Protect microcontroller from contactor noise

Discussion in 'General Electronics Chat' started by Scott216, Dec 9, 2013.

  1. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    I have an Arduino microcontroller controlling my hot tub. There is a big contactor that turns on the heating element (230 VAC) and I'm sure some nice electrical spikes are generated. I've already seen it mess up the LCD display on the Arduino. The Arduino output is connected to an optically isolated relay which in turn energizes the contactor. I'd like to find out what I should do to protect my Arduino from electrical noise the contactor makes when it energizes and de-energizes. I'm building a PCB that the Arduino will plug into, so I could put some protection/filter circuits there. Any suggestions?
     
  2. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    Are you using decoupling capacitors?
     
  3. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    I'm not using anything. The power supply for the Arduino is a pretty good one. I'm using Omron S8VS-01512 (http://bit.ly/18wTqWU)
     
  4. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    You need a R/C snubber across the coil.
    Another way is to switch the load with a Triac using a zero cross Opto.
    Or a off the shelf Opto22 SSR unit.
    Max.
     
  5. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    I'm betting the contactor and relay will have an instantaneous draw of more than 1.2 amps...

    Are you seeing a voltage dip when switching the contactors?

    Add some decoupling capacitors near your ICs.
     
  6. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    If you already have an opto to control the relay, it is a relatively simple to replace relay and contactor with a Triac, or for a mere $40.00 you can replace the lot. http://www.opto22.com/site/pr_details.aspx?cid=3&item=240A25
    Working with industrial controls, the worst I have found is Contactor coil kick back for upsetting low voltage equipment.
    Max.
     
  7. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    My power supply is 12 volts and the Arduino only needs 5, so it could have a pretty big voltage drop before it would be affected by that. So far the Arduino seems to be running okay and I don't think it's been affected, but I have only had it hooked up for testing and I'd like to make sure my setup is a little more robust. The only problem I've seen is my LCD display would become garbled. I changed how the display is wired to the Arduino and I haven't seed that problem again. I do have an optically isolated relay between the arduino and contactor coil.

    I thought about a solid state relay, but they get pretty pricey, especially double pole. My hot tub heater draws about 22 amps so I'd probably want a 30 or 40 amp relay.
     
  8. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    That would probably work. I'd probably want to use two of them, one for each 220VAC leg. I'd also need a big heat sink, right?
     
  9. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    Ideally you would want one in each leg for N.A. installations.
    It should be a lot cheaper if you use a zero cross opto such as the MOC3040 or later versions, and a couple of high current triacs.
    Max.
     
  10. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    I think for now I'll stick with the mechanical contactor and see how that goes. From what has been posted so far the recommendations are: 1) Use decoupling capacitors. That only helps if voltage drop to the Arduino is a problem, right 2) Put a R/C snubber across the contactor coil. Since I already have an opto-isolated relay controlling to the coil, does a snubber still provide some benefit?
     
  11. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    Preferably you want to kill the source, rather than do a patch after the fact, but any of those will help, it depends on what opto you are using, the coil could still be a problem, you can also get line snubbers to cure any contact arc causing it.
    Max.
     
  12. ScottWang

    Moderator

    Aug 23, 2012
    4,858
    767
    I think the electrical spikes was affected from the power line, you can using two 22~50uH Magnetic Core Inductors to in series with the +5V and GND.
     
  13. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    I agree, I'd like to kill the problem at the source if possible. I've never used a snubber. How do I go about figuring out what to get?
     
  14. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    Here are a couple
     
    bug13 and Scott216 like this.
  15. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    Thanks for the links to the snubber info. I'm a little confused about the recommendation made on page 2 where they say
    "Electrocube has determined that the best overall combination is 0.47-.50 Mfd @ 220"
    Is the 220 the line volts or resistance? I'm assuming it's the RMS line volts. If so, then how do I calculate the resistance? I tried the following, but it didn't seem right.

    If I use their formulas (see attached), which they say aren't necessarily the best way to go because of other factors. But here's what I get:
    The current through the hot tub heater is about 22 amps. So the capacitance is 48.4uF. Which is a lot more then their 0.47uF general purpose recommendation. For voltage I used 220 * 1.414 = 311 to get the peak. Based on this the value for resistance which would be 7.85 ohms

    I'm assuming in the formula that I need to take the 3.16 root of C; but I wasn't sure about that.

    Here's the Excel formula I used (amps = 22, volts = 311):
    =311 / (10 * ((22^2)/10)^(1/3.16) * (1 + 50/311))
     
  16. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,571
    2,381
    220 is RMS line voltage. I interpret it that the peak voltage is required for the voltage value for the capacitor.
    Also the snubber PDF's are based on inductive loads where there is stored energy that has to be dissipated, this does not occur in resistive loads, so I would consider the the only source to worry about is contact arc from a resistive load.
    Placing a snubber across each contact may suffice in your case?
    Max.
     
  17. Scott216

    Thread Starter New Member

    Jun 16, 2011
    18
    0
    That's what I want to do, but I don't know what size snubber to get.
     
Loading...