monitor 12 vdc automotive battery used to start backup generator

I have a Generac automatic standby generator which uses a 12 VDC automotive battery to initially start. I wish to use a Arduino board to monitor the status of this battery (monitor voltage and current draw) to determine whether or not the battery is failing and needs to be replaced.

First of all, I'm only an electronics hobbyist and would need probably need more "hand holding" than others. My ultimate goal is to successfully complete the project without damaging the Arduino.

I've done some internet research and determined that I would probably need an opto isolator such as the CNY17 as indicated in this http://electronics.stackexchange.com/questions/43498/how-can-i-use-a-12-v-input-on-a-digital-arduino-pin. Also, to measure current, I would probably need Hall effect current sensor, like the Allegro ACS712ELCTR-30A-T referred to in this link http://arduino.cc/forum/index.php/topic,12870.0.html.

Thus, can anyone provide some detailed guidance and direction so that I don't "brick" my Arduino?

Thanx

I rely on a small diesel generator all the time and reliable starting is essential. I don't bother monitoring the start battery. They don't usually fail suddenly and anyway I have another battery that could probably start the engine if essential.

I do monitor the charging of my storage batteries because repeated discharging and charging shortens battery life. Starter batteries are normally fully charged all the time.

I don't think you will get any useful information by monitoring the huge current drawn by the starter motor - probably over 60 amps. The problem is that it only takes a few seconds to start an engine so the total amp-hrs consumed is negligible. The usual test for a failing starter battery is a high-current test which probably draws 200 to 300 amps and causes the voltage to drop and weak cells to bubble excessively.

If the engine is only used very occasionally there is a risk that the starter battery will self-discharge which will be bad for the life of the battery as well as impairing starting capacity. A small solar panel to trickle charge it is usually a good solution.

If you insist on monitoring the starter battery I think it would be sufficient to monitor the voltage. The voltage is a reasonable indicator of the state of charge but it won't tell whether the battery is wearing out. With a suitable trickle charger even that is probably unnecessary - except perhaps to check that the trickle charger is working.

...R

My Generac generator "exercises" once a week for 12 minutes to recharge the battery and keep the oil circulated in the engine. I just installed a new battery and have attached a DC voltmeter to the battery. In the idle state, the voltage reads 14.0 volts. When the generator starts up, the voltage drops to around 12.6 volts and gradually works back up to 14.0 volts in the 12 minute exercise period. It's my understanding that a battery is considered "dead" when the startup voltage drops to around 11 .0 volts. So, I was thinking about logging this data in order to determine when the battery should be replaced.

I eventually plan to add text messaging to the project so that I can be notified when the battery is at a critical level and needs to be replaced.

My reason for implementing this project is that I am often out of town for about two months at a time. The last time I was out of town, my battery died and the generator did not "exercise" for almost two months. Thus, the oil wasn't circulated in the engine. Starting the generator after the battery replacement is rough on the engine since it is "dry" (the oil has not been circulated in the engine for a while).

This project is also a learning experience for me since I just retired and took up tinkering with electronics as a hobby. The software aspect should be a "no brainer" since I was a network administrator/systems programmer for 30 years.

Thanx

I guess all you'd need to do is take a voltage reading before and after you start it up... a failure would be easily detected from a low voltage drop and staying that way.

The voltage drop will be very dependent on the ambient temperature. The engine will need more power to start it in cold weather and the cold will also reduce the battery performance.

Does the engine have a pre-heater as it will also draw energy from the battery, are you satisfied that 12 minutes is long enough to replace the starting charge. I have no experience of using an engine that way.

How is the starting process managed. When a person starts an engine they know to stop cranking when the engine fires, and also if it fails to do so. Perhaps you would get useful information on battery quality by timing how long it takes to start? Again you would need to allow for temperature.

...R

Does the engine have a pre-heater as it will also draw energy from the battery, are you satisfied that 12 minutes is long enough to replace the starting charge. I have no experience of using an engine that way.

Nope. There isn't any preheater. The 12 minute "exercise" period is set by the manufacturer (Generac). I can only determine on what day and time that this exercise will begin. In other words, I cannot change the length of the exercise period. Essentially, what happens is the generator automatically starts up every Saturday at 4:30 P.M. and exercises for 12 minutes and then shuts down.

How is the starting process managed. When a person starts an engine they know to stop cranking when the engine fires, and also if it fails to do so. Perhaps you would get useful information on battery quality by timing how long it takes to start? Again you would need to allow for temperature.

The starting process is automatically handled by the generator. The generator no longer attempts to start the generator after four failed automatic start attempts. I would hear the "clicking" sound similar to what you would hear in a car when the battery is dead each time the generator attempted an automatic start.

Temperature here in Northeast Pennsylvania in the coldest months, January and February could range from approximately -15 to 40 degrees fahrenheit.

Edit: I forgot to mention that this generator is powered by natural gas. So, I don't believe a preheater such as that used in diesel powered setup would be necessary.

Perhaps your Arduino could keep track of how many start attempts are necessary. I presume normally it starts first time. So you might become concerned if it regularly takes 2 attempts and take action when it gets to 3?

I would also be inclined to listen to the start process as I would expect the length of starter-motor-run to be indicative of battery health.

Even if you have no access to the start process you can easily monitor it by measuring the voltage on the starter solenoid.

Of course another option might be to double-up the starter battery capacity. But personally I would install a solar panel trickle charger or use a mains powered trickle charger if that's an option.

After all, I think the goal is to prevent a failed start rather than to collect data :)

...R

I agree that monitoring the time from start of cranking until started is a better measure of battery and engine condition.

Most automatic generator controls monitor the alternator output to tell when the engine is started and the starter motor should be turned off. I have built controls for some big diesel generators for broadcast stations (pre-Arduino - it would be easier now!)..

Also I would suggest that you monitor the battery voltage, sampled very quickly for the time the engine takes to start. You are then really doing a high-current battery test. The battery voltage in the first 500 millliseconds of so of cranking will tell you the most about the battery condition. The length of time from start of cranking to high engine rpm as seen by alternator output tells you the overall system condition. If you were to graph the battery voltage in the first few seconds, you'd see a sudden drop to about 8 volts or so that rapidly increases

I don't know about natural gas, but I have converted a couple of gasoline fuel generators to propane with a commercial kit, and there was a solenoid valve that enriched the mixture for starting which I activated for about 5 seconds at startup.

I recently got a small 4KW gasoline generator for $250 at Harbor Freight. I'd like to figure out how to hack it to convert it to propane and build an Arduino control system for it. Biggest problem is to hack an electric starter on an engine that doesn't have one... Hmmmm.

Hang a pulley from the branch of a nearby tree (or a roof beam). Use an electric motor to haul up a heavy weight. Release the weight to pull the starter cord :) :)

...R

terryking228: Biggest problem is to hack an electric starter on an engine that doesn't have one... Hmmmm.

terryking228: I recently got a small 4KW gasoline generator for $250 at Harbor Freight. I'd like to figure out how to hack it to convert it to propane and build an Arduino control system for it. Biggest problem is to hack an electric starter on an engine that doesn't have one... Hmmmm.

Buy a starter motor for a small car and replace the pullchain with a gear?

Perhaps your Arduino could keep track of how many start attempts are necessary. I presume normally it starts first time. So you might become concerned if it regularly takes 2 attempts and take action when it gets to 3?

Well, I believe it depends on the battery. My original battery had approximately a 50/50 chance of starting the generator on the first try even when it was brand new. But my new battery starts the generator on the first try every time. This may be attributable to the Cold Cranking Amps (CCA) rating of the battery. The manufacturer recommends a battery with a minimum CCA of 525 CCA at 0 degrees fahrenheit. My original battery was at 550 CCA but my new battery is at 625 CCA's. Thus, my new battery had more "juice" to initially start the engine.

After all, I think the goal is to prevent a failed start rather than to collect data

Exactly. But I was looking at it from the perspective of collecting data in order to anticipate a failed startup due to a "dead" battery.

Also I would suggest that you monitor the battery voltage, sampled very quickly for the time the engine takes to start. You are then really doing a high-current battery test. The battery voltage in the first 500 millliseconds of so of cranking will tell you the most about the battery condition. The length of time from start of cranking to high engine rpm as seen by alternator output tells you the overall system condition. If you were to graph the battery voltage in the first few seconds, you'd see a sudden drop to about 8 volts or so that rapidly increases

Yep, I'm considering this solution since I've been monitoring my new battery with a DVM. It just hangs around 14.0 volts until the generator attempts to start. At that time, the voltage drops down to around 12.6 volts. So, I believe that as the starting voltage goes substantially lower than 12.6 volts, well, it's time to replace the battery before it fails completely. Please correct me if my logic is flawed.

Bob, Is the generator exposed or inside a building or inside an enclosure? Try using a solar-powered battery charger.

Bob, Is the generator exposed or inside a building or inside an enclosure? Try using a solar-powered battery charger.

The generator is outside the home. It is inside a steel enclosure.

I just reviewed the installation manual for my Generac model 52420 automatic standby generator and realized that the generator is equipped with a battery trickle charger. The battery receives a trickle charge from the mains utility power source when the generator is not running to prevent self discharge.

It also states that the trickle charge feature cannot be used to recharge a discharged battery. The primary purpose of the trickle charge is to help extend the life of the battery by maintaining the battery when the unit is not running.

With that said, my first (original) battery lasted 6 years before it died. I just replaced it two months ago.

What I'm trying to do is determine when the battery is coming to the end of its life and replace the battery before it fails. Generac dealers are very conservative and recommend replacing the battery every two years. IMHO, that's a waste of money. Case in point, my cheap battery lasted 6 years.

I often travel for two or three months at a time and really don't need a battery failure while I'm on the road. The backup generator supports a sump pump, furnace etc. So, I don't want to come home to a flooded basement or frozen water lines, cracked toilets etc.

To reiterate, I want to us the Arduino to track battery data over an extended period of time and evaluate that data to predict a battery failure. In other words, a just in time battery replacement.

This is the most useful piece of data. Next time just by a new battery after 5 years. :)

Seriously if you get 4 years out of a battery it is unlikely to be worth any intellectual or financial investment to extend its life. Battery costs about £80 = £20/year for 4 years = 0.2p per hour of your time writing programs.

...R

BobS0327: With that said, my first (original) battery lasted 6 years before it died. I just replaced it two months ago.

Is it possible to connect another battery to the backup generator?

Is it possible to connect another battery to the backup generator?

I guess anything's possible. But there isn't any room for a second battery in the generator enclosure. Thus, the battery would have to be installed outside the enclosure.

Also, I couldn't find any info on using a second battery on the manufacturer's website or documentation. My gut feeling is that the manufacturer may frown upon using a second battery since they're recommending that the battery be replaced every two years. This "modification" may also bring up issues if the generator is under warranty. It could possibly violate the warranty.

Caution, these sensors are for cases where you have a lot of current. And you don't want to short-circuit or drain your battery here !

So you should have short duration "tests" every x minutes/hours, and then leave that battery alone.

You don't need opto isolators necessarily, that depends on your circuit, whether you are connected to mains or not, etc.

For example if you power your arduino board from the battery, no need to isolate it from its own power supply !

And keeping the same battery for 6 years won't? :) :)

...R

BobS0327: [ My gut feeling is that the manufacturer may frown upon using a second battery since they're recommending that the battery be replaced every two years. This "modification" may also bring up issues if the generator is under warranty. It could possibly violate the warranty.

Bob, Do you have a device that keeps the battery warm in the winter? Do you have a device that keeps the engine block warm in the winter?

And keeping the same battery for 6 years won't?

I don't believe it will violate the warranty. I've reviewed the 3 year limited warranty and can't find anything that would indicate a requirement to replace the battery after two years. Also, the installation and owner's manual have no references to a requirement to replace the battery after two years. The 2 year recommendation can be found on the Generac website FAQs area and in 3rd party Generac websites. Also, dealers will make this recommendation. I believe it is recommended more to guarantee that the generator will start in an emergency situation rather than to avoid a warranty violation.

On the other hand, modifying the generator for a second battery probably would violate the warranty as it states "Products that are modified or altered in a manner not authorized by Generac in writing" will violate the warranty.

Do you have a device that keeps the battery warm in the winter? Do you have a device that keeps the engine block warm in the winter?

Yes, I have a Generac heating "blanket" that is wrapped around the battery to keep it in warm in the winter. I don't think I would need to keep the engine block warm since it is powered by Natural Gas. Also, Generac doesn't provide any products to keep the block warm in the winter.