I m using relay 230vac/10amp to switch on/off LED driver. But sometimes relay sticks. Loads are nowhere close to 230v*10amp=2300watts. I cant understand what is causing this. i tried connecting an NTC 10ohm, assuming it could be from inrush current. But still the relay contacts are getting stick. And after opening one of them, i found contact were just welded together. please suggest what can be done and what is causing this.
-
2\$\begingroup\$ We can't know which relay you are using, which LED driver you are controlling with the relay, and with what circuit you are driving the relay. When done incorrectly or relay ratings are exceeded, the contacts can weld together. A relay labeled with 230VAC and 10A may not be able to switch 230VAC at 10A, and the relay driver inrush current may be tens of amps if it is capacitive load. \$\endgroup\$– JustmeCommented Apr 3, 2022 at 10:26
-
\$\begingroup\$ Is it a CCS LED driver? Can you give us the link? Usually the cheapy but popular Arduino/Rpi 5V relay module specifying 230VAC 10A can only handle continuously 8A-. \$\endgroup\$– tlfong01Commented Apr 3, 2022 at 10:41
-
\$\begingroup\$ relay is being drived by nud3112 and led driver are not fixed. it depends on end led light type. \$\endgroup\$– Design4ecCommented Apr 3, 2022 at 13:49
1 Answer
This is happening because of the inrush current drawn by LED drivers. There are capacitors in the input section of LED drivers and they need to charge up as soon as you turn the relay ON. This initial charging incident will draw tens of amps for a few milliseconds and that ends up welding the contacts.
You have a few options to prevent this from happening:
Use LED drivers that have lower inrush currents.
Use relays rated to handle inrush currents. These will have two ratings - one for steady current for ex - 16 amps and a second rating for inrush current such as 80 amps.
Use an in-line inrush current limiter - there are commercial products available that reduce the inrush current drawn by these LED drivers. Search for them and see if they are right for your use case.
Use triacs (preferably with zero cross switching) - Triacs are solid state devices and have no moving parts. These should be able to handle the LED drivers without getting damaged. However, there will be a voltage drop across these components and these will get hot at higher currents. You might have to provide heat sink or cooling of some sort.
-
\$\begingroup\$ LED driver type is not in control, as it depends on user in this case. I read somewhere NTC can be used for this. What are your views on this ? \$\endgroup\$ Commented Apr 3, 2022 at 13:50
-
\$\begingroup\$ NTC needs to get cold to develop the resistance. Capacitors lose charge faster than that. So if there is a power cut and resume in 2 seconds, caps would be discharged and NTC still hot. That will cause relay failure. You can develop logic around it but it wont be worth it. You will be ruining user's experience. For ex - if there is a power cut, you dont turn on the relay until 10 seconds have passed. As a user, 10 seconds wait is too much. I am not gonna use your product. \$\endgroup\$ Commented Apr 3, 2022 at 15:26
-
\$\begingroup\$ Agree to this point . But I have no other options than to implement NTC as of now . And lots of applications used NTC for inrush current management. How did they manage , I was thinking \$\endgroup\$ Commented Apr 4, 2022 at 3:31
-
\$\begingroup\$ @Design4ec - are you making home automation products using relays and facing sticking issues with clients and want to solve the problem using a series NTC? \$\endgroup\$ Commented Apr 4, 2022 at 15:49
-
\$\begingroup\$ Wanted to test NTC for failed relays . As i read many times that NTC helps in these problems, which didnt actually worked for me. But found another solution. When used another make relay of same rating, everything worked fine. Even without NTC or anything. the earlier relay seems to have some problem. \$\endgroup\$ Commented Apr 11, 2022 at 16:35