|
| Author |
Message |
spencerin
Member
|
# Posted: 19 Apr 2014 04:08pm
Reply
I have this problem only when a cooling appliance's condenser (i.e. 5,000 BTU A/C unit, small refrigerator) kicks in - the inverter chirps on and off and makes it hard to sleep.
I have a 1,500W Cobra inverter connected to a ~300Ah 12V battery bank. When I come to my cabin, the battery bank is fully charged at the low 13Vs. Things run well until the condenser kicks in and then the voltage drops immediately to 11.6V and the inverter starts to chirp (as designed). Once the condenser stops, all's normal again.
This seems to me to be a capacity / battery bank issue and not an inverter issue, but I'm not certain about that, either. The system recharges relatively quickly - like within 10-15 minutes the system will recover back to upper 12 / lower 13Vs.
So, how do I overcome the chirps? New inverter, add batteries, find the speaker for the chirp and cut the wire?
|
|
MtnDon
Member
|
# Posted: 19 Apr 2014 06:33pm - Edited by: MtnDon
Reply
Battery cables.
If too small they can cause a voltage drop under the heavy load. I ran into that with a Xantrex inverter a short time ago. A 1000 watt load caused the warning buzzer to trigger. changing from #2 AWG to 2/0 AWG solved the problem. That was an extreme jump in cable size but I had the 2/0 on hand.
Do the cables get warm under load? Also check each connection point to see if any of them have high resistance and get hot.
If you want, list the cable lengths, the watts under load and I'll calculate the wire gauge needed for minimal voltage drop.
|
|
creeky
Member
|
# Posted: 19 Apr 2014 06:50pm
Reply
it's a load issue.
i thought I stretched the capacity of an inverter/battery bank. but a 5k a/c and a fridge. nice. on 300 amp hrs with a modified sine wave inverter.
a pure sine inverter would handle the surge of the condenser better. so maybe it's time for a better inverter. bigger battery bank. and more solar/genny-charger input.
|
|
razmichael
Member
|
# Posted: 19 Apr 2014 06:51pm
Reply
Not sure what size you are using but I fully agree with MtnDon that size does matter! I think this issue came up a couple of years ago on the forum. I suspect some of the manufacturers really understate the cable size to not scare potential customers - assuming they will never likely use the inverter to it's max. The Cobra 1500 suggest #4 for less than 4 feet which is really not going to work for heavy loads. It does say "as short and thick as possible"! I compare this to my Xantrex HF-1800 that states 2/0 for max of 5 feet (the HF-1000 recommends #2).
|
|
MtnDon
Member
|
# Posted: 19 Apr 2014 08:36pm - Edited by: MtnDon
Reply
Quoting: spencerin The system recharges relatively quickly - like within 10-15 minutes the system will recover back to upper 12 / lower 13Vs
Hmmm. How long does the system run in alarm mode? If only a few minutes at most I would expect a quicker recovery. But if the run time is longer that could still be okay.
How old and what type are the batteries? Is this alerting under load new because the load was never applied before or was it working okay and then the alarm began out of the blue?
You could try this as a battery check... assuming you have more than one battery. A load of about 1000 watts is applied to each battery separately for 15 seconds. Record the battery voltage at the 14 second mark. Disconnect the load at 15 seconds. Wait 15 seconds with no load. Check the battery voltage. It should have recovered to within a couple tenths of a volt compared to the first, pre load, reading. Check each battery. If there is a bad one it will stand out.
|
|
ILFE
Member
|
# Posted: 22 Apr 2014 12:04am - Edited by: ILFE
Reply
Here is a chart that I use for my cable length and voltage drop calculations:
Image attached, but here is a direct link to the file as well.
|
|
spencerin
Member
|
# Posted: 22 Apr 2014 06:21pm - Edited by: spencerin
Reply
MtnDon (and others),
The length of the cable from the battery bank to the inverter is 3', and I can't shorten the cables any more. Now, they are the #4 AWG Cobra cables.
I haven't touched the cables to see how hot they may get, but I do have the circuit fused, and the fuse hasn't blown yet. The watts under load is usually pretty low, like <600, most of the time, and not continuous. As far as the length of time running in alarm mode, not very long - once the condenser stops running, the chirping stops. It is just strange that the voltage drops quickly and then bounces back fairly quickly.
This has been going on from day one for about 2 years now, and I haven't noticed any drop in battery performance over that time. So, perhaps it is the cables after all. I just bought another ~100 Ah deep cycle battery to add to the bank, so that will help.
What gauge of cable to do you recommend?
|
|
MtnDon
Member
|
# Posted: 22 Apr 2014 08:26pm - Edited by: MtnDon
Reply
3 feet of cable is good. It would be difficult to get shorter.
600 watts? That's the 120 VAC draw I assume. Add 15 to 20% for inverting loss, so maybe 700 watts of DC is being used..
If the voltage is 12 VDC that would be about, 700 / 12 = 58 amps. If the voltage drops to 11.5 that 700 watts = 61 amps......
My calculator shows for, 12 VDC, 60 amps, 4 AWG copper wire, keeping to a maximum of 3% voltage drop, a 60 amp load should work okay out to 9 feet from battery to inverter.
Note that the chart posted above calls for 6 AWG wire. I have thought for some time that the charts and tables we are given are overly optimistic. Maybe they don't want to scare us with huge cable sizes and prices. I dunno.
From time to time I even think my calculator may be optimistic as the #2 cable I used was theoretically sufficient for the Xantrex inverter I mentioned above.
Theoretically #2 AWG wire would have 0.60% drop at 60 amps with a 3 foot distance. IF you occasionally ran higher loads I'd go with an even larger size; 1/0 should let you run up to the maximum rating of the inverter. If you don't run heavier loads I'd probably use the #2 AWG size.
Welding supply shops often sell welding cable by the foot. Wireandcabletogo.com has decent online pricing. solarseller.com has good prices on cable lug ends.
Are your cables factory made? Or home made? Home made soldered lugs can introduce extra resistance. Solarseller has a pretty good hammer crimper.
It really does sound like the cables have too much resistance for the load. That makes the voltage drop over a few seconds to where the alarm kicks in. If you were close by I'd loan you some #2 cables as a trial.
If the batteries are any good at all they would be capable of delivering the amperage being used. I would check the cables for warmth the next time you are there. You should not be able to feel any temperature rise; if you can that is electricity being wasted. Also look for any hot connections. An IR thermometer is handy for that.
Any chance of borrowing an additional set of cables like you have? Connect them all up in parallel and the alarm should cease. Maybe using a pair of auto jumper cables... carefully to not short anything.
This problem is yet another reason I prefer a 24 VDC system over a 12 VDC system. You can use smaller and cheaper gauge wires as the current draw is half.
|
|
|
spencerin
Member
|
# Posted: 22 Apr 2014 09:29pm - Edited by: spencerin
Reply
MtnDon,
Thanks for the recommendations. I think the first thing I will do is run a heavy load for a minute and then touch the cable (after shutting the load off of course) for warmth. And then I'll go from there.
I do have another question, about the configuration of the wiring of the battery bank. I have 3 batteries wired in parallel. If you number them #1, #2, and #3 from right to left, I have the charge controller wired to both terminals of #1 and the inverter wired to both terminals of #3 (with #2 joining #1 and #3). Does this matter at all for the equalization of the charging or drawing from the batteries?
I've seen some diagrams recommending that the positive wire of the charge controller be connected to the positive terminal of #1 and the negative wire of the charge controller be connected to the negative terminal of #3 (instead of #1). And vice versa for the inverter. What difference does it make, if any? (I didn't think it did.....)
|
|
MtnDon
Member
|
# Posted: 22 Apr 2014 10:23pm
Reply
Shutting the power off and them temperature checking by hand is fine, but with a 12 VDC system you have nothing to worry about touching one of the insulated cables at a time. I don't mind holding and then turning on the load so I can feel the actual rise, if present. But either way is fine. Respect DC but don't be overly frightened of it. What is to be avoided and very careful about, is a short circuit, pos to neg as that can weld metal.
~~~~~
Three parallel connected batteries. As you did, let's number 1, 2, 3. The most common Good method is to run battery interconnects from the + on #1 to #2 and then another to #3. Repeat for the negative terminals. Then connect the charger(s) and the load(s) using the Positive on battery #1 and the Negative on battery #3.
The reason being that will help to prevent discharging or charging one battery more than the others.
The reason that can happen in parallel connections is that even batteries from the same manufacturing run can and do have different internal resistances. A battery with a higher internal resistance will be slower to recharge, meaning more current will go to the battery with the least resistance. Similarly with discharging, the battery with less resistance will discharge more than the others. Eventually after many use cycles you can end up with one battery that is growing "old" faster than the others. Equalizing can help some.
There is no way to predict the outcome between different sets and different users. I have some parallel strings and after 5 years can npot really tell any deterioration. Others seem to have problems the first year.
~~~~
Using heavy copper buss bars is another connection method and one that I particularly like. The heavier the bars are the less their resistance. There is a Positive and a Negative buss bar. The batteries are connected to the buss bars; all the + to the Positive bar, all the negatives to the Negative buss bar. The chargers and loads are then connected to the buss bars. Use or oversize cables for the battery to buss connections is recommended; 2/0 or larger. Making each cable the same length is often recommended. The idea behind all that is to make the least possible resistance in the cables, the connections and the buss. The batteries are the variable that the owner can not control so we try to make the best out of everything else.
A side benefit of buss bars is that it is simple to remove one battery from the system if necessary, without disturbing the other battery connections.
|
|
spencerin
Member
|
# Posted: 23 Apr 2014 12:14am
Reply
MtnDon,
To clarify this statement ("Then connect the charger[s] and the load[s] using the Positive on battery #1 and the Negative on battery #3."), you mean that the + from both the charge controller and inverter share the same + #1 terminal, and the - from both share the same - #3 terminal? Am I reading that right?
|
|
MtnDon
Member
|
# Posted: 23 Apr 2014 09:50am
Reply
That is correct. Electrically, the power "in" and the "out" paths of the batteries can share the same terminals.
|
|
spencerin
Member
|
# Posted: 30 Apr 2014 09:23pm - Edited by: spencerin
Reply
An update for the sake of updating -
Ran a ~750W appliance for a minute straight and held the wire while the appliance ran. Maybe the insulation is really good, maybe I didn't run the appliance long enough, or maybe I need to run a higher wattage appliance, but there was no increase in temp in the wire.
So, what did I do? I took apart the inverter, took out the speaker, and said, "time to move on." Everything else about the inverter works as designed, and the silence is priceless.
|
|
ILFE
Member
|
# Posted: 9 May 2014 02:27pm
Reply
Personally, if there is any question, I would:
1. Make sure appropriate fuse / breaker protection is in place.
2. Run the appliance for several minutes.
If no temperature rise is detected, as you stated, keep the speaker disconnected and carry on.
|
|
|
