controlling meanwell with relay on 120VAC side - relays keep getting stuck closed!

andrew.box

100 W
Joined
Apr 15, 2011
Messages
154
Location
kansas city, USA
Hey folks, I'm trying to drive a meanwell 12V 18A PSU, model S-250-12, using a standard type relay like one would use with arduino to switch the 120VAC mains to the power supply on and off. Not an SSR or anything, regular relay that goes 'click'. Well, I've had 3 relays now fail closed with this project which leaves the PSU and load turned on all the time, not what I want to do!

After the 2nd failed I figured it may be inrush current welding the contacts closed, relays are 10A rated and my load on the 12v side is only 5A but inrush maybe would be higher, so I put a 2 ohm 25A rated thermister in series with a leg of the AC mains between the relay and meanwell. But it happened again, I just found the relay stuck closed. I'm stuck on this as well, it can't be 3 bad relays at this point.

Anyone ever had this issue or have a better idea that me what's going on?

Full disclosure - this isn't being used as a charger in this project but rather to drive a 12v resistive heating element for an incubator for a DIY creature rearing setup, but I didn't know where else to turn and figured this may be useful to someone building an arduino controlled meanwell based charger someday. If it's too off topic feel free to remove.
 
Not sure if it’s arcing. I don’t hear anything. This one is enclosed though inside a whole temp controller unit. I put an clamp style meter on it when I first built it and it was showing 5A on the DC side. I think there’s inrush because my workshop lights dim ever so slightly when the relay clicks closed. I’m guessing it’s a capacitor charging since the load itself isn’t really that big. 60-70w or so.

Thanks for the reply!
 
if it's this one
https://www.mouser.com/ProductDetail/MEAN-WELL/S-250-12?qs=A0aqJZyP9eUB6nG3oMOuGg%3D%3D
then the spec sheet linked there says it's a 45 inrush current at 250vac, which means up to twice that for 115vac.

so you have to use a relay capable of at least 90a or it will arc while closing the contacts, and weld them together.

it's probably a lot easier to switch the output than the input.

if yours is not the same model, look up your model on mouser or meanwell's site, and verify the inrush current you'll have to deal with.
 
amberwolf said:
if it's this one
https://www.mouser.com/ProductDetail/MEAN-WELL/S-250-12?qs=A0aqJZyP9eUB6nG3oMOuGg%3D%3D
then the spec sheet linked there says it's a 45 inrush current at 250vac, which means up to twice that for 115vac.

so you have to use a relay capable of at least 90a or it will arc while closing the contacts, and weld them together.

it's probably a lot easier to switch the output than the input.

if yours is not the same model, look up your model on mouser or meanwell's site, and verify the inrush current you'll have to deal with.

Oh wow, yeah that’s probably the issue. So even with the 2 ohm thermistor you think it would draw more than 10A or so? Maybe the thermistor isn’t getting long enough to cool down, I guess that would do it.

Yeah, I’ll just build myself something to switch the 12v load side. I think that’s the way to go. I never noticed they have specs published for the inrush. Thanks a bunch! I guess I just naively assumed it was a few amps and nothing too extreme since they’re found so commonly in all kinds of devices.

Thanks again.
 
andrew.box said:
So even with the 2 ohm thermistor you think it would draw more than 10A or so?
you'd have to calculate it based on the voltage you're running it on.

if it's 115vac, then 115v / 2ohm = 57.5a.

you'd also have ot look at the ntc / ptc 's specs to see how long it takes to heat up and become lower resistance, to see if that is as long or longer than the inrush current time will be, since the inrush will be slowed down and take longer due to the lower current flow.

then you have to get a relay that can handle the actual inrush current that's still present.
 
amberwolf said:
andrew.box said:
So even with the 2 ohm thermistor you think it would draw more than 10A or so?
you'd have to calculate it based on the voltage you're running it on.

if it's 115vac, then 115v / 2ohm = 57.5a.

you'd also have ot look at the ntc / ptc 's specs to see how long it takes to heat up and become lower resistance, to see if that is as long or longer than the inrush current time will be, since the inrush will be slowed down and take longer due to the lower current flow.

then you have to get a relay that can handle the actual inrush current that's still present.

Ah, right. Just ohms law if voltage is in RMS. Been a while since I’ve done much circuits work. Forgot my basics. Yeah, 45A is still too high for the presumably 10A relay in that device. Will change tactics accordingly. Thanks for the help.
 
andrew.box said:
I guess I just naively assumed it was a few amps and nothing too extreme since they’re found so commonly in all kinds of devices.
if you mean, the meanwells are found commonly (or rather, psus of various types similar to the meanwells), then yes, but they aren't being switched by relays in hardly any of them--that's a very rare thing.

almost always the dc side is switched, and usually it's done via a shutdown line, like on a computer, that simply "tells" the psu to stop outputting power.

unfrotunately your meanwell is one of the oens that doesn't have this option wired to the terminal block (dunno if it has it internally, but based on the block diagram in the spec sheet, if you can find the ovp section's optoisolator input, you could splice into that signal to control it's shutdown / power on without a big relay.
 
do you need actual isolation or do you just want to turn it off?
 
then just use a SCR.
 
12V SPST relay sounds like the tool
 
amberwolf said:
then the spec sheet linked there says it's a 45 inrush current at 250vac, which means up to twice that for 115vac.
Did you mean half? Actually it's more like a quarter due to the CV^2 factor. Keep in mind the input capacitance does not change when voltage does in a universal supply.
 
A SSR would work well for that application. The inrush of the capacitors charging draws enough current to weld the contacts.
 
andrew.box said:
On the low voltage side, or even to switch the mains?
I'm thinking to switch the mains. The big AC ones can handle the inrush and they aren't super expensive.
 
fechter said:
andrew.box said:
On the low voltage side, or even to switch the mains?
I'm thinking to switch the mains. The big AC ones can handle the inrush and they aren't super expensive.

Thanks, I'll look into it. Good option to have even if I just switch at the 12v for this project.
 
I see some of those have a DC rating, only 0.25 A, as opposed to 40x the AC current.

Is DC switching under load really **that** much more stressful to mechanical relays?
 
Yes, because under AC the voltage (causing the arc) drops to zero at the frequency of the AC.

DC never stops arcing, cuz it's DC, until the contacts are far enough apart that under that specific voltage the arc won't sustain.

But when the contacts are *closing*, then either one can weld the contacts closed and prevent breaking the circuit, if the arc current/temperature is high enough to cause the contacts to heat to molten temperature at the point of contact at the time they close.
 
So for this project https://endless-sphere.com/forums/viewtopic.php?f=14&t=98599&p=1444529#p1444529

just focusing on the contactor bit, what would y'all recommend for 4Vdc

A. for initial experimenting at 10A maximum

B. later for production use, at 30-35A

?

Any potential derails about the rest of that thread's ideas should best get posted over there
 
Back
Top