Confused about snubbers, oscilloscope readings

Discussion in 'The Projects Forum' started by ebeowulf17, Jun 8, 2015.

  1. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    A while ago I got lots of great advice on this forum about how to reduce motor-related electrical noise in the espresso machines I work with. Improved grounding and snubbers made a world of difference for us. Along the way, I tried to do reading to better understand what was going on, and the general consensus seemed to be that motors mostly cause problems when switching off, not switching on. I also heard that the worst of the electrical noise wasn't directly from the back-emf of the motor, but from the arcing across relay contacts that the emf could produce - I was under the impression that the snubber preventing the arcing. None of this ever quite explained our scenario, where we most often saw glitches in our low voltage system when the motor turned on, not off.

    Last week I finally got an oscilloscope and tried to "see" the noise and test which of a few noise fighting strategies was most effective. Below are pics of my oscilloscope settings and the resulting traces.

    20150602_075707_Test-setup_2_small.jpg

    [​IMG]

    It seems like there are two very different noise signatures here. When the motor starts, there's a large, solid drop in the trace, about 4ms long and down 200mV at its worst. When the motor stops, there's a fuzzy "cloud" of high frequency noise for about 2ms, maybe 100-150mV p-p.

    From what I've seen and heard, I'd have expected the snubber to totally wipe out the cloud of noise when the motor stops, but it seems to have little to no effect on that. Instead, it does a remarkable job of eliminating the dip upon motor startup that most people seem to think shouldn't exist to begin with.

    I'm pretty confused and want to better understand all this. Can anyone help me make sense of this?

    Sorry forgot to describe setup: Oscilloscope is connected to +5VDC power and DC ground wires that power all of the control systems in an espresso machine which also includes carbonator pump motors (240VAC, centrifugal switch start.) Wiring schemes include switching 240V to motor with DPDT relay, switching only hot leg to motor through relay and leaving neutral maintained, switching both legs with relay and adding snubber, and finally switching just the hot leg with an SSR.

    One of our clients had noticed that using SSRs eliminated their noise issues, but later decided that the reason it helped was the maintained neutral, not the SSR itself. These tests would seem to contradict that.

    As always, sorry for rambling. Thanks in advance to anyone who gets through all this.

    [EDITED to reduce file size of pictures]
     
    Last edited: Jun 8, 2015
  2. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    Is the low level logic picking up the switching devices directly, what is the nature of the relay coils Ac? DC? Voltage?
    Is there any suppression across all coils?
    Is the low level logic galvanically isolated or are commons connected?
    Max.
     
    ebeowulf17 likes this.
  3. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    1) The relay controls are as follows: Main control board sends 5VDC signal through opto-isolator to control some switch (almost positive it's just a tiny triac) which controls the 240VAC to the relay. So there should be reasonable isolation with the opto, I think.
    2) Not so sure about suppression - there are a pair of capacitors on the 240 side of the relay switching circuit, but I don't remember any resistors, so I don't think it's making a snubber... unless one of the pieces I thought was a cap is a snubber. How small do they make integrated snubbers? These things are sub M&M size.
    3) Frame ground and DC common (what I called DC ground earlier) are connected at the point where the power cord enters the machine and ties to the frame. Here's a sketch I made back when I was working on all this the first time around:
    [​IMG]

    Also, Max, thank you very much for all the help you provided on my last go-around. I really appreciate it. I learned a lot then and I hope to learn a lot more now!
     
  4. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    Are you actually experiencing any problems right now, or are you just trying to ascertain the source of the original problems?
    Often when suppression steps are taken and cures a particular problem, there still may be some electrical noise on the machine but it now has little or none effect.
    Max.
     
  5. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    Mostly I'm just interested in better understanding what I'm seeing for my own curiosity.

    We *think* that we've solved the problems with snubbers, but can't be sure. We would see flickering LEDs here in our shop quite often, but never experience any machine malfunctions as a result. The snubbers definitely stop the LED flickers.

    The actual glitches and malfunctions happen almost exclusively to a client of ours in Australia, but even there it was so intermittent that it's hard to pin down. Their techs believed motor noise was the problem, so we started watching for any signs of motor noise over here.

    I guess my fear would be that the real problem for them comes from the noise when the motor stops, which our snubber seems to have no effect on. If that's the case, then SSRs might be the better solution.

    They've installed snubbers now in their most problematic locations, so we're waiting to see if that does the trick. Hopefully it will. Either way, I'm still interested in better understanding what the various patterns I'm seeing on the oscilloscope really mean, and whether we're really still one step short of truly killing the noise issue.
     
  6. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    I am also a member on another site for systems that implement AC motor control from a PC parallel port and many think that using a SSR will solve all noise problems due to isolation, the often overlooked factor is that the PP on a P.C. is not galvanically isolated due to the PC. P.S. 0v being grounded as well as the motor through the neutral.
    Generally following the bonding practices here http://www.automation.siemens.com/doconweb/pdf/840C_1101_E/emv_r.pdf?p=1 and by setting up a star point ground, solves most problems.

    I just realized I had posted this in your previous thread.:oops:
    Max.
     
    Last edited: Jun 9, 2015
  7. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    Yeah, that was a great article, a bit over my head in some portions. It helped me convince everyone here to adopt a much-improved (though still imperfect) grounding scheme which we've already implemented.

    Regarding the SSRs, the oscilloscope data makes it look like they really knock out a lot of noise that the snubbers don't. I'm not sure how or why they do it, and I don't know if it matters or not, but I'm always tempted to make things the best they can possibly be, and it appears as if SSRs are the best in this situation.

    I just find it odd that our experiences here seem so different from the norm. We expected negligible switching noise when starting motor, but that's when we see a huge voltage drop. The snubber seems to negate that drop, but I don't understand how. At the same time, the snubber seems to do nothing to the noise when stopping the motor, which I was led to believe would be its strong suit.
     
  8. Brevor

    Active Member

    Apr 9, 2011
    279
    54
    Where did you place the snubbers?
     
  9. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    If this drop is on the AC supply then it is due to the short circuit current that is inherent in a AC Induction motor started across the line.
    The motor represents an almost dead short until the inrush is over, any low resistance connection or small diam on conductors will aggravate this at start up.
    Max.
     
    ebeowulf17 likes this.
  10. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    Snubbers are connected parallel to the motor across the switched power to the motor. We've tried two different physical arrangements: in one case the snubber gets ring terminals and screws down onto the power posts in the back of the motor housing. In the other, we connect the snubber to the output of the relay, along with the cord to the pump. They should be electrically identical - the only difference being 6 ft of 16 gauge cord between the motor and the snubber in the latter arrangement.
     
  11. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    We're seeing the drop on the 5VDC supply (which is fed by the AC.) I wouldn't be too surprised if the DC drop is related to the AC drop you described. Although, if that's the case, can the snubber (or SSR for that matter) somehow reduce the inrush current? It doesn't make sense to me, but the snubber sure seems to kill that drop.
     
  12. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    Turns out I was a little off in remembering the switching circuit. A very similar circuit elsewhere in the machine has two caps in parallel to the triac-switched power to a coil like I remembered.

    However, in the case of the driving circuit which controls the motor-switching relay, instead of caps, there's a varistor in parallel with the relay coil. If my understanding is correct, the varistor would protect the triac and other components from back-EMF induced over-voltages, but not provide any noise filtering per se.

    So maybe back EMF noise from the relay coil, instead of from the motor, explains those fuzzy noise clouds when the motor is shut off. That would explain why adding snubbers to the motors has no effect on that particular noise.
     
  13. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,498
    2,364
    I have seen more problems attributed to lack of snubbers on Contactor coils than motor currents.
    Max.
     
  14. ebeowulf17

    Thread Starter Active Member

    Aug 12, 2014
    678
    79
    Ok, so that probably explains the high frequency noise when stopping the motor (and opening the relay.) Sometime soon I'll try snubbing the relay and put the scope back on it to confirm.

    That still leaves the question of what the big dip is when the motor starts. Inrush current causing a drop in voltage makes perfect sense to me, except that adding snubber to the motor eliminates the dip. Could a snubber somehow reduce the inrush current? It's not obvious or intuitive to me how that would work.

    Thanks again for all your insights and assistance.
     
Loading...