Charging batteries from alternator

[Col_T]

If I understand correctly, a LA battery will take whatever current the alternator can generate in 'bulk' charge mode. In 'absorption' mode, the battery is no longer able to take unlimited current, so the alternator output current drops, and the voltage rises, until the current tapers down to a fully charged point, often defined as 1 or 2% of the battery's AH rating at 14.4v.

 

Is this is correct, the alternator is worked pretty much flat-out during the early stage of charging but then eases-up to come to a controlled near shut-down.

 

Charging an LiFePO4 from an alternator seems to be slightly different, in that the LiFePO4 (lithium) battery can take pretty much whatever the alternator can throw at it right up to fully charged, at which point whatever BMS is fitted should 'disconnect' the alternator to prevent the battery being over-charged. The alternator does not like this sudden 'chop' at all, which is why many lithium installations include LA battery(s) - the BMS switches out the lithiums, the alternator charging current goes to the LA so the alternator is happy.

 

So, now to the question. How does the alternator deal with this if the LA battery is already fully charged when the BMS chops the lithiums from the charging circuit?  The alternator will not be able to gently ease-up as it would if it was charging an LA only installation, it would be more like going from 'bulk' straight to 'float' - will this damage the alternator at all?

 

I'm not thinking of going down the lithium route, just pure curiosity!

[Keith M]

What happens to the alternator if it is not spinning at a speed so that it does not overheat

The best way to recharge lithium is to use a battery to battery charging system.

[Steve42]

Even if the LA battery is fully charged is should absorb the load dump.

[WotEver]

1 hour ago, Col_T said:

How does the alternator deal with this if the LA battery is already fully charged when the BMS chops the lithiums from the charging circuit?

As Steve says above, the alternator is not suddenly seeing an open circuit which would allow the voltage to temporarily soar up to a destructive level. The LA battery will prevent that. 

[Dr Bob]

1 hour ago, Col_T said:

 

So, now to the question. How does the alternator deal with this if the LA battery is already fully charged when the BMS chops the lithiums from the charging circuit?  The alternator will not be able to gently ease-up as it would if it was charging an LA only installation, it would be more like going from 'bulk' straight to 'float' - will this damage the alternator at all?

 

 

I am not sure what you meant by 'gently ease up'. I am no expert on alternator circuits - you need the experts like @Sir Nibble or Nick or @MoominPapa or wotever. MP explained it to me in a post a month ago but I cant find it.

As I understand it, the alternator is spinning putting out its current at a certain voltage level. If the connection to the Lithiums is broken, the current stops flowing and the voltage starts to rise very rapidly (ie milliseconds) and can reach values of over 100V. That peak can then destroy the diodes in the alternator. With LA's in circuit, the rising voltage is constrained regardless of whether they are full or not. I asked MP the question if having a much bigger alternator would risk making this worse but apparently not.

 

eta....wotever beat me to it agreeing with Steve.

Edited March 14, 2020 by Dr Bob

[WotEver]

1 minute ago, Dr Bob said:

As I understand it, the alternator is spinning putting out its current at a certain voltage level. If the connection to the Lithiums is broken, the current stops flowing and the voltage starts to rise very rapidly (ie milliseconds) and can reach values of over 100V. That peak can then destroy the diodes in the alternator. With LA's in circuit, the rising voltage is constrained regardless of whether they are full or not. 

Yes, this is correct. 

[Tony Brooks]

8 minutes ago, Dr Bob said:

I am not sure what you meand by 'gently ease up'. I am no expert on alternator circuits - you need the experts like @Sir Nibble or Nick or @MoominPapa or wotever. MP explained it to me in a post a month ago but I cant find it.

As I understand it, the alternator is spinning putting out its current at a certain voltage level. If the connection to the Lithiums is broken, the current stops flowing and the voltage starts to rise very rapidly (ie milliseconds) and can reach values of over 100V. That peak can then destroy the diodes in the alternator. With LA's in circuit, the rising voltage is constrained regardless of whether they are full or not. I asked MP the question if having a much bigger alternator would risk making this worse but apparently not.

I was about to agree as Wotever posted.

 

I think the OP also has the cart before the horse in the understanding of charging. It is correct as to what the current does but that is a result of having the voltage regulated and the batteries effective "resistance" to current flow rising as it charges. The alternator regulator is only controlling voltage.

 

I think if  were going lithium not only would I keep the starter LA in parallel with the lithiums but I would doctor the alternator so I could open circuit the rotor feed or return you could then shut it down in an instant without causing that voltage spike.

Edited March 14, 2020 by Tony Brooks

[Col_T]

Thanks for the replies.

 

I think I may not have phrased the question correctly. I understand that the LA battery is retained as a load dump when the BMS switches out lithiums out of the charging circuit; I was more interested in the effect that has on the alternator which, in my mind, goes from working flat-out to idling in fractions of a second e.g. is there any 'shock' loading that may damage the alternator over time?

 

@Tony Brooks - my understanding, or perhaps misunderstanding, is that the battery chemistry effectively controls how much current it can take in 'absorption'. Is the alternator regulating voltage to produce a constant output wattage, or does it sense the current and just output that at the highest voltage it can?

 

[MoominPapa]

My understanding is this. When a load dump occurs, it takes a finite time for the excitation current in the alternator to drop, due to the inductance of the field coil. In the period between the load dump and excitation falling, the alternator is still running with the old excitation, but with a much smaller current, and larger load resistance. This causes the voltage at the alternator terminals to spike to a high level, and that breaks things.

 

So, why doesn't switching off the heated rear windscreen in your car (a big load dump) spike the electrics and break things? The answer is that a LA battery has a very non-linear voltage-current curve. Anyone who has played with an adjustable power supply to charge a battery knows this: the current is zero until the PSU voltage reaches the open-circuit battery voltage. Another volt or two gets you to a normal charging current. An extra 10v generates very large charge currents, well beyond the battery specs, if your PSU is meaty enough to achieve that.

 

When the load dump happens, the battery is still in circuit, and as the voltage starts to rise, the charge current through the battery rises rapidly, and that limits the voltage rise. A voltage spike which can do damage to anything on the system that's sensitive gets turned into a current spike through the battery and a small voltage rise. Even if the current spike is very large, it's of very short duration, and the chemical and thermal inertia of the  battery means it hardly notices it. This applies to Lithium cells as well as LA cells: they have the same sort of current-voltage curve at higher than normal voltages. If you disconnect a charging battery, you've simultaneously generated a load-dump AND removed the device that is protecting the system from load dumps. That's which it's a bad idea to pull the wire of the car battery terminal while the engine is running, and why you can't safely do charge termination for a Lithium bank by just disconnecting the bank 

 

BUT If you have a LA battery and a Li battery in parallel, when you disconnect just the Li bank, the LA stays connected. Disconnecting the Li bank dumps the load on the alternator, but the LA battery is still in circuit and absorbs the spike, just like when you dump load by switching off the headlights in your car. This is why combined LA/Lithium systems are useful as retrofit in places where you can't add smart control of the charging sources: it allows you to do charge termination on the Lithiums (and to implement under/over-voltage and temperature protection) without having to do it by retrofitting controls to dumb alternators and chargers and solar controllers.

 

It certainly works for me.

 

MP.

 

 

10 minutes ago, Col_T said:

Thanks for the replies.

 

I think I may not have phrased the question correctly. I understand that the LA battery is retained as a load dump when the BMS switches out lithiums out of the charging circuit; I was more interested in the effect that has on the alternator which, in my mind, goes from working flat-out to idling in fractions of a second e.g. is there any 'shock' loading that may damage the alternator over time?

 

What kills alternators is excess voltage across the diodes. A battery in circuit dumps the energy of the load dump by diverting the current through the battery, so the voltage doesn't rise far enough to affect the diodes.

 

MP.

 

Edited March 14, 2020 by MoominPapa

[Sir Nibble]

3 hours ago, Dr Bob said:

I am not sure what you meant by 'gently ease up'. I am no expert on alternator circuits - you need the experts like @Sir Nibble or Nick or @MoominPapa or wotever. MP explained it to me in a post a month ago but I cant find it.

As I understand it, the alternator is spinning putting out its current at a certain voltage level. If the connection to the Lithiums is broken, the current stops flowing and the voltage starts to rise very rapidly (ie milliseconds) and can reach values of over 100V. That peak can then destroy the diodes in the alternator. With LA's in circuit, the rising voltage is constrained regardless of whether they are full or not. I asked MP the question if having a much bigger alternator would risk making this worse but apparently not.

 

eta....wotever beat me to it agreeing with Steve.

You won't exceed the peak inverse voltage of the diodes (typically 200V for a 12V machine) with a 12V lead acid battery connected. 

[Tony Brooks]

2 hours ago, Col_T said:

Thanks for the replies.

 

I think I may not have phrased the question correctly. I understand that the LA battery is retained as a load dump when the BMS switches out lithiums out of the charging circuit; I was more interested in the effect that has on the alternator which, in my mind, goes from working flat-out to idling in fractions of a second e.g. is there any 'shock' loading that may damage the alternator over time?

 

@Tony Brooks - my understanding, or perhaps misunderstanding, is that the battery chemistry effectively controls how much current it can take in 'absorption'. Is the alternator regulating voltage to produce a constant output wattage, or does it sense the current and just output that at the highest voltage it can?

 

1. As I understand it its just the high induced voltage surge that does damage but some makers fit a free wheel into the pulley that will allow the rotor to over-run the pulley but if the alternator is under load I bet it will slow down PDQ, especially as the induced voltage will also cause  a burst of high current.

 

2. Correct, once the regulator starts to come into play its the batteries that control the current flow at that particular voltage. Increase the regulated voltage and the current will rise, decrease it and the current falls.The regulator just controls voltage, that's all.

[WotEver]

5 hours ago, Col_T said:

Is the alternator regulating voltage to produce a constant output wattage, or does it sense the current and just output that at the highest voltage it can?

Neither really.
 

The alternator has a maximum power that it can output. As an example let’s use 1000W. If the battery SoC is very low then both its internal resistance and terminal voltage will be very low. So the alternator works flat out, supplying its maximum current, and the battery sucks this up while its voltage slowly rises. Eventually the battery voltage will rise to the level at which the alternator regulates - somewhere around 14.4V. The alternator regulator now keeps that voltage at 14.4V while the batteries happily take 70A charging current. Slowly, the current demanded by the batteries reduces as their internal resistance rises correspondingly. All the time the alternator regulator simply keeps the voltage at 14.4V.
 

So the power output of the alternator slowly rose to 1000W and then started reducing again. At no point does the internal regulator care what the power output is, if merely prevents the voltage from rising above its set level. 

[Col_T]

Excellent. Very many thanks to everyone that has contributed.

 

I would say this is CWDF at it's best - the question answered, with loads of relevant, detailed information - I couldn't ask for more.

 

Thanks to all.

[Dr Bob]

.......and no jokes? We're slipping!

[Alan de Enfield]

1 hour ago, Dr Bob said:

.......and no jokes? We're slipping!

Are you 'fishing' ?

[WotEver]

18 minutes ago, Alan de Enfield said:

Are you 'fishing' ?

He wasn’t codding. 

[Alan de Enfield]

Just now, WotEver said:

He wasn’t codding. 

Skating on thin ice tho'

[BEngo]

But he was not out of plaice.

 

 

[WotEver]

1 hour ago, Alan de Enfield said:

Skating on thin ice tho'

Which is a bit dodgy when you’re getting broad across the bream. 

[Dr Bob]

6 minutes ago, WotEver said:

Which is a bit dodgy when you’re getting broad across the bream. 

I'm not keen on broad beans.

[Alan de Enfield]

1 minute ago, Dr Bob said:

I'm not keen on broad beans.

Much more comfortable than narrow beans.

[MtB]

 

Oh for cod's hake, can we drop the fish puns?

 

Although I suppose without them, this plaice would have no sole....

[Dr Bob]

1 hour ago, Mike the Boilerman said:

 

Oh for cod's hake, can we drop the fish puns?

 

Although I suppose without them, this plaice would have no sole....

I may have to think up some tank jokes.

[philjw]

 

You'll be on the right track

[nicknorman]

Anyway, fish jokes aside, modern alternators tend to have zener diodes rather than plain high PIV diodes. Our Iskra has zeners which are either 36 or 38v, can’t remember. Automotive stuff is specified to be able to withstand a transient of 40v or thereabouts. So really, such an alternator should be relatively happy with a battery disconnect - the transient energy being absorbed by the zeners. However of course the boat’s connected systems probably aren’t designed to cope with this sort of transient! And personally I wouldn’t want to routinely put my alternators zeners into reverse conduction!

 

Which brings me to a second point which is that it is not good practice to rely on the “ultimate battery protection” of a disconnect, to routinely occur at the end of charging a Li battery. The charging should be designed not to exceed normal cell voltages, the disconnect should only come into play when something has gone horribly wrong. In other words, it should be an emergency system not a normal system. Without 2 layers of protection it is likely to eventually end in tears, just like the 737 Max (only with fewer fatalities, hopefully).

Edited March 15, 2020 by nicknorman