
    Usually well designed circuits have them built in.
    Max.
     
  9. patrick99e99

    Thread Starter New Member

    Aug 19, 2013
    17
    0
    so, if I took a 20μf electrolytic capacitor and hooked 5v to the + terminal, and GND to the - terminal, and then sent the GND from that point over to the 50v step up driver board (the schematic I provided in my initial post), do you think that will solve this problem?
     
  10. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,548
    2,373
    It certainly will not hurt, I would also place one on the 50v, as previously mentioned, your physical wiring layout of power supply common's also can also have an effect.
    Max.
     
  11. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,013
    1,531
    I know I'm pretty dumb and even dense in understanding this stuff. But what even is the purpose of this circuit? This is a real question about this circuit, trying to learn from this.

    The way I see it, when there the 50V supply is active all of the mosfets are turned on. Why even use this circuit?

    I thought something that's called a "driver", should be used to drive something? What I mean is shouldn't the gates be controlled by some signal other than the 50V supply?
     
  12. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    The combination of D1, R1 and R2 could be causing the problem.
    Try deleting R1 and replacing D1 with a 10kΩ resistor.
     
  13. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,013
    1,531
    I was serious about this question. I've never come across a circuit like this that uses the supply voltage as a gate turn on before. Could someone explain for me?
     
  14. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    Maybe it is a circuit designed to be used with a relay contact, switch or push-button.
    Short J1-1 to J1-3 to activate the output relay.
     
Loading...