Running small Air Conditioner on an inverter

[markopolo]

EDIT: I split the posts below from the Rear Window Air Conditioner topic: http://www.classbforum.com/phpBB2/vi...hp?f=12&t=2156
I think it will be easier to follow both topics now
__________________________________________________ __________________________________________________ ____________

I don't have a generator in my van so a smaller air conditioner appealed to me because I can run it off an inverter for a short time.
I have a 1,000 watt pure sine wave inverter wired into the vans electrical system with an automatic transfer switch. My 2 AGM house batteries total 180ah capacity.

Fedders air conditioner specs:
BTU Cooling: 5200
Voltage: 115
Amps Cooling: 4.5
Watts Cooling: 485
Plug Cap. Amps: 10
Weight: 38.6 lbs

I tested this setup today:

With the van engine running and the rear window air conditioner running off my inverter the voltage at my house batteries was 13.9 volts with all battery banks combined with a manual battery selector switch. The voltage was 13.7v using only the isolator to combine the banks. To me that means my alternator at idle supplies the current needed to run the air conditioner.

The inverter showed the output watts at 420w. That varied a bit from 370 to 410 to 390 etc.
With the motor off and running off my inverter only on my house batteries the DC amp draw got as high as 35.5ah. It varied from 31.5 to 35.5.
AC amps usually showed 3.95ah. It varied a bit from 3.68 to 3.95.

In the rear of the van (setup like a RT 190 Versatile) the starting temperature of 85F dropped to 79 in 5 minutes. At 10 minutes the temperature was 74.5F and at 15 minutes it was 71.6F. I turned of the A/C at 15 minutes. The 88F front cab temperature didn't drop much during the 15 minute test although it had continued to drop when I checked at the 20 minute mark (a/c had been off for 5 minutes by then). I was parked in direct sun.

The air conditioner compressor will cycle on and off as needed depending on the temperature. If I set the temperature on the A/C at 72F I could get up to two hours compressor cooling time which might equal 3 to 4 hours of a cool rear section of the van if I further block off that rear area. That might make the difference of a good nights sleep or not. This will really only be practical if I'll be driving for a few hours the next day or will soon have access to grid power to recharge the battery bank.

Please post any comments or advice.

2nd Edit: The van was parked in full mid-day summer sun and I have 120 watts (or maybe more) of solar panels on the roof. The input from the solar panels would have affected the test results above.

[pjc188]

Hello Marko,
I have become a member about 5 minutes ago.
I wanted to ask you about starting the window AC on batteries.
I have tried to do what you did 2 months ago, but it did not work.
As soon as i turned the AC on (5000 btu), the inverter started bipping and the red light came on and it just wouldn't start the little ac.
A friend told me that this is because the voltage from the batteries dropped down rapidly trying to provide the electrical load the ac needed, and the inverter's protection kicked in because of that.
Could you please give me an advice on that issue?
Thanks

[markopolo]

On batteries only: (van motor not running) The same thing will happen with my system if I try to start it in "Cool" mode. My inverter will beep and shut-off.
If I start the A/C unit in "Fan" mode and let it run a minute or two and then select "Cool" it works.

The thing for me to remember is to put the A/C in "Fan" mode before I turn it off. It will remember that setting on the next start-up.

That 15 minute test I ran was actually on batteries only. I know my report wasn't clear on that. My results could have been influenced a bit by some voltage input input from the solar panels on my roof because I was parked in full sun.

Today, with the van parked in my garage, and no input from the solar panels, I could not measure the DC amp draw because it exceeded my clamp on ammeter 40 amp limit. On the original 15 minute test the amp draw didn't exceed 35.5 amps DC. I only thought of the solar panel affect today.

My inverter is a bit small for the startup load. It is rated 1000 watt continuous /1100 watt five minutes /2000 watt surge. It is a Pure Sine Wave type.
I have two AGM house batteries. All test were done with fully charged batteries.

Some sort of "hard start" capacitor installed in the A/C might help. Particularly, when the batteries are not at full capacity.

Anyone know if a car amp audio 12v capacitor would help with the start-up loads when using an inverter?
Or would a 110V capacitor like the Supco SPP6E be more effective? http://www.supco.com/eclass.htm

[pjc188]

Thank you for your reply.
You mentioned your inverter is a little bit small for that startup load.
Do you think if i use 2000 running 4000 surge it will not have any problems starting the compressor?My AC is 4.5 amp running - couldn't find the start amps, even after trying ro contact manufacturer.

[markopolo]

For me, running the air conditioner on the inverter is a "no-other-choice" option. I wouldn't want to spend too much money on the project because you are always going to be limited by the size and condition of your battery bank and also recharging that battery bank. A small and quiet Honda or Yamaha generator could run the air conditioner all day and night as long as you have gas to put in the tank.

Back to your question Yes, a 2000 watt inverter would have a better chance.

I read (on a few websites) that startup loads can be 3 to 7 times the rated running watts. But, for small air conditioners like mine, my guess is that the surge can't be much more than the 15 amp AC rating of a typical household circuit. My air conditioner specs show only a 10amp ac plug rating. So the surge from my air conditioner might only be around 1200 to 1800 watts or so. I think it is the duration of the surge and not just the load that causes my inverter shutdown. Mine displays: OLP - which is AC output Over load shutdown. So, an inverter with a 2000 watt continuous rating should handle that startup much better.

What size inverter are you using?
How many batteries?
What gauge wire are you using to connect the inverter to the battery bank?
What is the length of those wires?
Does your air conditioner start-up if you turn on the fan first, then wait bit, and then start cooling?

[markopolo]

Update - my 1200 to 1800 watts surge guess looks to be way too low based on this guy's testing:

http://www.wind-sun.com/ForumVB/show...377#post122377 and
http://www.wind-sun.com/ForumVB/show...917#post122917

[youtube:1w9p0gqr]SxV7rv5wTUg[/youtube:1w9p0gqr]

22.6 amps AC inrush current on the 5000 btu Goldstar a/c he tested. Maybe 2500 to 2700 watt surge.

[markopolo]

Originally Posted by markopolo

........... I could not measure the DC amp draw because it exceeded my clamp on ammeter 40 amp limit. .........

It can measure up to 400A DC - I just needed to read the selector switch properly



43.5 amp hr draw today on battery only...........

Inverter shows 450 watts AC, voltage dips to 12 volts (from 12.7 volts)

[youtube:2kp5ov5i]HamfusD6gRo[/youtube:2kp5ov5i]

[pjc188]

Hi Marko,
You mentioned the voltage drop to 12v when running the 5k btu ac.
Isn't this drop close to the inverter's point of low voltage shut off?
Please give more details about voltage drop, amount of batteries (or amp/hour and voltage drop)
Also, what's the minimal safe voltage to operate the batteries?
Thanks again

[markopolo]

The voltage dipping to 12v during a 43 ah load is to be expected. It will drop even more as the battery capacity decreases. It was a very short test and the voltage bounced back to 12.7v as soon as the load was removed.
On my inverter there is a low voltage alarm at 11 volts and low voltage shutdown will occur at 10.5 volts. Those are not resting voltages. They are when the inverter is in use. My inverter needs to see 11.6 volts to restart after a shutdown.

The number of batteries matters. 43 ah from 1 battery is harder on it than two batteries each supplying 21.5 ah. 3 batteries each supplying 14.33 ah is even better etc.

The resting voltages in the red and yellow areas on the following chart are what you want to really avoid. Your batteries will last longer if you don't go below 50% (resting voltage) and always make sure to recharged them as soon as you can.



if you get a chance to answer these questions it might help:

What size inverter are you using?
How many batteries?
What gauge wire are you using to connect the inverter to the battery bank?
What is the length of those wires?
Does your air conditioner start-up if you turn on the fan first, then wait bit, and then start cooling?

[pjc188]

Thanks again,
So if i understood, you want to avoid less that 50% discharge of your batteries.
But, as soon as you tested the voltage of a full battery, the voltage (under load) got to 12v immediately, which according to the chart you have posted, is very close to the "yellow area" which has the low voltage you wanted to avoid - so is it means that you have a very short time of use of the battery, before getting to the yellow zone, or before the inverter will shut off?

My inverter is 2000/4000 peak
I have 2x6v in series - total 210 amp/h in 12v
6 gauge cables
2 feet long
AC will start cooling immediately - if you switch from fan to cooling, it will turn off the fan completely and then start the compressor and the fan together

[markopolo]

The chart shows resting voltages. It illustrates battery capacity. It is unrelated to voltage under a load. Take the load off and the batter voltage (and therefore capacity) will be higher.
The source is here: http://www.windsun.com/Batteries/Batter ... 20Capacity Great info.

Your setup seems ok but the 6 gauge wire is small. I went small on mine also but it is 4 gauge with a 1000 watt inverter. That would not cost too much to change.
The 250 amp initial surge on a 4' long round-trip circuit with 6 gauge wiring run causes a .49 volt drop on a starting voltage of 12.7 volts (3.86% drop). The voltage drop with 2/0 (00) gauge would only be .19v (1.5%)

I picked 250 amps DC as the surge because of the test that other guy did that showed 22.6 amps AC inrush. 22.6 x 110v AC = 2500 watts (approx). Divide that by 10 for the DC amp hr draw.
I picked 10 as the number to divide by (instead of 12v) because my video and ammeter photo pretty much demonstrates that. It showed the inverter output as 450 watt AC and a 43.5 amp hr DC draw. That's roughly a 10x multiple. The loss is caused by wiring and inverter inefficiencies. Inverters are 90% efficient or thereabouts.

Air conditioner run-time is going to be limited. 210 a/h x 50% = 105 ah. I'd guess that you could get two hours of actual cooling (compressor + fan time) then your batteries would very much need to be recharged. I'm roughly factoring in a "penalty". Peukert's law

If the battery is discharged in a shorter time, with a higher current, the delivered capacity is less.

The 105 amp hr rating on your batteries may have been calculated at the 20 hr rate meaning that after a 20 hour test the ending voltage was 10.5 volts and the battery delivered 105 amp hrs.

Lot's of inverters shut off at 10.5 volts but that is not 10.5 volts at rest. It is 10.5 volts under a load. They are two very different measurements. 10.5 volts at rest is most likely an unrecoverable battery. Good for its trade in recycling value.

[pjc188]

Hi again Marko,

Lot's of inverters shut off at 10.5 volts but that is not 10.5 volts at rest. It is 10.5 volts under a load. They are two very different measurements. 10.5 volts at rest is most likely an unrecoverable battery. Good for its trade in recycling value

Marko, there is an error in the info i gave about the batteries - they are rated at 225 a/h @ 20 hr rate, and 180 a/h @ 6 hr rate.

So if i have for example 6 batteries (6 volt, 180 a/h, 6hr rate), that will give me total of 540 a/h @ 12v (again, 6hr rate) - this is in resting mode.
So actually i do not have 540 amps at load mode, since under load there is a drop in the voltage, which means the voltage is closer to the shut off point of the inverter.

My question is: According to your video, fully charged batteries will start at 12v under load, while at resting mode they will start at 12.7v - does that mean that it will get to the shut off voltage point much faster - therefore much less time to rely on the batteries?

Is there any way you could estimate the difference in a/h i could rely on the batteries under load, if i know the resting a/h capacity?

Thanks again,
Moses

[markopolo]

You are doing that math right. Here's an explanation for anyone who doesn't understand series and parallel battery connections:

In the SERIES CONNECTION, batteries of like voltage and Amp-Hour capacity are connected to increase the Voltage of the bank. The positive terminal of the first battery is connected to the negative terminal of the second battery and so on, until the desired voltage is reached. The final Voltage is the sum of all battery voltages added together while the final Amp-Hours remains unchanged. The bank's Voltage increases while its Amp-Hours, Cranking Performance and Reserve Capacity remain unchanged.

In the PARALLEL CONNECTION, batteries of like voltages and capacities are connected to increase the capacity of the bank. The positive terminals of all batteries are connected together, or to a common conductor, and all negative terminals are connected in the same manner. The final voltage remains unchanged while the capacity of the bank is the sum of the capacities of the individual batteries of this connection. Amp-Hours Cranking Performance and Reserve Capacity increases while Voltage does not.

------------------------------------------------------------------------------------------------------------

Originally Posted by pjc188

So actually i do not have 540 amps at load mode, since under load there is a drop in the voltage, which means the voltage is closer to the shut off point of the inverter.

The voltage drop will vary depending on the load.

Originally Posted by pjc188

My question is: According to your video, fully charged batteries will start at 12v under load, while at resting mode they will start at 12.7v - does that mean that it will get to the shut off voltage point much faster - therefore much less time to rely on the batteries?

Again, The voltage drop will vary depending on the load. Mine were under a 43.5 ah load. And, my batteries are fairly new so they still get to 12.7v when fully charged. The numbers will be different for every system.

Originally Posted by pjc188

Is there any way you could estimate the difference in a/h i could rely on the batteries under load, if i know the resting a/h capacity?

Here's my guess. 6 batteries will share the load very well. I'll think of yours as 3 12v 180ah batteries. Each pair can deliver 30ah for 6 hours. So combined they're rated to be able to deliver 90ah for six hours (approximately - they might start higher and finish lower).
Lets assume you are comfortable drawing them down to 50% so roughly 270ah available at the combined 90ah rate. You'll be drawing them down at approx half that rate (my second test showed 43.5ah draw) so the batteries should easily perform to their specified rating. In theory, you'll should get 6 hours of cooling (compressor + fan) time.

The compressor will cycle on and off so you probably won't ever use 270ah overnight unless it is very hot and the air conditioner runs continuously. But, if you did use 270ah, then you have a big problem - how do you replace the 270ah you took out? You'll probably have to drive 6 or 7 hours and that might get you back to 90% charged. That might be very hard on a regular alternator.

I still see it this way:

The air conditioner compressor will cycle on and off as needed depending on the temperature. If I set the temperature on the A/C at 72F I could get up to two hours compressor cooling time which might equal 3 to 4 hours of a cool rear section of the van if I further block off that rear area. That might make the difference of a good nights sleep or not. This will really only be practical if I'll be driving for a few hours the next day or will soon have access to grid power to recharge the battery bank.

[markopolo]

Some new data

There are 4 batteries in my van; 2 house and 2 starting for the diesel engine. I have a marine style multi-battery switch that lets me choose #1. the house battery or #2. the engine battery or #3. Both banks combined or #4. All off. I can momentarily combine both banks and it lets me start the and run the air conditioner with the a/c on cool mode. It is a "make before break" switch so I can then switch back to #1. house battery only without interrupting the current flow. The current drawn from the engine batteries would be measured in amp-seconds; too small to matter. The inverter shows 12.2 volts when all 4 batteries are paralleled by the multi-battery switch and running the air conditioner on the inverter.

There's more
Started the van. House and Engine battery banks get paralleled by the isolator. Additionally, I then set the multi-battery switch to "Both".

With the van engine running at idle, my Scangauge reports .08 GPH (gallons per hour) fuel consumption and 624 RPM.

With the van engine running at idle and the rear window air conditioner running off my inverter, my Scangauge reports .11 GPH (gallons per hour) fuel consumption and 634 RPM. 13.5 volts reported by the inverter.

With the van engine running at idle and the rear window air conditioner running off my inverter and the van dash air on "max air" blowing cold, my Scangauge reports .45 GPH (gallons per hour) fuel consumption and 651 RPM. Voltage shown by the inverter is still 13.5 volts.

There was no input from the solar panels for this test.
The alternator, at idle, can carry the load of both the rear window air conditioner and the vans front dash air conditioner. Combining the two air conditioners will allow more rapid cooling. Might be useful.

Daytime running lights come on when the van motor is running. Turning them off in my van is easy, just press the emergency brake down. Even just one click will turn off the daytime running lights. Useful if I'm parked.

I don't think the Fedders 5200 btu air conditioner in my van exchanges air. I think it just recirculates the air in the van.
From the specifications:

I haven't slept with it running but maybe I should leave the Fantastic fan open just a bit if i ever do.

Was thinking about C0 (carbon monoxide) with the van motor running. I do have a new from last year C0 detector in the van.

Anyway that's the new data, I'm not suggesting it as a solution for anyone. It really needs to be well thought out by anyone trying something like this.

[pjc188]

Hi Marko,

Lets assume you are comfortable drawing them down to 50% so roughly 270ah available at the combined 90ah rate. You'll be drawing them down at approx half that rate (my second test showed 43.5ah draw) so the batteries should easily perform to their specified rating. In theory, you'll should get 6 hours of cooling (compressor + fan) time.

So If i'm going to use around 270 a/h, so batteries will be @ 50% discharge - under load, this condition will not shut off the inverter?
In other words - as far as i understood, we will not know the voltage after approx. 6 hrs of cool mode.

I there any way to be more precise on this issue?

Also

...they might start higher and finish lower

- did you mean the battery voltage?

Thanks again,
Moses

[markopolo]

I don't think the voltage will go low enough to shut off your inverter. You'll only be using 50% of the battery bank capacity. The wiring and temperature will have an affect. You won't really know until you try.
Deep cycle batteries can be discharged 80%. Limiting that to 50% is easier on the batteries.

they might start higher and finish lower

- I was referring to amps and how they calculate the 6 hr rating. I don't know how they do the calculation.

I plugged in your numbers here: - excel spreadsheet - http://www.smartgauge.co.uk/calcs/peukert.xls


The calculator above shows 14 hours to completely discharged at 45ah. So approx 7 hours to 50%.

I'm very curious to know how you plan to recharge the batteries.

[pjc188]

Hi Marko,
Thanks again for your reply - super appreciated

About ccharging my batteries: this will be done by mainly solar power, and when it's not enough, by small generator.
I am planning to make a 48v system in the van instead of standard 12v, and here is why:
I calculated that i will be able to put 5 solar panels on my van's top - they are approx. 3 feet wide each.
Each provides on a sunny day about 7-8 a/h per hour (about 5 hrs of the day the sun is highly effective).
So 5 of these panels times 7 a/h is 35 a/h per hour x 5 hrs of sun = 175 a/h.
The windoe AC we want to run takes around 43.5 to 50 a/h on 12v - so we have a little more than 3 hrs to run the AC with the a/h the solar system will provide - 175 a/h (provided by solar panels) divided by 50 a/h (consumed hourly by the AC) will give roughly 3.5 hrs of use only.
Now - let me explain why i'll use 48v system - instead of 12v.
The 48v panels are almost the same size as the 12v.
48v panel, on a sunny day, provides about 5 a/h.
5 panels will produce around 25a/h.
At 48v, the AC will use 10-12 a/h, instead of 43.5-45 a/h on 12v.
So now we have a system that provides us 25 a/h per hour, times 5 hours of a full sun (5 hours full sun on average) - 25a/h x 5 hrs = 125 a/h.
So if the AC will use even 15 a/h, the system at least, will provide enough power to run that AC for at lease 8 hrs.
Here is the calculatio: 125 a/h (provided by solar panels) divided by 15 a/h (consumed hourly by the AC) will give roughly 8.33 hrs of use.
That's, in my opinion - not bad at all.

Please verify and let me know what you think.
Thanks again,
Moses

[markopolo]

It's clever and you're thinking "outside the box" for sure. I doubt that using roughly the same amount of batteries could yield such dramatically different results in actual use. And you have the complexities of combining typical 12v RV items (pump, lights, fans, furnace, fridge controls, C0 and propane detectors, etc.) and a 12v van alternator with a 48v system with 48v inverter.

So that would be 8 x 6v batteries in series but still only 225 a/h capacity @ 20 hr rate (I don't know if the voltage affects the rating). Only 112 a/h useable if discharged to 50%. Divide by 12 hr air conditioner use = approx 9.3 hours on battery.

They might recharge then next day - I'm not sure - because the 25a/h combined from the panels would equal only 3.125 a/h per battery. Closer to a "trickle" charge than to a "bulk" charge. It can take a long time to get a battery 100% charged.

That solar array would more than power the air conditioner during a sunny day but you'd probably be outside anyway. Parking in the shade, rainy days, winter sun, performance loss from panel heat build up all need to be factored in. 8 batterys would take up a lot of space and you need to access them for maintenance. 8 AGM's would be $$$$.

The Peukert calculator shows 8 x 6v 225 a/h capacity @ 20 hr rate batteries in a series parallel configuration to give 12v would let you draw 45ah for 10 hours to get to 50% discharged.

12 volt series/parallel


I love solar and have had it on 4 RV's but it hard to beat a generator for powering an Air Conditioner.

[pjc188]

They might recharge then next day - I'm not sure - because the 25a/h combined from the panels would equal only 3.125 a/h per battery. Closer to a "trickle" charge than to a "bulk" charge. It can take a long time to get a battery 100% charged.

Hi Marko,

The 25 a/h is per hour - if we average 5 hrs full sun, we'll have 125 a/h.
8 x 6v will give us a battery bank that the system will see as one large 48v battery.
If the system will push 125 a/h to a 48v "battery", i think it will bulk charge and trickle charge - even with losses, you'll still get between 7-8 hours of comp+fan.
Assuming the AC will cycle on and off - looks like pretty good 9-10 hours of AC when reaching 50 batt discharge.

Oh, another thing i forgot to mention is that even in the hours that the sun is not at her peak (early morning, early evening) we can still get extra amps.

Anyway, the only problem is the cost...

[markopolo]

Yes, 25ah (per hour) from panels divided by 8 batteries = 3.125ah (per hour) per individual battery. It just seems low to me. I don't have any experience with 48v systems though.

Yes, the cost................