 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Relays and 60v DC
| Author | Message | ||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Recently converted my solar hot water from thermal panels to solar panels. I'm using 6 60 cell panels wired in series parallel to drive a 48v 1000w heating element. The open circuit voltage from the panels is about 65v. Initially I tried a 100 amp relay to switch the panels out once the water tank thermostat reached temperature. This relay vaporised its contacts the first time it released. Now I guessed the near instantaneous rise in voltage across the relay contacts started a sustained arc, fed by high voltage DC. So I grabbed another relay, and fitted a suppression capacitor/resistor across the contact. This was a 100 ohm resistor in series with a 0.1uF capacitor. Again, the relay contacts vaporised. I dug up a big 600v 40a MOSFET I had recovered from a dead inverter welder, mounted it on a heat sink with a fan. Its working fine, but the fan needs to constantly run. I want to get back to using a relay. Its simple, and no fan. The capacitor in series with resistor across the contacts didn't work for the last relay, so I'm rethinking the problem. Maybe a diode, a 0.1uF capacitor and a 1k resistor in parallel, across the contacts? Yes there would be some current flowing through the resistor when the contacts are open, but only a few milliamps, certainly not enough to have any real effect on my tank of water. The capacitor should dampen the initial spike, the diode should stop the first back emf spike, and the resistor should load the circuit enough to stop the high voltage arc. Any thoughts? I'm quickly running out of relays. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
wiseguy Guru  Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Gizmo, welcome to mechanical DC switching. The issue you are having is not too dissimilar to plasma welding. Once the arc begins the little ball of flame in the contact gap conducts almost like a solid conductor, maintaining the heat and plasma arc, successfully turning your relay contacts to gas. There are a few methods to counteract this issue, one of the simplest would be to use the mosfet in parallel with the relay contacts then with a bit of simple circuitry turn the mosfet on just before you open the relay contacts and leave it on for say 25 - 100mS after the contacts open before turning off the mosfet. If the relay is a change over maybe some passive circuitry using the normally closed contact could turn off the mosfet. The mosfet wouldn't even need a - heatsink for this. That way the relay is not passing current when the contacts open, I believe it is better to allow the relay contacts to pass current when closing to keep them "clean". I first discovered this when a project needed switching ~ 100V DC through a 30A relay rated for 415VAC. I thought 100VDC should be a cinch. I turned off the relay but my 20A was still flowing - looking at the relay revealed a blue greenish glow inside just befoe the plastic case combusted. I have seen an LY4 relay with 4 changeover contacts all wired in series that worked but not recommended. other systems use magnets or air jets to extinguish the arc. Edited 2020-04-02 10:24 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Mike has summed up the problem very well. The mosfet will turn on super fast and beat the relay by a mile. So when the contacts close there will be only a very low voltage across the mosfet to short out. The trick is then to make the relay release first so that the mosfet does all of the turning off. Mosfet turn off needs to be pretty rapid (but delayed) to prevent very high momentary dissipation in the mosfet, so it will need some kind of monostable time delay circuit driving the gate. Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1578 |
I have played/ experiment with relays quite a bit and come to some conclusions. I visualise DC as having Momentum. It only travels one way unlike AC so once it's gets going, is like a train and does not want to stop. Instead of trying to stop the runaway, you merely divert it to somewhere it can do no harm. The trick I found with DC is not to turn it off but to Divert it to another ( Dummy) load. I found that the Dummy load can be less than 50% of the original load and you won't get arcing. I use DPDT relays with the Dummy load on the normally Closed side. If the input is low enough and the relay contacts rated high enough, You can go to a dead short on the second contacts and the relay will switch fine without arcing. Doing this pulls the panels down off their curve so you end up with less power being shorted than what is supplied to the load anyway. You could also in this case easily use something like a 500W halogen tube as the Dummy load or an old bar radiator etc. I would use the water heater thermostat to switch the low side of the relay to take the load off it and transfer the current to the DPDT relay. Playing with the heater thermostats, I have found them to be pretty weak over all and only just up to the job they are designed for. You can wire both contacts of the relay in parallel for greater current carrying capacity as well as use a relay for each set of poles. I Disconnect both poles with a double AC breaker on the dc side of one of my solar inverters which is 3.6 KW and WELL overclocked on the panels and done it for over 2 years with no problems. That's my test setup so I have probably switched that under full load more than 50 Times. Switching both poles in my experience goes a long way as well with preventing DC arcing. I found the relays with the diversion load/ short will switch with no more than a fat spark the same as they show with a decent AC load . I ran one for days on a flip flop and saw no degradation of the contacts. I was running around 270V @7A with 2x 500W floods as the Dummy loads. I think that one may have been enough but I wanted to test the base theory not the low limit for the ballast. The LY? type relays with the base are Cheap on fleabay and even if you swapped the relay out every year would still be a cheap and effective way of doing this. Parallel the contacts and put a relay on each pole and for what you are switching, I'd be pretty certain you wouldn't have a problem without a dummy load at all. |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
I have a spare solar panel here, I'll do some testing over the next few days. I really want to keep it simple. I understand diverting the power to a dummy load to keep the open circuit panel voltage under control, but a capacitor across the contacts should do the same. A capacitor will act as a short circuit for a few milliseconds, then slowly ( relatively speaking ) let the voltage across the relay contacts rise to 60v. A low value ( 5 ohms ) series resistor should limit discharge current when the contacts close. Once the sun comes out, I'll test a sudden open circuit, then one with a capacitor of several values, a resistor, and a diode. See if I can film it with my phone. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
Glenn, see this article for selecting the RC values for a arc suppression snubber, Link. Cheers Mike Edited 2020-04-02 20:04 by Solar Mike |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Thanks Mike, good read. Looks like I needed a much much larger capacitor than the one I was using. My voltage is 60, and it could be drawing up to 22 amps when it goes open, so really I should be using a resistor around 47ohm and a cap in the 15 to 22uF range. I like the idea or two contacts in series, especially in case one set gets welded. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
It depends on how much voltage and current you plan to break. The capacitor needs to slow the rate of voltage rise sufficiently across the slowly opening contact to prevent an arc from forming. Its a matter of physical distance across the opening contact, instantaneous voltage, and time. An RC network will certainly help with that up to a certain point in a very light duty application. But keep in mind that with the contact open, the capacitor will then be fully charged, and when you close the contact again, there will probably be an almighty "splat". Its not usually possible to find compromise values of both the resistor and capacitor that can completely eliminate both the arc and the splat. If you wish to try this method, its going to require a pretty high value electrolytic in series with a suitable diode, directly across the contact, so that the electrolytic is able to charges at full available current through the diode when the contact opens. If the capacitor is large enough, that should prevent an arc from forming. When the contact closes, the diode prevents the electrolytic from rapidly discharging. So you need to add a resistor in parallel with the diode so the electrolytic can discharge slowly and safely (over a few seconds) through the closed contact. Cheers, Tony. |
||||
| dwyer Guru Joined: 19/09/2005 Location: AustraliaPosts: 574 |
hi Glen Have you try by putting whole relay into a jar of Castrol Transmax Auto Trans Fluid TQ 95 1L that use for Ford Fairlane Falcon Fairmont AU that design for electronic auto transmission Cheers dwyer |
||||
 Print this page |
