Myth of alternator damage cause by BMS disconnect?

[goboatingnow]

Originally Posted by team karst

At this level of detail, it might be instructive to further define a TVS. Or even more common TVSS.
Those acronyms better describe their purpose, but not what they are. You have MOV’s, special semiconductor diodes, and gas discharge devices. They each have their purpose for protection. In fact, I will hereby call my start Pb
battery a TVS in this application.

I’m specifically talking about automative qualified high power TVS diodes.

[team karst]

Good choice, ive used those in my last boat. And, of course, several in parallel will work even better.
I happened to see a design yesterday that was a over voltage crowbar. It purposely blew a small fuse to protect a downstream device. That can be highly effective for long duration OV events.
Lets say your solar controller runs amok, and the house BMS opens up the battery bank totally. What happens to the house Vnow?

[rgleason]

The alternative is a far more precise solution based on high current mosfet switching , coupled with precision IC over voltage sensing . This can drive either a series mosfet or a shunt mosfet.

Shunt solutions have the advantage of quenching the inductive spike and hence protect all devices in the circuit including the alternator.

Hence today you could easily fit a shunt protective device and remove any issues with unexpected alternator disconnects. This is a far better solution then deploying a “buffer battery “

I suppose this is the approach being taken by Balmar with their forthcoming Balmar APM-12 and APM-24? But these appear to be just protecting the alternator.

How would you suggest designing and fitting a shunt protective device for a small 12vdc system with a 160a alternator and 200-300 LFP?

Would this be based on your description "high current mosfet switching , coupled with precision IC over voltage sensing . This can drive either a series mosfet or a shunt mosfet."


See this Perfect Switch Mosfet Dual Rectifier Isolator (Gen 3.0) for charging two types of batteries
https://perfectswitch.com/isolators/...fier-isolator/

Rectifier-relay


LVD


Li Battery PowerGate Designed to protect low or high voltage

[rgleason]

Here is the example that I found some years ago
https://perfectswitch.com/applications/marine-lifepo4/


"The endless search for power" 8 pages


https://perfectswitch.com/ad-09262017-1-r5/


https://perfectswitch.com/ad-07102019-1-r2/




Also http://www.getlithium.com/29l_Perfec...n4-300Amp.html

[RaymondR]

Originally Posted by PaulCrawhorn

No one is discussing LiPo batteries here.

So let's do something about the confusion being caused by the acronyms. We are talking about Lithium Iron Phosphate sometimes acronymed as LFP and sometimes as LiPo. Is that correct?

That is a completely different technology, absolutely nothing to do with LFP.

And obviously any lead battery needs to be regularly replaced, "any old" simply means plain, "nothing special".

If you want to spend a lot feel free, but in this specific function there is no need.

I've had lead acid batteries fail when paralleled with other lead acid batteries and they tend to cook but because of the fairly high internal resistance of the current source and consequent large voltage drop I noticed the failure before a complete melt down occurred and was able to isolate the defective unit. Not sure this would occur with lithium, and son't really care to carry out the experiment on a fairly flammable plastic boat.

And since most can just use their normal Starter for this,then there is no extra cost in money or time & attention.

Time, care, and attention and lot's of conversation are often required to properly address issues in the marine sphere and one should not stint on any of them.

This is just scaremongering over a well settled question.

AHHA, the old scientific consensus raises it's ugly head again.It's becoming the "universal panacea" for folks who don't want to reason their opinions with others.

[RaymondR]

OK, I'm getting old and maybe having a "silly old bastard" moment but will sometime explain to me why trying to mitigate the effects of a large voltage excursion is superior to preventing the excursion by removing the excitation of the alternator so that no excursion occurs.

Controlling the excitation of the DC generator to control the current to the DC motor was the method used for output power control with the old diesel electric locos very successfully prior to the introduction of the AC/DC SCR control systems.

It might be a problem with alternators using internal regulators (although the ones I use could be readily modified) but with those using external regulators having the BMS charge relay disconnect the current to the external regulator and thereby the excitation current to the alternator rotor and prevent a large voltage excursion in the alternator output rather than having it disconnect the alternator from the battery and thereby causing a large voltage excursion appears to be a viable solution.

Someone please do me a favour and explain why this would be a bad idea or would not work.

[goboatingnow]

You are right. The best method is to control the field excitation . But this isn’t necessarily easy on some alternators and many people don’t want to “ break into “ their alternators in the first place.

Hence the issue of protecting both the alternator and the down steam circuitry from an abrupt disconnect

Of course an alternative method is not to have an abrupt disconnect. For example a mosfet switch can be programmed to turn off over a number of milliseconds this significantly reduces the Ldi/dt computation.

[goboatingnow]

Originally Posted by rgleason

I suppose this is the approach being taken by Balmar with their forthcoming Balmar APM-12 and APM-24? But these appear to be just protecting the alternator.

How would you suggest designing and fitting a shunt protective device for a small 12vdc system with a 160a alternator and 200-300 LFP?

Would this be based on your description "high current mosfet switching , coupled with precision IC over voltage sensing . This can drive either a series mosfet or a shunt mosfet."


See this Perfect Switch Mosfet Dual Rectifier Isolator (Gen 3.0) for charging two types of batteries
https://perfectswitch.com/isolators/...fier-isolator/

Rectifier-relay


LVD


Li Battery PowerGate Designed to protect low or high voltage

Yes but it’s unclear what the turnoff speed is. You actually want a slow turnoff

[rgleason]

In the example "The endless search for power" 8 pages

Battery Banks
1. Lithium House Bank
2. FLA/AGM Starter Bank

BMS
Needs to control charge and load relays, and moniter and balence cells.
The article describes the Perfect Switches and BMS in greater detail.

Three Buses
1 Ground Bus
2. Charge Bus - Solar & Alternator
3. Load Bus - House panel

BMS will disconnect LFP Charge bus relay when fully charged.
BMS will disconnect LFP Load bus relay when depleted.

Alternator Disconnect Protection
"Charge bus relay connects with the SLA battery before it switches off from the lithium bank. So if the charge bus switch is opened, the alternator immediately sees the lead acid battery in the system and adjusts its charging according to the new battery. The lithium battery is removed from the charging bus but is still running the loads through the load bus. As the charge level falls back to within the programmed range the relay brings the charge bus back on line and the SLA battery is disconnected. "

Charge and Load Bus Relays
"The two relays connected to the charge and load bus are for “system failure” incidents. Each of the charging sources should be programmed to keep the lithium pack within its limits without throwing the relay. In the best possible design we want the whole system to operate without ever throwing one of these relays."

Make before break Switch and Charging Start Battery
"We haven’t used a battery isolating relay. Instead, we worked with the good people at Perfect Switch to develop a “make-before-break” single pole double throw (SPDT) Mosfet solid state relay that will comfortably handle 250 amps at 12 volts continuous. This won’t allow us to simultaneously charge the lead acid battery through the charge bus.

"However, we have a solar charge regulator, which will put a small charge to the lead
acid battery independent from the charge bus. ...keep it well charged as the current draw on this battery will be quite low. In addition, we have placed a manual override switch in the system, which will allow us to bring the SLA battery on to the charge bus if it ever requires a quick top-up.

[wsmurdoch]

Originally Posted by Dieseldude

The varistors seem like a good idea. They will insure that spikes above 14 volts get shorted to ground. But when they start conduction at 14 volts, a voltage drop out can occur. For a 12 volt system, a slightly high varister voltage value might be better, say 16 volts to avoid needless triggering, as a fully charged battery can approach 14 volts.


A varistor alone can be a fire hazard though. They are intended to react to fast transient voltages only, as their heat dissipation ability is limited. A sustained over voltage will cause them to overheat, and this could cause a fire. A fuse in series with the varistor is essential so that it will blow before the varistor will burn. The varistor should be positioned away from flammable materials and/or enclosed inside a metal housing. You would need to check the specs for your varistor's maximum current rating, and choose a fuse of no greater than that value. The old electronics word for such a circuit was a "crow bar", as when they trip, it is like placing a conductor such as a crow bar across the circuit. If the varistor max current is exceeded, the fuse simply blows. But then the over voltage protection is gone. If anyone has placed varistors in their electrical system, it is extremely important to insure that each is properly fused and well isolated for fire protection, or you could have a catastrophic fire. A fuse holder, fuse, some wire, crimp connectors, terminal strip, and perhaps some sheet metal or small metal electrical box are a very small cost and effort compared to effects of a catastrophic fire. BEWARE! A varistor without fuse protection can be very dangerous!

Another problem is that you may have no way of knowing when and if the varistor has triggered. An over voltage alarm would be helpful. I am not familiar with what is available off the shelf for these, but it would be good safety to have one in the system. If it trips, you will know to check your varistor and its fuse.

Notice that the 14V on the Littlefuse V22ZA3P TVS specification sheet is 14V RMS alternating current with its 20V peak. The DC number is 18V as shown in the next column.

The V vs A curve for the device is shown in Figure 10 of the specification sheet. While the TVS has a significant capacitance, I was hoping that Goboatingnow would point me toward a MOSFET based device with a flatter V vs A curve.

I agree with your recommendation of a fuse. Otherwise the TVS or its leads are the fuse. Sizing the fuse would be a challenge remembering that the device is designed to handle 1000A for 20 µs (and that is its reason for being), but a quick blow fuse sized to protect the leads might be a good start.

[PaulCrawhorn]

Originally Posted by RaymondR

We are talking about Lithium Iron Phosphate sometimes acronymed as LFP and sometimes as LiPo. Is that correct?

No, LFP is just a simplified version of LiFePO4.

LiPo is never used to refer to that specific chemistry. It is not even a chemistry indicator, just a lightweight and very vulnerable packaging technology, common in the RC world.

All of the various more energy-dense chemistries used with LiPo (e.g. LCO, LMO, maybe NMC, LMC these days) are nominal 3.6-3.7V, max 4.2V and very fire prone, not safe enough for this use case.

[PaulCrawhorn]

Also LiPo lifespan is like 2-5% that of LFP.

[Dieseldude]

Originally Posted by wsmurdoch

Notice that the 14V on the Littlefuse V22ZA3P TVS specification sheet is 14V RMS alternating current with its 20V peak. The DC number is 18V as shown in the next column.

The V vs A curve for the device is shown in Figure 10 of the specification sheet. While the TVS has a significant capacitance, I was hoping that Goboatingnow would point me toward a MOSFET based device with a flatter V vs A curve.

I agree with your recommendation of a fuse. Otherwise the TVS or its leads are the fuse. Sizing the fuse would be a challenge remembering that the device is designed to handle 1000A for 20 µs (and that is its reason for being), but a quick blow fuse sized to protect the leads might be a good start.

I had not looked at the current spec. Th problem is that if an over voltage condition lasts much longer than 20 micro seconds, the thing will get hot and might burn. A quick blow fuse might blow too soon. Since the concern is about a sustained over voltage, I would be inclined to use a slow blow fuse so that it will not blow during a quick voltage spike, but it will blow for a sustained over voltage.

[tanglewood]

When the alternator output gets disconnected, doesn't it create a reverse EMF? I.e. it's a reverse voltage spike? No?

[team karst]

Originally Posted by tanglewood

When the alternator output gets disconnected, doesn't it create a reverse EMF? I.e. it's a reverse voltage spike? No?

I think of it this way:

Alternator is chugging along, delivering 50 Amps. Suddenly, the output is disconnected. Since the regulator has a time constant, plus the field current can’t collapse instantly, plus stator windings, being coils, therefore inductive, they hold energy, there becomes a high positive voltage at the B+ port to keep the current flowing for a time. This transient energy may manifest itself in switch arcs, and other undesirable voltage excursions.