Zombie Relay: Is it deffective? Is it me? Is it undead and might kill me while I sleep?

[WoodsieLord]

** before I forget: my grid runs at 220VAC, single phase. **

I have bought an 8 pin relay. The purpose is this: I want a small fridge to run on solar by day and use the grid by night. This would save power during the day and preserve my already-very-small lead acid bank when no power is being harvested.

The fridge load is quite small (150W ~ 200W, when running) and I've been running it off solar for quite some time without issues (except that the bank gets drained a bit too much for my liking).

Relay characteristics:
- It states it can switch 10A @ 220V AC.
- The coil is also rated at 220V AC.
- Not stated, but it's supposed to be designed for continuous use.
- It came with a DIN socket.




I installed the Relay so that when the coil is not energized, the two Normally Closed pins connect the grid to the fridge. When the coil is energized, the fridge gets connected to the inverter through both Normally Open pins.

The problem: When energizing the coil, the relay goes "zombie" after a while (a while being between 20 min ~ 1 hour sometimes a bit more).
'zombie': The relay won't close normally open nor normally closed either!. (fridge powers off)

I've tried many times, leaving the relay in the 'zombie state ' from a few minutes to at least 3 hours. The relay gets a bit hot but not to the point of burning my fingers.
When I don't energize the coil and the fridge runs on the normally closed pins, the relay works cool indefinitely.

The relay quality is not great, but I can't seem to find a better one.

now...
. . . If the relay could not handle the current, it would overheat even when the coil is off and runs on the normally closed pins.
. . . if the coil was mislabeled and was actually 24V or 12V, using it with 220V for such long test periods would have burned the coil. But the coil keeps coming back to life after I let it cool down.

Since this type of product features no kind of warranty, I felt adventurous: I removed the plastic casing and drilled lots of holes to let it vent. This ramped up the working time a bit (previously it would go zombie after ~20 minutes, now it might take 1 hour or so).

Any ideas on what's going on or a workaround? Thanks in advance!

 

[WoodsieLord]

after messing around for quite a while I ended up replacing the relay and the new one, different brand, seems to work fine

 

[HRTKD]

after messing around for quite a while I ended up replacing the relay and the new one, different brand, seems to work fine

You're going to leave us in suspense and not tell us the brand/model of the new one???

 

[chrisski]

I am by no means an expert on AC, and I realize that the picture I'm posting is a 120 volt relay and 30 amps, but the relay you show is kind of tiny and I'm wondering if it is typical for such a small relay to be rated to be rated to 2kw - 2.5kw. The pic I'm showing is for a 30 amp contractor relay.

 

[Hedges]

A little DPDT relay is risky for switching between two sources. It can bridge and feed the grid from inverter or vice versa.
I tried to use a similar relay (bigger and 30A) to wire two 5kW heating elements in series/parallel. It shorted across the mains.

Big knife switch would be more positive. But not so economical/practical.

Two relays, and a control circuit waiting for one to be positively open (continuity through NC contacts) before energizing the other (and vice versa) might be safer. Some difficulties doing that with same contacts, so maybe a 3PDT relay, with 3rd contact to detect relay position. Hope it isn't welded and twisted.

 

[WoodsieLord]

Sorry! This is the one that works fine:

EDIT!: I got the pictures off internet but the one I have says 6A instead of 10A**




I'm sure my monitoring solution won't tell the real story in regards to the startup peak, but the overall consumption of the fridge is quite low.



I've installed the relay using a DIN socket.


@Hedges you have me worried now: At first (before purchasing these relays), I thought of using two simple relays (1 neutral & 1 live) but I was afraid of what would happen if one relay would fail (I would be bonding neutral from grid with live from inverter through the fridge, or vice versa )... so I started searching for relays that would use the same physical mechanism to bond two pair of connectors independently. These are the ones I could find. You think these ones also have a high risk of accidentally interconnecting the grid with the inverter output ?

Thanks !

 

[Hedges]

Well, PG&E required a visible blade disconnect between my GT PV inverters and the breaker panel back when I first put them in. (Optional now, they can yank meter to be 100% sure isolated). I think that is for when utility workers touch lines.

Circuit breaker interlocks are a UL listed way of backfeeding the breaker panel. I have a couple of those. In that case there is just the (not visible) breaker contacts isolating the disconnected one. But one is switched off and the other on after a delay (in my main panel.) Another one sold as a generator kit has its own box, and the interlock is such that turning one on snaps the other off, very fast transfer.

Some generator switches use two relays.

I don't know if my relay actually twisted to cross-connect, or just drew an arc. It didn't trip the local 50A breaker but it did the remote 70A one (which carried additional loads). One contact was fried, and after that I used the remaining contact to just enable one element. 5kW or 5kW + 5kW.

McMaster-Carr

McMaster-Carr is the complete source for your plant with over 595,000 products. 98% of products ordered ship from stock and deliver same or next day.

www.mcmaster.com

I'd just want to be pretty sure grid and inverter output never shall meet.
It may not be perfect, but I was thinking an extra contact (e.g. 3PDT relay) could have the 3rd contact carry power to activate coil of other relay only after first one opened.

Sunny Island uses a 3 pole relay rated 30A per pole to disconnect from grid. Max current is 56A. I had thought they had 3 in parallel and were derating considerably, but maybe 2 in parallel and third signals relay position (i.e. not stuck)