1

I am trying to design my electric vehicle charging station. There are some examples on net but they are designed by using relays at live and neutral outputs.

The main purpose of this design is to cut both neutral and live output.
Trying to implement this with triacs and zero-crossing optocouplers. On proteus sim it seems to be working but I would like to be sure if I need to add/change this design.

Is it OK to use TRIACs like this? Any comment/help would be appreciated.

Note: Proteus screenshot below is just to simulate power side of things. Ground fault detection and pilot signal circuit will be added on final design.
Also ground at bottom added for crash proof on sim. Won't be there on the PCB.

Proteus Screenshot

Edit 1:

Thanks for all comments, really helping me out.

After MikroPower's answer, I made below changes: sch v2

So I think I need to control both high current relays.

  • When enabling output:
    SW1 & SW3 ON, after few secs SW2 ON & SW1 OFF.
  • When disabling output:
    SW1 ON & SW2 OFF, after few secs SW1 & SW3 OFF.

Also tried reverse mains on sim, and seems to be working. sch v2 reversed

So what do you think about last design?

greybeard
  • 1,469
  • 1
  • 4
  • 17
Onur Akıncı
  • 11
  • 4
  • Triacs, in my experience, generally fail short-circuit. What will happen your setup if that happens? Relays can be more fail-safe in this regard. – Transistor Mar 26 '23 at 21:41

2 Answers2

4

It's not a good idea to put TRIACs in series, this will not work.

You possibly want to protect the switches from damage and burning with this method. What you can do is to use one TRIAC to stop the current flow and disconnect the neutral and live wire with an ordinary relays. If there is no current flow, then the contacts of the relays will have a very long endurance.

With a latching relay you could save some more energy.

The TRIAC has a voltage drop of 1.5 V and with a load of 16 A there is 24 W of heat that you have to cool away. A method to stop this voltage drop is to use a relays in parallel to the TRIAC.

  • If you want to close the connection, first you close the TRIAC and then you close the relays to prevent the voltage drop.
  • If you want to open the connection, first you open the relays and then the TRIAC.

With this method you have a minimum of energy loses and a high endurance of the metallic contacts.

winny
  • 13,064
  • 6
  • 46
  • 63
MikroPower
  • 1,338
  • 2
  • 11
  • @greybeard - I think if he is switching an inductive load he will add a snubber. He want to make the circuit as easy as possible only to understand does it work like desired or may be not. – MikroPower Mar 26 '23 at 21:43
  • It's possible to switch multi-digit Amperes and forget about the inductance of a multi-meter cable, coiled or not. – greybeard Mar 26 '23 at 21:45
  • I am using mostly "snubber-less triacs" for 230V AC, like BTA16 from ST. They are rated for 600V - 800V, but I am always not sure and add a snubber only to be on the save side. – MikroPower Mar 26 '23 at 21:57
  • main purpose of this design is to cut both neutral and live output. Because my ev charger will be a portable one with standard EU mains socket, so live and neutral lines in circuit can be visa versa. Relays or relays+triacs can be used as well but i was concerned about contact failure as you said. – Onur Akıncı Mar 26 '23 at 22:23
2

NOT a good idea, relays is chosen correctly can provide safety isolation, no semiconductor switch does that (And yes, I have been bitten by an output fed from a turned off triac, the things leak a little).

You have a safety function here, and you want to use devices known to have a tendency to fail short circuit?

A relay or contactor ideally with force guided contacts is in reality the right thing here, as it can not only provide the safety isolation gap required but can also be designed to support a monitor output such that a welded relay contact can be detected and can cause a secondary protective system to trip.

Do watch the requirements on the earth monitoring stuff by the way, some of it is gnarly as all get out, involving sensing DC faults and the like.

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
Dan Mills
  • 17,266
  • 1
  • 20
  • 38
  • 1
    I've replaced your "pithy anglo-saxon term" (quote: Larry Niven :-) ) with a quaint and time tested one. You may wish to alter it to some other site-acceptable term. || I understand that that terminology is unremarkable in many quarters, but as a site we need to not step onto the slippery slope of variably acceptable terminology. And the material needs to be acceptable in any forum, although few liable to read it would be liable to be offended. – Russell McMahon Mar 28 '23 at 04:27
  • (Not offended, but displeased: I'd rather not see such language, and not on SE.) – greybeard Mar 28 '23 at 06:03
  • @greybeard Indeed. The code of conduct does not allow such terminology, even though it's excessively widespread "out in the world". Lary Niven used "TANJ" (There ain't no justice" where more common epithets may otherwise be used. On one occasion he allowed himself to remark as I noted above re what his character said but which was not printed. – Russell McMahon Mar 28 '23 at 09:53
  • TANJ? Going to have to remember that one. – Dan Mills Mar 28 '23 at 09:55
  • @DanMills [TANJ](https://www.google.com/search?q=tanj+%2Bniven&rlz=1C1CHBF_enNZ834NZ839&sxsrf=APwXEdfzpPVMOoSHkn0UsQLUJN-UcYckFA%3A1679999785292&ei=KcMiZIjAEYmOhwP7_pKQBg&ved=0ahUKEwjI-LLztv79AhUJx2EKHXu_BGIQ4dUDCA8&uact=5&oq=tanj+%2Bniven&gs_lcp=Cgxnd3Mtd2l6LXNlcnAQAzoKCAAQRxDWBBCwAzoGCAAQHhANSgQIQRgAUJ0IWLAsYO0vaAFwAXgAgAHABIgBgBSSAQkwLjkuMi41LTGYAQCgAQHIAQjAAQE&sclient=gws-wiz-serp) – Russell McMahon Mar 28 '23 at 10:37