Myth of alternator damage cause by BMS disconnect?

[tanglewood]

How big a problem is this, really? I think it stems from battery switches that suddenly leave a running alternator with an open circuit output. The sudden change in current induces a reverse voltage that can spike quite high and cause damage to the rectifier diodes.


But the situation is different with a BMS disconnect. At least I think it is in many, if not most circumstances. The difference is that you are removing the battery from the house power circuit, not removing the alternator from the house power circuit. The difference is the house power load, which remains connected to the alternator when a BMS disconnects the batteries, is there to absorb any inductive kick-back.


So the alternator DOESN'T go open circuit. Yes, the load can change dramatically based on the relative sizes of the house loads and battery charge load. But it doesn't go open circuit. So for those of you with more current EE practice than me, what do you think? I know to actually quantify this you need to know a lot of specifics about a particular boat, and it will be different for every boat. But it's surely LESS of an issue, and perhaps no real issue at all.

[PaulCrawhorn]

A cheap small lead batt kept hard-wired to the alt output is good enough to buffer the resulting spike/surge.

[donradcliffe]

In my RV the 180 amp alternator goes to the starting battery, which is connected to the LFP house battery by a voltage sensitive relay that opens below 13.5 volts. No problems after 10 months of living aboard.

[goboatingnow]

Well it’s got nothing to do with “ open circuit “ it’s all to do with the speed of current change

The typical alternator has about 350 uH of inductance

So V= L di/dt

So for say a 40 A change in 10 uS then the inductor will generate ( assuming it’s not clamped ) and output change of 140V

Whereas the same change of current but over 100mS generates 0.14v.

Now modern alternators typically have avalanche diodes and fast acting regulators. Fast spikes are limited to 35V

So you can see if the BMS cutoff is designed to say take 100 mS ( abs hence can dissipate heat during that phase ) then no alternator spike will occur in the first place.

Such slow switch off is easier to design into mosfet switching

Relays and so forth have switch bounce in the order of 5 uS or less per bounce so you see these are not great

[CarlF]

I agree that this risk is wildly exaggerated in today's typical cruising boat installation.

To have a BMS disconnect is evidence of an improper installation.

I have never had my BMS disconnect for the simple reason that all charging sources are set well below the voltage that the BMS would disconnect. In my case they are set to a 14.1v bulk and 13.4v float (from the battery manufacturer) while the BMS won't disconnect until 14.6v. And the maximum charge amps for my solar panels plus alternators running together is less than 50% of the maximum charging amps for the bank.

And since my bank consists for five drop-ins in parallel, all five would have to disconnect to bother the alternator.

Still, if the very small risk bothers you it's easy to:

a) install a Sterling Alternator Protector for less than $100

b) Parallel a lead/acid battery as suggested previously

c) do a+b

[tanglewood]

Originally Posted by CarlF

I agree that this risk is wildly exaggerated in today's typical cruising boat installation.

To have a BMS disconnect is evidence of an improper installation.

I have never had my BMS disconnect for the simple reason that all charging sources are set well below the voltage that the BMS would disconnect. In my case they are set to a 14.1v bulk and 13.4v float (from the battery manufacturer) while the BMS won't disconnect until 14.6v. And the maximum charge amps for my solar panels plus alternators running together is less than 50% of the maximum charging amps for the bank.

And since my bank consists for five drop-ins in parallel, all five would have to disconnect to bother the alternator.

Still, if the very small risk bothers you it's easy to:

a) install a Sterling Alternator Protector for less than $100

b) Parallel a lead/acid battery as suggested previously

c) do a+b

I agree with the above. Just trying to unravel the actual effect of an on-going electric load on the alternator after a battery disconnects. That is, after all, what a & b above are doing, just via different mechanisms.

[Cadence]

Originally Posted by CarlF

I agree that this risk is wildly exaggerated in today's typical cruising boat installation.

To have a BMS disconnect is evidence of an improper installation.

I have never had my BMS disconnect for the simple reason that all charging sources are set well below the voltage that the BMS would disconnect. In my case they are set to a 14.1v bulk and 13.4v float (from the battery manufacturer) while the BMS won't disconnect until 14.6v. And the maximum charge amps for my solar panels plus alternators running together is less than 50% of the maximum charging amps for the bank.

And since my bank consists for five drop-ins in parallel, all five would have to disconnect to bother the alternator.

Still, if the very small risk bothers you it's easy to:

a) install a Sterling Alternator Protector for less than $100

b) Parallel a lead/acid battery as suggested previously

c) do a+b

An electrolytic capacitor might be better than a lead/acid battery since it doesn't require maintenance.

[team karst]

Originally Posted by Cadence

An electrolytic capacitor might be better than a lead/acid battery since it doesn't require maintenance.

That might work. The cap smooths out the V transition from full load to zero load. It will become a race to the finish; either the VR will reduce the field current in time to save the cap from OV, or, it won't. A strict calculation on the VR time constant would be in order. Even a 10000uF 25V cap won't take 60 inbound Amps for long, depending on its V rating.

[goboatingnow]

Originally Posted by team karst

That might work. The cap smooths out the V transition from full load to zero load. It will become a race to the finish; either the VR will reduce the field current in time to save the cap from OV, or, it won't. A strict calculation on the VR time constant would be in order. Even a 10000uF 25V cap won't take 60 inbound Amps for long, depending on its V rating.

10000uF would be far too small values in excess of 1F would be needed , batteries are equivalent to 100F

The VR can do nothing to regulate the load dump response. The output spike is not a function of the field current at that point and is not influenced by it.

[tanglewood]

Originally Posted by goboatingnow

Well it’s got nothing to do with “ open circuit “ it’s all to do with the speed of current change

The typical alternator has about 350 uH of inductance

So V= L di/dt

So for say a 40 A change in 10 uS then the inductor will generate ( assuming it’s not clamped ) and output change of 140V

Whereas the same change of current but over 100mS generates 0.14v.

Now modern alternators typically have avalanche diodes and fast acting regulators. Fast spikes are limited to 35V

So you can see if the BMS cutoff is designed to say take 100 mS ( abs hence can dissipate heat during that phase ) then no alternator spike will occur in the first place.

Such slow switch off is easier to design into mosfet switching

Relays and so forth have switch bounce in the order of 5 uS or less per bounce so you see these are not great

Thanks. This is what I was looking for. But as the voltage spike takes place, the current to the house loads will increase, draining off the energy stored up in the alternator windings. Doesn't that have a mitigating effect on the magnitude of the voltage spike? Isn't this exactly how a connected lead battery or clamping diode reduces the spike, but conducting current in response to the voltage increase? Obviously the characteristics of a clamping diode and lead battery are different from the house load, but isn't the same principal at play?

[lmxr]

Originally Posted by tanglewood

Thanks. This is what I was looking for. But as the voltage spike takes place, the current to the house loads will increase, draining off the energy stored up in the alternator windings.[...]

That's what happens indeed: the more stuff connected, the lower the voltage spike. The only issue is that most stuff does not tolerate even low voltage spikes, nor does it tolerate an increase in current through it.

[goboatingnow]

Originally Posted by lmxr

That's what happens indeed: the more stuff connected, the lower the voltage spike. The only issue is that most stuff does not tolerate even low voltage spikes, nor does it tolerate an increase in current through it.

No electronics connected to a “ car “ alternator should fail on load dumps. The iso tests require all auto electronics to cope. Boat electronics should and generally does follow and has TVS diodes to handle excess voltage. The cheap stuff doesn’t

[lmxr]

Originally Posted by goboatingnow

No electronics connected to a “ car “ alternator should fail on load dumps. The iso tests require all auto electronics to cope. Boat electronics should and generally does follow and has TVS diodes to handle excess voltage. The cheap stuff doesn’t

I agree about auto electronics being quite well protected. Less sure about boat electronics though. Would you mind sharing a video with a demonstration that boat electronics indeed can handle a load dump?

[goboatingnow]

Originally Posted by tanglewood

Thanks. This is what I was looking for. But as the voltage spike takes place, the current to the house loads will increase, draining off the energy stored up in the alternator windings. Doesn't that have a mitigating effect on the magnitude of the voltage spike? Isn't this exactly how a connected lead battery or clamping diode reduces the spike, but conducting current in response to the voltage increase? Obviously the characteristics of a clamping diode and lead battery are different from the house load, but isn't the same principal at play?

The issue is the spike voltage not the current. But it’s actually quite easy to design in TVS diodes into the electronics of loads to make them withstand the spike.

The issue has been older alternators tended to destroy their diodes.

Today most integrated alternator ICs are rated to withstand 300 V

But yes you are correct the bigger the load ( or on effect a current sink the less issue there is with inductive kick back. ( the alternator regulator is also working to reduce kick back

Put a scope of a boats 12v system abs watch all sorts of nasties fly around

[RaymondR]

Or disconnect the excitation rather than the alternator output.