1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

Transient suppression for on-load tap-changing of 120 VAC ~10 A transformer

Discussion in 'High Voltage' started by jelliott, Dec 8, 2016.

  1. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    I've automated the very crude charger in my brother's 35-yr-old electric car, with relays that switch the 120 VAC input from one transformer tap to another, and ultimately pull power from the transformer entirely when the batteries are fully charged. Not surprisingly, this can produce substantial transients that I'm concerned may eventually damage the transformer. (The relays have already had to be replaced, despite being rated for higher currents than they’re actually carrying.) I’m looking for advice on the best way to suppress these transients. I had a SiBOD device that I found on eBay between each transformer input tap and the 120 VAC power source, which seemed to help (i.e. the car became less likely to trip the household circuit breaker when switching rates or shutting off). But then one of them shorted out such that it provided a path for power to continue flowing to the transformer after the relays had pulled power to shut it off. And since replacing that damaged one, the car has tripped the household breaker at least once when it tried to shut off, so this solution is clearly imperfect. I’m vaguely familiar with the concept of RC snubbers, but wouldn’t know how to properly spec one even if I did have complete specs on this transformer (which I don’t).

    Thanks in advance.
     
  2. Ramussons

    Ramussons Member

    Joined:
    May 15, 2013
    Messages:
    180
    Likes:
    13
    Location:
    Asia
    How about using Using Opto Coupled Triacs with Zero Crossing Switchover in place of the Relays?
     
  3. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    If I were to do it all over again, I'd probably try design a solution like that, or something similar (to provide zero-crossing switchover). But for the near term I was hoping that someone might be able to suggest something relatively straightforward that I could add to the existing circuits to suppress the transients sufficiently that my brother can plug the car in and walk away without worrying about tripping the household breaker when the charger changes rates or shuts itself off.
     
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT

    Hi jelliott,

    Can you post a schematic of your charging system and let us know about your mains supply and the charging currents/voltages involved?

    As far as I know, there are two reasons why a domestic supply will trip:
    (1) Excess current
    (2) Inbalace between the live and neutral currents

    It is also possible that your supply has a ground current trip.

    It would be important to determin the cause for tripping before a solution can be suggested.

    spec
     
  6. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    Here's a rough schematic of the charging system, showing the relays:
    Charger.gif

    The mains supply is typical North American 120 VAC @ 60 Hz. The charger provides the vehicle batteries between 5 A (full battery) and 20 A (depleted battery) at ~50 V (depleted battery) to 59 V (full battery) DC.

    The breaker in question will not trip on ground current. I don't know definitively whether it's tripping on excess current or on a live/neutral imbalance. But I do know that the breaker trips correspond to the switching performed by the relays, and that their occurrence was reduced by the addition of the SiBOD devices (not shown) between the transformer inputs and the mains power.
     
  7. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
  8. michael8

    michael8 New Member

    Joined:
    Feb 26, 2016
    Messages:
    3
    Likes:
    0
    Some thoughts:

    a. The rate-change relay isn't open before close and is thus shorting
    part of the transformer winding. The shorted transformer turn
    causes a large current flow (while the relay switches).

    b. The rate-change relay flashes over (sometimes) so it's opening
    before closing but the opening arc flashes to the other contact
    with the same results as (a).

    How easy would it be to use two relays with a time delay between them
    (and with some sort of interlock too so they both can't be on) instead
    of the rate-change relay?
     
  9. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
  10. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    Barring flash-over phenomena, the design of the relay precludes (a). But I'm reasonably certain that (b) is occurring (if you're standing around with the hood open after dark, you can see an arc when it switches--at least you could before I added the SiBOD devices). However, that doesn't explain the occasional breaker trips that occur when then shutoff relay pulls power--there's seemingly something else about the resulting transient spike that trips the breaker.

    Not easy, from my point of view. (I'm a mechanical engineer by training, and the analog circuit I've built to energize these relays in response to battery voltage is pretty much the limit of my self-taught electrical engineering abilities.) Additionally, I suspect that the transient voltage spike that comes from pulling power from the transformer would still be a problem, even if I could implement your staggered switching solution.

    If you look at the design of commercially-available electromechanical on-load tap-changers, it would seem that what I really ought to do is replace each of my relays with no fewer than four separate relays, to sequentially 1) Connect the mains power (in parallel) to the same input tap that it's already powering, but via a resistor, 2) Disconnect the direct connection between mains power and the first input tap, 3) Connect the mains power to the other input tap via a resistor (thus briefly shorting the transformer winding through the resistors), 4) Disconnect the connection made in step #1, 5) Connect the mains power to the other input tap directly, and 6) Disconnect the connection made in step #3. But I figure there ought to be a solid-state means of mitigating the transient voltage spike without resorting to that level of complexity.
     
  11. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
    Last edited: Dec 13, 2016
  12. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    I'm intrigued. At first glance, it seems like it wouldn't be applicable, since these filters are seemingly designed to prevent the washing machine from creating EM interference with other appliances. But I can also see how it might be successful in attenuating the transients that I'm dealing with. Would using one of these filters prevent the car from being charged by a GFI outlet that trips on ground current (i.e. many outdoor outlets in North America)? Hmm...
     
  13. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
    I don't follow what you are saying here because the objective is to stop transients being fed from the relays back to the consumer unit and activating the trip. The area of which you speak is electromagnetic compatibility (EMC) which covers radiated, conducted and electromagnetic coupling and electrostatic coupling both from the generator and susceptibility point of view.

    The filters have attenuation in both directions and are use on pretty much all UK washing machines. The filter's primary function, though, is to prevent the hash generated by the various motors and solenoids in washing machines from getting fed back down the power lines- the very function that you seem to need.

    Being a Brit, I'm not familiar with GFI outlets so I cant be authoritative, but I cant see that there would be a problem. The filters do not trip the ground current detectors (sensitive type) fitted to some UK domestic installations.

    This type of filter is ubiquitous, being fitted to most electronic and electrical equipment s, in one form or another. In fact, some IEC chassis mount mains inlet receptacles have a similar filter built in, and I know for sure that theses are used in the States.

    The easiest way is to suck it and see- you can get a filter for next to nothing from a scrap washing machine. If you are concerned about earth current, just leave the earth connection of the filter unconnected.:)

    By the way, I think there are some mains input filters that do not have an earth terminal, but maybe not on washing machines.

    UPDATE: I think the capacitors connected to earth on the schematic, which may be worrying you, are low value RF types which would not feed much current down the earth line- if I remember correctly that is.

    spec
     
    Last edited: Dec 14, 2016
  14. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    I was thinking that my problem had more to do with huge inductive transient voltage spikes briefly driving enough current through something (transformer windings?) to trip the breaker on current, rather than transients tripping the breaker directly. But that was just speculation on my part, so this filter is probably a good idea.

    Yeah, it looks like I can get one of these filters built into an IEC inlet receptacle that's otherwise identical to the on already on the car, for < US$10, so I'm going to go ahead and try it! (But I probably won't get the part in the mail until after the Christmas holiday, so it may be a while before I report back.
     
  15. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
    The filter built in to an an IEC receptacle will probably saturate (but there is no harm in trying it)- you need a much bigger filter like those fitted to washing machines.

    spec
     
  16. crutschow

    crutschow Well-Known Member Most Helpful Member

    Joined:
    Mar 14, 2008
    Messages:
    10,610
    Likes:
    478
    Location:
    L.A., USA Zulu -8
    How about adding one zero-switching SSR in place of the shutoff relay and momentarily open that during the relay tap change.
    That should prevent any relay arcing or transient surges.
     
  17. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    I guess I overlooked the relevant parameter(s) when I was shopping around for one of these filters to try--besides eyeballing the physical size of the thing, what should I be looking for to find one with more capacity? Capacitance?
     
  18. jelliott

    jelliott Member

    Joined:
    Nov 1, 2011
    Messages:
    101
    Likes:
    1
    Something like that certainly seems like the proper way to do this. But that would require design and fabrication of an entirely different (and probably much more sophisticated) control system; for the near term I was hoping that someone might be able to suggest something relatively straightforward that I could add to the existing circuits to suppress the arcing/transients. I suppose maybe I could turn each of the double-throw relays shown in the above schematic into a pilot relay that switches off one zero-switching SSR and turns on another one.
     
  19. spec

    spec Well-Known Member Most Helpful Member

    Joined:
    Oct 27, 2015
    Messages:
    5,828
    Likes:
    709
    Location:
    Somerset UK GMT
    Just go for a filter designed for a higher power inductive load, like a washing machine.

    spec
     
  20. crutschow

    crutschow Well-Known Member Most Helpful Member

    Joined:
    Mar 14, 2008
    Messages:
    10,610
    Likes:
    478
    Location:
    L.A., USA Zulu -8
    That should work also.
     
  21. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,042
    Likes:
    959
    Location:
    NJ
    Hello there,

    When something does not work right we look for the mechanism that is causing the problem. Once we find the mechanism, we can solve the problem with almost no doubt. This means no guessing.

    In this case there are only a couple reasons how this could trip a breaker. One is high primary current, and the other one is high secondary current which then causes high primary current. To find out which one, we would measure the current using a scope. That would tell us immediately which one is causing the problem.

    Without that measurement however it looks like the only things that can cause the problem is either:
    1. The relay arc over causes a short between contacts and therefore makes it look as if there is a partial short on one winding.
    2. Either relay opening along with partial arc over causes a boost converter syndrome which causes a high output current on the secondary which of course causes high primary current.

    Luckily the reason for either or both of these conditions is because the transformer primary is temporarily open circuited. To solve the problem then the solution is to stop the primary from becoming completely open circuited.

    Since either relay has to open for at least a short time, that means we should apply a load to the PRIMARY either just before we switch the main relay or just keep it applied continuously if we can stand the efficiency loss.
    The size of the load will vary depending on primary current, but even an incandescent light bulb might do the trick. That should get you going at least for the time being.

    Remember that if the primary is loaded correctly there will be no arc over.
     

Share This Page