diy solar

diy solar

ATS causing power surge inverter damage.

TorC said:
Reading this, I'm wondering if Vgrid->batt and Vbatt->grid are set too close for the battery bank size. Solution is either a bigger battery bank or to set those points further apart.
Click to expand...
This question has been brought up by me and others in this thread. Note that the MOES device senses the battery voltage and will swap as a function of the two set points. It is not a failover to grid or vice-versa. I have my low voltage disconnect set at 11.4 V. I do have ~0.15 V lower reading on the MOES than at the battery...and not because of wire size...just an error of the MOES but ok as it is just a relative error. This LV cutoff translates (after rest recovery) to only about a 30% DOD. 12.3 V as measured at my (isolated by my marine switch) flooded lead acid battery. My reenable set to 13.5 V.

That said, your question (TorC) got me thinking more about these inverter failures. My system is basically set up to switch off at some point after the panels stop receiving sunshine and the battery goes below 11.4 V. This time of year I lose sun on the panels about 4 pm and the battery (if fully charged by then) will ride out until about 9 pm before hitting the 11.4 V. Because of how my system is set up there's no chance that I will have a rapid reenabling of the battery/inverter. I also look at the weather every day and if it's not likely to get a good charge on the battery I will keep my loads down to ~50 Watts.

So I'm curious if anyone with a failed inverter has had the device switch to grid during the day under a heavy load. In particular, if the load is excessive the battery can drain rapidly especially if the SOC is low to begin with. And couple that with the possibility of clouds coming and going there is a possibility of the sun popping out of a cloud simultaneously with the MOES switching to grid (due to heavy load and cloud cover) and back (strong sun will raise panel and thus battery voltage). Also, curious as to voltage set points as indicated by TorC. Please post your settings at time of failure. Note that I have almost 2 volts between the two set points. That may not be enough if MOES is switching during the day.

But as likely as not the continuous pounding of high inductive loads can be the source of your inverter failures. Break-delay-make if your system runs this way.
 
After rereading the thread while a bit less tired, I agree the relay issue sounds more likely the root cause. Setpoints too close or a load higher that solar output on a battery bank designed only for brief UPS load time rather than full grid time may be contributing, possibly only by more cycling that means more chances for something to go wrong.

If it's the relay issue it probably is much less likely, or even not an issue at all if it's paired with a grid-tie capable inverter, as I'd expect the inverter to be designed to handle an out of phase connection. An off-grid only inverter may just use dumb start the sine wave when it's turned on and force that without paying any attention to what else is happening. Unfortunately, with it being cheap equipment, it's much more likely the latter inverter will get paired with it.

If they really used two relays separately controlled that seems very suspect. A DPDT relay would be far safer, IMHO. Otherwise it should be set up so the output of one triggers the other so it takes the wrong welded contact to backfeed the inverter or generator.
 
experimenting,i know for a fact the moe ats will not trip at 200w and i know it will at 460w,been running for two days with my fingers crossed at 290w,going to keep creeping up the watts until it trips,i will keep you posted
 
I'm not sure what fire hazard means by "know for a fact the moe(s) ats will not trip at 200w". By that, do you mean that it won't destroy your inverter? Or is it tripping your inverter but not killing it? I would hold off on this experiment as I think I might have a solution. Until then I'd try to lower your loads anytime the Moes is about to switch back to the inverter. If I remember (most nights, and only temporarily) I've been turning off the Moes and my inverter when the Moes switches to grid. If nothing else it limits further battery DOD from standby inverter power draw. I don't think the Moes draws any power when off but does draw about 5 W from the battery when inverter selected. I measured with my new (and favorite new toy) clamp current meter 0.4 A @ 12V. This is most likely primarily the energized coils on the relays.

I gave this some more thought and research over the w/end. I stand by the likely scenario that only switching from grid to the inverter is causing the problem. The relays "throwing" from inverter to grid essentially is a just like an off switch for the AC side of inverter and thus no transients should be going into the AC output of the inverter when switching to the grid. If anyone more knowledgeable than me is following this thread, please chime in!

As for a remedy, I think there may be a device out there (varistor...https://en.wikipedia.org/wiki/Varistor...please read). Note the "limitations" and "hazards". It may be as simple as selecting the correct varistor and placing in parallel with the two inverter inputs at the MOES (L and N). Due to the potential failure modes, a short (with fire!) and thus should be in its own enclosed metal box with an inline fuse and open (benign) so somehow you'd have to know if that happened so you could replace to continue protecting the inverter.

Again, looking for someone with more knowledge than me on this potential solution. It seems that the selection of the varistor needs to be ~20% above operating voltage and sized to absorb (dissipate as heat) the transient. Note that this is essentially a very short duration event (order of milliseconds) and thus no heat sink required. The device itself (metal-oxide) is made with ceramic balls (microscopic) coated with the metal-oxide and thus distributes the heat readily/uniformly within the device.

So, I'd wait to see if a) someone chimes in or b) I have more time to look into this myself. Pressed for time on other projects right now.
 
SunnySoCal said:
I'm not sure what fire hazard means by "know for a fact the moe(s) ats will not trip at 200w". By that, do you mean that it won't destroy your inverter? Or is it tripping your inverter but not killing it? I would hold off on this experiment as I think I might have a solution. Until then I'd try to lower your loads anytime the Moes is about to switch back to the inverter. If I remember (most nights, and only temporarily) I've been turning off the Moes and my inverter when the Moes switches to grid. If nothing else it limits further battery DOD from standby inverter power draw. I don't think the Moes draws any power when off but does draw about 5 W from the battery when inverter selected. I measured with my new (and favorite new toy) clamp current meter 0.4 A @ 12V. This is most likely primarily the energized coils on the relays.

I gave this some more thought and research over the w/end. I stand by the likely scenario that only switching from grid to the inverter is causing the problem. The relays "throwing" from inverter to grid essentially is a just like an off switch for the AC side of inverter and thus no transients should be going into the AC output of the inverter when switching to the grid. If anyone more knowledgeable than me is following this thread, please chime in!

As for a remedy, I think there may be a device out there (varistor...https://en.wikipedia.org/wiki/Varistor...please read). Note the "limitations" and "hazards". It may be as simple as selecting the correct varistor and placing in parallel with the two inverter inputs at the MOES (L and N). Due to the potential failure modes, a short (with fire!) and thus should be in its own enclosed metal box with an inline fuse and open (benign) so somehow you'd have to know if that happened so you could replace to continue protecting the inverter.

Again, looking for someone with more knowledge than me on this potential solution. It seems that the selection of the varistor needs to be ~20% above operating voltage and sized to absorb (dissipate as heat) the transient. Note that this is essentially a very short duration event (order of milliseconds) and thus no heat sink required. The device itself (metal-oxide) is made with ceramic balls (microscopic) coated with the metal-oxide and thus distributes the heat readily/uniformly within the device.

So, I'd wait to see if a) someone chimes in or b) I have more time to look into this myself. Pressed for time on other projects right now.
Click to expand...
still working,was 200w no problems,been at 290w for 2 days no problems works great,i think,i know the bigger the load the more of a chance the inverter will trip,just trying to find that sweat spot
 
FH...please clarify. Your inverter has been going into an overload trip? I think in an earlier post you indicated you'd fried 3 inverters. As I've said, if at all possible, turn off loads (especially inductive as in motors) prior to switching back to inverter. For now, at least until perhaps we can implement a surge suppression. I'd hate to see you smoke another inverter!


Are you in a similar situation as me with the transition to grid occurring in the evening when your battery is a low voltage set point on the MOES?
 
yep over load trip,error opp,but it only happened at 460w and it switches back and forth all day especially on cloudy days,again no problems at 200w,waiting a couple more days at 290w to see if it trips,this 3000w inverter just trips so it dosent blow up so far,if we could stop the surge or slow down the relay switching,i think it would work.im running led lights,oh when it did trip,it was in the morning when i got up to check it
 
That's most likely too high for your low voltage disconnect. On my system I was disconnecting at 12.0 volts (my old home brewed device) but that was obviously under load. Within 30 minutes or so (the "recovery" time) the battery would read 12.7 volts...which is fully charged. That was by design as I wanted to keep my (expensive, but a gift!) Trojan battery running forever. If you look at the curves for cycle life vs DOD they don't even go to 0% DOD!

1643046827841.png

Note that that relationship between disconnect voltage and DOD is going to be a function of the load at disconnect. I'd recommend determining what your "rest state" voltage is at 12.2 low voltage setting at your 290 W load. Do this, but don't allow your MOES to trip as it will start charging immediately if in sunlight. You should let the Moes read 12.2 (I think it disconnects at 12.1 with a 12.2 setting), turn off the Moes (ie goes to grid as source), turn off your inverter, turn off your PV panels (hopefully you have a disconnect), and if you have a battery disconnect switch (you should!) turn that to off as well. Read the battery voltage in ten minute increments until it stabilizes. I hope you have a VOM...if not you can get a decent one for your needs under $30.

Point is, I think that your LV is set too low and your inverter may be switching in and out way too frequently. Chart below is for AGM or FLA though I usually use 12.7 for full charge for FLA but this one had two types of batteries so posted it instead not knowing what type you are running...


1643047255580.png
 
SunnySoCal said:
That's most likely too high for your low voltage disconnect. On my system I was disconnecting at 12.0 volts (my old home brewed device) but that was obviously under load. Within 30 minutes or so (the "recovery" time) the battery would read 12.7 volts...which is fully charged. That was by design as I wanted to keep my (expensive, but a gift!) Trojan battery running forever. If you look at the curves for cycle life vs DOD they don't even go to 0% DOD!

View attachment 81077

Note that that relationship between disconnect voltage and DOD is going to be a function of the load at disconnect. I'd recommend determining what your "rest state" voltage is at 12.2 low voltage setting at your 290 W load. Do this, but don't allow your MOES to trip as it will start charging immediately if in sunlight. You should let the Moes read 12.2 (I think it disconnects at 12.1 with a 12.2 setting), turn off the Moes (ie goes to grid as source), turn off your inverter, turn off your PV panels (hopefully you have a disconnect), and if you have a battery disconnect switch (you should!) turn that to off as well. Read the battery voltage in ten minute increments until it stabilizes. I hope you have a VOM...if not you can get a decent one for your needs under $30.

Point is, I think that your LV is set too low and your inverter may be switching in and out way too frequently. Chart below is for AGM or FLA though I usually use 12.7 for full charge for FLA but this one had two types of batteries so posted it instead not knowing what type you are running...


View attachment 81079
Click to expand...
i agree but its not tripping,just experimenting,leaving it their for now,switching back and forth all day on a cloudy day is proving my point on the load 200w
 
If you are there when it's going back and forth then fine assuming it's at least a few seconds between switching...but if not it could literally be switching back and forth quite rapidly...generating a big beating on your inverter(s). Were these the settings you used when your other inverters fried? When I had my home-brew every once in a while the relay would "buzz" as in going back and forth rapidly (several to many times per second kind of thing). Why my circuit was probably not the best!
 
no if i remember it was 12v cut off and 14v reconnect and differant batteries 4 50ah 12v gel,now 2 6v 230ah in series
 
SunnySoCal said:
Ok...but I suggest you still do the test I suggested unless your goal is 0% DOD.
Click to expand...
silly me did your test in reverse,13.7v witch is my float,disconnected everything,10mins,13.33,20mins,13.27,30mins 13.15 reread your post started over but got side tracked,12.1 disconnected everthing voltage goes to 12.7 instantly 45mins later its at 13.2 not sure if thats what you want or if its a good thing
 
Close...but "recovery" voltage (as in rest voltage with no load connected to battery) is going to be both a function of the load power and where you are on the discharge curve (ie DOD). I'm surprised that it jumps "instantly" unless you still have your panels engaged? But I haven't performed this sequence so will do so when I get a chance.

So this test is two-fold (as in two "data points")...and you can iterate as one of the outcomes will give you your DOD at any given MOES shut off voltage, the other insuring you aren't causing the relay to chatter. So I would repeat the test once you have a fully charged battery:

With a fully charged battery
Apply your loads (300 W ok)
Monitor your MOES until it closes in on your shut off voltage (12.2 Volts as indicated on your MOES).
Once it gets to 12.2 V as quickly as possible turn the MOES unit to Off, disconnect your PV panels, and turn off your inverter. If you have a battery switch throw that for good measure to get the most accurate reading.

Determine what your "rest" battery voltage is. Compare it to the chart in my previous post to determine DOD.

Most likely based on your reported results the MOES relay(s) aren't "chattering" (rapid switching) but reco just the same to be on safe side. You might consider doing the experiment twice...once after twilight (no chance of charging) and once in full sun (maybe the "instantly" result).

You might find that you can lower your LV cutoff on the MOES if prior thought was that you were at 60% DOD. I'm performing similar experiment and have found (with ~80 W load) that 60% DOD is happening with about a 11.5 V LV cutoff (as in the battery recovers to ~12.2 V).

Note that it's been reported that the MOES voltage is off a bit but that's ok as it's a relative. Sometimes mine reads 0.1 V lower and sometimes 0.2 V lower...so I figure somewhere around 0.15 V lower reading.
 
ok i went out to pick the kids up from school came back and the voltage is at 13v been at least a hour and a half,just hooked it all up again, as soon as it disconnects it always goes to 12.7,everthing connected or disconnected.sits at 12.6 when i check it in the morning with the draw of the inverter and controller,moes ,before it starts charging still dark,getting off toppic,havent had a problem with tripping the inverter for awhile so happy for now,cheers
 
SunnySoCal said:
Close...but "recovery" voltage (as in rest voltage with no load connected to battery) is going to be both a function of the load power and where you are on the discharge curve (ie DOD). I'm surprised that it jumps "instantly" unless you still have your panels engaged? But I haven't performed this sequence so will do so when I get a chance.

So this test is two-fold (as in two "data points")...and you can iterate as one of the outcomes will give you your DOD at any given MOES shut off voltage, the other insuring you aren't causing the relay to chatter. So I would repeat the test once you have a fully charged battery:

With a fully charged battery
Apply your loads (300 W ok)
Monitor your MOES until it closes in on your shut off voltage (12.2 Volts as indicated on your MOES).
Once it gets to 12.2 V as quickly as possible turn the MOES unit to Off, disconnect your PV panels, and turn off your inverter. If you have a battery switch throw that for good measure to get the most accurate reading.

Determine what your "rest" battery voltage is. Compare it to the chart in my previous post to determine DOD.

Most likely based on your reported results the MOES relay(s) aren't "chattering" (rapid switching) but reco just the same to be on safe side. You might consider doing the experiment twice...once after twilight (no chance of charging) and once in full sun (maybe the "instantly" result).

You might find that you can lower your LV cutoff on the MOES if prior thought was that you were at 60% DOD. I'm performing similar experiment and have found (with ~80 W load) that 60% DOD is happening with about a 11.5 V LV cutoff (as in the battery recovers to ~12.2 V).

Note that it's been reported that the MOES voltage is off a bit but that's ok as it's a relative. Sometimes mine reads 0.1 V lower and sometimes 0.2 V lower...so I figure somewhere around 0.15 V lower reading.
Click to expand...
sooooooooo i was reading a very simalar thread by you and another member.i found it interesting that he was running led lights with a draw of 200w and his inverter is tripping,mine doesnt and im still at 290w with no problems,how do you explain that, https://diysolarforum.com/threads/ats-switching-from-grid-to-battery-trips-inverter.33113/
 
You must log in or register to reply here.

Similar threads

U
ATS ignorant
Replies
4
Views
343
UGT
U
J
Midnight solar classic 200/ATS/GP-SW3000-24 3000-Watt Inverter
Replies
2
Views
362
JamesM
J
S
Safe cheap transfer switching with LVD in small system
Replies
15
Views
307
SWSL
S
H
Is Inverter Power breaking Modems, Electronics?
2
Replies
28
Views
1K
Zwy
Zwy
J
Pure Sine Wave and Surge Protector Power Strips
Replies
3
Views
1K
GridWorks Green Solar
G
Back
Top