ACRs, reserve batteries and tales from the land of Duh

[tgsail1]

Found an interesting effect a last week that others might have experienced. We have a 400 AH GC main bank, 105 AH 12V reserve battery, 105 A L-N alternator (internally regulated, connected to the 1-2-B switch), a Yandina ACR (100 A)and BMV-700 monitor. Went for a weekend cruise and decided it was a good opportunity early in the season to start on the reserve bank. Started great, had a great weekend monitored our usage over the course of a day and since I wasn't using solar, I could get good readings. Next day, -60AH later, noticed I never switched the battery over to house- oops. Decided to run the engine to recharge the battery. Thing 1: Engine starts (still on reserve) and the alternator is putting out a whopping 70A to the poor little 105 AH battery. Thing 2:ACR switches on, the alternator output dropped to 60A (higher voltage at the terminal), but near as I can tell, the house batteries began charging the reserve battery. Thing 3:After putting 40AH net back in to the system, turned off the engine, the ACR continued to connect the batteries together for about 5 min during which time the main bank continued to charge the reserve bank limited only by battery acceptance and the 100A limit of the ACR.  At the end of it all, the resting voltage on the house bank had clearly dropped.

This is an odd situation- messing up one bank shouldn't make you mess up the second one. If the situation was reversed and the house bank was low, the reserve bank would have been dragged down- not a good idea.  A few things would improve the situation: 1) direct connection from the AO to the house bank would have limited the alternator voltage a bit and the massive current to the reserve bank, but not the leak from the house bank to reserve.  2) I should have started the engine for charging on the house bank, 3) Given 1) and 2) I think an Echo Charge would not have allowed this to happen at all. 

I guess in an ideal world, someone would make a global regulator- controlling the alternator, providing two isolated outputs to separate banks (the smaller bank doesn't need to be efficient), but recognizing when another source of charge is present (on a separate input terminal) and charging the reserve bank from that source. I'm not aware of such a regulator.

[mainesail]

Went for a weekend cruise and decided it was a good opportunity early in the season to start on the reserve bank. Started great, had a great weekend monitored our usage over the course of a day and since I wasn't using solar, I could get good readings. Next day, -60AH later, noticed I never switched the battery over to house- oops.

How was the -60Ah determined? If running off the reserve bank the Ah counter, ideally connected to the house bank, should not have been recording an Ah deficit? Was this a guesstimated -Ah consumption?

Decided to run the engine to recharge the battery. Thing 1: Engine starts (still on reserve) and the alternator is putting out a whopping 70A to the poor little 105 AH battery.

Similar question, how was the current output of the alternator determined? If properly wired, the Ah counter reads the current flowing into or out of the house bank, not the start bank? Did you DC clamp-meter the alt output?

Thing 2:ACR switches on, the alternator output dropped to 60A (higher voltage at the terminal), but near as I can tell, the house batteries began charging the reserve battery.

If the house bank resting voltage was higher than the charging voltage of the start battery, then that could be possible, for house bank help raise the overall system voltage to a higher overall parity. However, the combiner does not combine with the house bank until 13V and a 30 second delay so this is already above the house banks natural resting voltage and at +0.3V above resting it really becomes a charging voltage for the house bank as well as start..

Even at a 13.0V + 30 seconds to combine you still should have been in bulk. Once the 13.0V threshold has been passed a 30 second delay starts before combining occurs. By the time you combined you may have been at 13.1V +/-. I am guessing either there is a bad regulator voltage sense issue, causing the current limiting you saw, or the alt was just heating up and putting out less current. With a terminal voltage of 13V for combine this is well above the 100% SOC voltage point for the house bank meaning no current should flow out of it and into the other bank and most likely a few mA were flowing into the house bank. How and where you measured all this is important to determine what you saw....

Thing 3:After putting 40AH net back in to the system,

How was this +40Ah determined? In the description you have described discharging and recharging a bank that you should not have been able to see any + or - current or + or - Ah data for? Perhaps it's possible your bank 1 & 2 are backwards..?

turned off the engine, the ACR continued to connect the batteries together for about 5 min

The exact same way the combiner combines, it opens or un-combines. Once the 13V threshold has been met a 1 minute timer starts before the relay opens. With lightly loaded healthy batteries it would not be uncommon to hold a surface charge on the plates above 13.0V for 4 minutes or more. AGM's can do this for 30+ minutes and in cold weather disconnected can hold 13+ for weeks. Again this is nothing more than a surface charge and is not quantifiable usable capacity. So essentially 13.0V + is a temporary surface charge and above the resting 100% SOC voltage of the battery. 

during which time the main bank continued to charge the reserve bank limited only by battery acceptance and the 100A limit of the ACR.

Actually you were limited by Ohm's law. When batteries are at voltage parity very little current can move between the banks. The combiner creates what amounts to voltage parity because it physically parallels the batteries..
 
While some very small currents were flowing this "charge" or discharge of the house bank is very minimal because the batteries are in parallel and at virtual voltage parity. In order for current to flow into a battery there needs to be a voltage differential between the SOC voltage and the charging voltage.. The smaller this voltage differential the less current can flow and this is all governed by Ohm's law.... At parallel voltages very little current moves between the banks during the time between discontinuation of charging and the 13.0V threshold and 1 minute delay.

At the end of it all, the resting voltage on the house bank had clearly dropped.

If the Ah counter was wired correctly and that bank had only been used via the combiner then the net -Ah should have been displayed. If the batteries had a surface charge from the shore charger then wiping this out via combining can happen but the net decrease in quantifiable Ah capacity will be very small.

This is an odd situation- messing up one bank shouldn't make you mess up the second one. If the situation was reversed and the house bank was low, the reserve bank would have been dragged down- not a good idea.

Again the cut-in & cut-out voltages are well within a voltage range, above resting full voltage,  that it makes very little net difference to either bank. For example if the charge source were very small, current wise, and wired to the smaller, usually full, starting battery the relay would simply fire on at 13V then drop out again once house voltage dragged the new parity voltage down below 13V.  The solar array would then need to charge the start battery back to 13.0V + 30 seconds before it would try charging the house again. This is called "relay cycling" and can happen with small solar arrays fed to a start battery first, and with large differences in bank size, and why all charge sources should really feed house first.

A few things would improve the situation: 1) direct connection from the AO to the house bank would have limited the alternator voltage a bit

If the house bank was already full it would have just brought it to 13V+ very quickly then combined with the depleted start battery. The start battery still would have taken the same current if the terminal voltage was the same, which in parallel it would have been except for the minor voltage drop across the relay. The battery will only take high current for a short duration, even AGM batteries. At 70A you were essentially charging at .66C and target voltage would have been attained very, very rapidly. Heck at .4C even an AGM bank from 50% SOC is only in bulk for 19 minutes or so. A flooded battery with much higher internal resistance, charged at .66C, and you're talking somewhere under 2 minutes in bulk from 50% SOC....

and the massive current to the reserve bank,

Your battery was simply taking bulk current on its way up to the regulators target voltage.  At 70A it would have quickly risen to the regulation voltage limit and then the battery would have dictated current acceptance at that voltage.

but not the leak from the house bank to reserve.

Again Ohm's law dictates the transfer of current at parity voltage. With no charging source present and voltage parity mA of current move between the banks at best. You essentially only have 1 minute of paralleling beyond the battery dropping to 13V which is above the 12.72V +/- resting voltage of a 100% SOC flooded battery.

If we assume your house bank was at 100% SOC, at 12.7V, and you discharged the reserve battery to 50% SOC or about 12.2V and figure that in order to get any sort of reasonable current to flow into the 12.2V battery we need at least +.25V on top of that, then the voltage differences at the point we parallel the batteries (without a charge source present) = 0.25V differential. If we then assume a .007 Ohm internal resistance of these batteries, pretty typical, means an initial in-rush or current surge from one battery to the other of somewhere around a very, very short peak of 36A of current. As the voltages equalize or attain parity, this happens very quickly, lets assume 12.600V and 12.596V at .007 Ohm we can see that only 0.57A can move between the batteries with no charge source present.

The longer the batteries remain in parallel the more voltage parity is attained until we are looking at voltage differences out to the thousandths or 12.6000V vs. 12.5998V....

With this it is easier to see that 1 minute of delay at 13.0V, after charging has been removed, really takes almost nothing from either bank because by design this is all occurring above the 100% resting SOC voltage point..

2) I should have started the engine for charging on the house bank,

Really not a big difference but you don't need to start on your reserve battery, unless you simply want to test it. Course doing so created your issue.

3) Given 1) and 2) I think an Echo Charge would not have allowed this to happen at all.

Echo charger turns on at 13.0V too but is limited to 15A (though you'll rarely see 15A) and it would have taken significantly longer to recharge your reserve bank. Both ACR's and Echo type chargers turn off or isolate at well above "full battery voltage".... 

I guess in an ideal world, someone would make a global regulator- controlling the alternator, providing two isolated outputs to separate banks (the smaller bank doesn't need to be efficient), but recognizing when another source of charge is present (on a separate input terminal) and charging the reserve bank from that source. I'm not aware of such a regulator.

These are called ACR's, Combiners, Echo Charger's, Duo Chargers and Sterling Battery to Battery chargers. If you wire everything to house first these devices work beautifully. It is simply getting over the assumed hurdles of how they work, or what they do or don't do, that is usually the toughest...

[Stu Jackson]

A.     ....105 A L-N alternator (internally regulated, connected to the 1-2-B switch), a Yandina ACR (100 A)and BMV-700 monitor.

B.    A few things would improve the situation: 1) direct connection from the AO to the house bank would have limited the alternator voltage a bit and the massive current to the reserve bank, but not the leak from the house bank to reserve.  2) I should have started the engine for charging on the house bank, 3) Given 1) and 2) I think an Echo Charge would not have allowed this to happen at all. 

C.   I guess in an ideal world, someone would make a global regulator- controlling the alternator, providing two isolated outputs to separate banks (the smaller bank doesn't need to be efficient), but recognizing when another source of charge is present (on a separate input terminal) and charging the reserve bank from that source. I'm not aware of such a regulator.

This is a shorter version of what Maine Sail covered.

A.  Assuming the alternator is connected to the C post of the 1-2-B switch, get it off there and route it to the house bank.

B.  1)  Absolutely.  We've been "preaching" this for the past 18 years or so.  See the "Electrical Systems 101" topic.  The leak isn't an issue, since battery acceptance governs what any bank can take.   2)  Yes, just another reason to not have the alternator charge through the switch.

C.  They do, it's called a dual output alternator.  You don't need one.  What you have is fine, but until you move the AO to the house bank, you will constantly be having to "manage" you system.  Once you do so, you can simply leave the switch on the house bank and forget about all that management crap.   Also, Maine Sail is right, the ACR performs just this function.

Good luck.

[tgsail1]

How was the -60Ah determined? If running off the reserve bank the Ah counter, ideally connected to the house bank, should not have been recording an Ah deficit? Was this a guesstimated -Ah consumption?

Can't remember for sure, but I believe I have the reserve battery ground located at the same point on the shunt as the house bank.I know that is not optimal, but it let me charge both batteries off one charger two outputs before the ACR was installed. I should really move it now.

Similar question, how was the current output of the alternator determined? If properly wired, the Ah counter reads the current flowing into or out of the house bank, not the start bank? Did you DC clamp-meter the alt output?

As stated above, measured with the BMV-700 shunt. It was measuring the net into both batteries

If the house bank resting voltage was higher than the charging voltage of the start battery, then that could be possible, for house bank help raise the overall system voltage to a higher overall parity. However, the combiner does not combine with the house bank until 13V and a 30 second delay so this is already above the house banks natural resting voltage and at +0.3V above resting it really becomes a charging voltage for the house bank as well as start..

Even at a 13.0V + 30 seconds to combine you still should have been in bulk. Once the 13.0V threshold has been passed a 30 second delay starts before combining occurs. By the time you combined you may have been at 13.1V +/-. I am guessing either there is a bad regulator voltage sense issue, causing the current limiting you saw, or the alt was just heating up and putting out less current. With a terminal voltage of 13V for combine this is well above the 100% SOC voltage point for the house bank meaning no current should flow out of it and into the other bank and most likely a few mA were flowing into the house bank. How and where you measured all this is important to determine what you saw....

Fair enough. So the reserve battery is sitting very low at 60% discharge might have a voltage of 12V. The alternator comes on pumping as much it can. It will switch to absorption when the voltage at the alternator (there is no remote sense) reaches a set point. When the house battery switched on, does it not raise the terminal voltage? Couldn't that fool the alternator into absorption? Alt heating is certainly possible. It was toasty.

How was this +40Ah determined? In the description you have described discharging and recharging a bank that you should not have been able to see any + or - current or + or - Ah data for? Perhaps it's possible your bank 1 & 2 are backwards..?

See answer above, measured through shunt

The exact same way the combiner combines, it opens or un-combines. Once the 13V threshold has been met a 1 minute timer starts before the relay opens. With lightly loaded healthy batteries it would not be uncommon to hold a surface charge on the plates above 13.0V for 4 minutes or more. AGM's can do this for 30+ minutes and in cold weather disconnected can hold 13+ for weeks. Again this is nothing more than a surface charge and is not quantifiable usable capacity. So essentially 13.0V + is a temporary surface charge and above the resting 100% SOC voltage of the battery.

Interesting. Assume that the ACR is not defective in turn-on. Then for the house bank surface charge to not dissipate quickly, the reserve bank at less than 100% SOC, would have a terminal voltage high enough to limit current flow between the batteries until the higher SOC battery bleeds off and the ACR opens.

Actually you were limited by Ohm's law. When batteries are at voltage parity very little current can move between the banks. The combiner creates what amounts to voltage parity because it physically parallels the batteries..
 
While some very small currents were flowing this "charge" or discharge of the house bank is very minimal because the batteries are in parallel and at virtual voltage parity. In order for current to flow into a battery there needs to be a voltage differential between the SOC voltage and the charging voltage.. The smaller this voltage differential the less current can flow and this is all governed by Ohm's law.... At parallel voltages very little current moves between the banks during the time between discontinuation of charging and the 13.0V threshold and 1 minute delay.

This is exactly the point of my question. If the resting voltages of the two banks are different, when connected, current will flow until the voltages are equal. The current flow is limited by wire and internal resistance, so it takes time for this equalization to occur.

If the Ah counter was wired correctly and that bank had only been used via the combiner then the net -Ah should have been displayed. If the batteries had a surface charge from the shore charger then wiping this out via combining can happen but the net decrease in quantifiable Ah capacity will be very small.

Wired as described, the AH counter measures energy into both banks. The last charge the HB would have seen was the alternator (during the cruise over because the ACR would have paralleled the banks) about 20 hours earlier. I doubt my HB are in such good shape they can hold a surface charge that long.

Again the cut-in & cut-out voltages are well within a voltage range, above resting full voltage,  that it makes very little net difference to either bank. For example if the charge source were very small, current wise, and wired to the smaller, usually full, starting battery the relay would simply fire on at 13V then drop out again once house voltage dragged the new parity voltage down below 13V.  The solar array would then need to charge the start battery back to 13.0V + 30 seconds before it would try charging the house again. This is called "relay cycling" and can happen with small solar arrays fed to a start battery first, and with large differences in bank size, and why all charge sources should really feed house first.

Fair enough. I guess the point is that even if the house bank is fully discharged, start bank fully charged, and the ACR is switched on, alternator turned off, it will not take long to drag the start battery to under 13V where the ACR will open. At 12.9 V the battery is still at full charge. What's curious is that it took about 5 min for this to happen in my case- something to look into.

If the house bank was already full it would have just brought it to 13V+ very quickly then combined with the depleted start battery. The start battery still would have taken the same current if the terminal voltage was the same, which in parallel it would have been except for the minor voltage drop across the relay. The battery will only take high current for a short duration, even AGM batteries. At 70A you were essentially charging at .66C and target voltage would have been attained very, very rapidly. Heck at .4C even an AGM bank from 50% SOC is only in bulk for 19 minutes or so. A flooded battery with much higher internal resistance, charged at .66C, and you're talking somewhere under 2 minutes in bulk from 50% SOC....

That would explain the drop in current from the alternator. I didn't realize the terminal voltage could rise that quickly. I think the voltage drop across the relay may not be that minor though. The Yandina relays have 6 ft ~10 ga(?) wires which are intended to limit the current ~100A, so there must be some 10mohms resistance in the wire.

Again Ohm's law dictates the transfer of current at parity voltage. With no charging source present and voltage parity mA of current move between the banks at best. You essentially only have 1 minute of paralleling beyond the battery dropping to 13V which is above the 12.72V +/- resting voltage of a 100% SOC flooded battery.

If we assume your house bank was at 100% SOC, at 12.7V, and you discharged the reserve battery to 50% SOC or about 12.2V and figure that in order to get any sort of reasonable current to flow into the 12.2V battery we need at least +.25V on top of that, then the voltage differences at the point we parallel the batteries (without a charge source present) = 0.25V differential. If we then assume a .007 Ohm internal resistance of these batteries, pretty typical, means an initial in-rush or current surge from one battery to the other of somewhere around a very, very short peak of 36A of current. As the voltages equalize or attain parity, this happens very quickly, lets assume 12.600V and 12.596V at .007 Ohm we can see that only 0.57A can move between the batteries with no charge source present.

The longer the batteries remain in parallel the more voltage parity is attained until we are looking at voltage differences out to the thousandths or 12.6000V vs. 12.5998V....

With this it is easier to see that 1 minute of delay at 13.0V, after charging has been removed, really takes almost nothing from either bank because by design this is all occurring above the 100% resting SOC voltage point..

Ok a lot here..Not sure why the calc in your example is not .5V/.007= 71 A, but the point is taken that the protection between the batteries is provided by the setpoint on the ACR being above the 100% SOC point.

I guess in an ideal world, someone would make a global regulator- controlling the alternator, providing two isolated outputs to separate banks (the smaller bank doesn't need to be efficient), but recognizing when another source of charge is present (on a separate input terminal) and charging the reserve bank from that source. I'm not aware of such a regulator.

These are called ACR's, Combiners, Echo Charger's, Duo Chargers and Sterling Battery to Battery chargers. If you wire everything to house first these devices work beautifully. It is simply getting over the assumed hurdles of how they work, or what they do or don't do, that is usually the toughest...

I meant a single smart controller to handle combining and charging (whether alt or solar or shore)- a battery master controller if you will. Not sure what difference wiring AO direct to the HB would have made. It would have put the charging source on the other side of the ACR which might have slowed the reserve batt current a little beyond that?. The remaining curiosity is why the ACR remained active for 5 min after charging.

C.  They do, it's called a dual output alternator.  You don't need one.  What you have is fine, but until you move the AO to the house bank, you will constantly be having to "manage" you system.  Once you do so, you can simply leave the switch on the house bank and forget about all that management crap.   Also, Maine Sail is right, the ACR performs just this function.

With my current setup (factory wiring plus ACR), I do not have to manage the system. I normally leave it on the house bank and don't worry about it-this case was an exception where I was testing the reserve batt. What's more, the AO (which is connected to the starter and then to the 1-2-B ) does not require running an additional heavy wire off the engine and, most importantly, can be easily isolated from the batteries for engine work without installing an additional switch. The only downsides are that I have to be careful not to switch to "off" while the engine is running- pretty easy since I don't touch it and the switch is out of the way, and there is a little more drop between the AO and the battery.