LiFePO4 Batteries - Okay Tear Me Apart ;-)

[senormechanico]

Point taken.
I just am a little concerned with the world's idea that anything bad happening to anyone is someone else's fault.

[s/v Jedi]

Yup, I checked and you can interrupt the field without any problems regardless of output power. This is the most important charge control to work with because most alternators used for charging house batteries have an external regulator. You don't need to have the solenoid in that case "to be sure" and none of the smart alternator regulators do that... it's just a waste and a risk.

I just wouldn't support alternators with internal regulator. Disconnecting their output is not a supported control mechanism. You can't do anything with a capacitor because you have no access to the field terminal nor control the field current. It is incompatible. They can be modified for use with an external regulator though and this will be easier than using a magnetic clutch.

So it's real easy; a relay in the F wire is enough.

cheers,
Nick.

[S&S]

Originally Posted by s/v Jedi

Yup, I checked and you can interrupt the field without any problems regardless of output power. This is the most important charge control to work with because most alternators used for charging house batteries have an external regulator. You don't need to have the solenoid in that case "to be sure" and none of the smart alternator regulators do that... it's just a waste and a risk.

I just wouldn't support alternators with internal regulator. Disconnecting their output is not a supported control mechanism. You can't do anything with a capacitor because you have no access to the field terminal nor control the field current. It is incompatible. They can be modified for use with an external regulator though and this will be easier than using a magnetic clutch.

So it's real easy; a relay in the F wire is enough.

cheers,
Nick.

Even with brushless? If you have a source, I'd like to see it.

[s/v Jedi]

Originally Posted by S&S

More from the battery side. I'm not sure that these will tolerate a voltage spike. Also, I'm running a brushless alternator so I have diodes rectifying the field current that I have to be concerned with.

Hi S&S,

I am quite sure that there will not be a voltage spike when you interrupt the field current. The only thing that happens is that the magnetic field will collapse over a short time period and the alternator output will follow that curve down to zero without ever going up first. Even if there would be something; the batteries themselves are the best capacitor to dampen that because there is no power behind it. It is the power that would damage anything, not just a high voltage (a high voltage collapses as soon as it is shorted by the very low internal resistance of the batteries).

With a brushless alternator there isn't a problem either. The rectifier diodes are for the field current feeding the primary/power alternator. You never disconnect anything there. The field terminals you work with are connected to the exciter alternator (brushless means there are two alternators in one housing and the exciter is just used to generate the power for the field of the output alternator). So you never disconnect an output.

I understand the confusion because all the smart regulators advertise their slow start/stop sequence. But you must realize that this is to save the diesel engine driving the alternator (give it's governor time to adjust to power take off), not for the alternator itself. Not a problem for a propulsion engine because it has much more hp than required to spin the alternator.

p.s.: still hoping for a BMS explanation!

ciao!
Nick.

[S&S]

Originally Posted by s/v Jedi

Hi S&S,

I am quite sure that there will not be a voltage spike when you interrupt the field current. The only thing that happens is that the magnetic field will collapse over a short time period and the alternator output will follow that curve down to zero without ever going up first. Even if there would be something; the batteries themselves are the best capacitor to dampen that because there is no power behind it. It is the power that would damage anything, not just a high voltage (a high voltage collapses as soon as it is shorted by the very low internal resistance of the batteries).

With a brushless alternator there isn't a problem either. The rectifier diodes are for the field current feeding the primary/power alternator. You never disconnect anything there. The field terminals you work with are connected to the exciter alternator (brushless means there are two alternators in one housing and the exciter is just used to generate the power for the field of the output alternator). So you never disconnect an output.

I understand the confusion because all the smart regulators advertise their slow start/stop sequence. But you must realize that this is to save the diesel engine driving the alternator (give it's governor time to adjust to power take off), not for the alternator itself. Not a problem for a propulsion engine because it has much more hp than required to spin the alternator.

p.s.: still hoping for a BMS explanation!

ciao!
Nick.

Thanks for the explanation Nick! Running a thermocouple for temp control should be equally straightforward.

A BMS does two things:
It provides a resistive bleed to help balance the cells and,
2, it provides interrupts to prevent overdischarging or overcharging based on bank overvoltage, bank undervoltage or cell voltages out of balance.

[s/v Jedi]

Originally Posted by S&S

Even with brushless? If you have a source, I'd like to see it.

Here's where I got my info on the brushless alternator: Alternator - Wikipedia, the free encyclopedia
I know Gord doesn't like us to quote Wikipedia but in this case I think he'll let it pass ;-)

A brushless alternator is a mechanical/electrical piece of beauty... There are two alternators, a small one and a big one and they are mounted on a single shaft. The small one is the exciter. It's armature is it's output winding while the field coil is stationary. The trick is that the output isn't taken off the armature like with a regular alternator. It is fed to a rectifier that is mounted on that shaft (it turns with the shaft) and the output is connected to the armature winding of the second (big) alternator. That one is build in reverse, with the field winding on the armature and it's output windings stationary. No brushes!

Now, to control that beauty you only need to control the field winding on the exciter (small alternator) part of it. The rectifier on the output is just like the rectifier on every other alternator from an electrical point of view... the only difference is that it spins with the shaft.
So, there's no more sources needed to be sure: you can control it like if it's a normal alternator with brushes. Oh wait... just like with my explanation for the regular alternator: just imagine that there would be a spike on the output of the exciter part... that one is always connected to the field winding of the power alternator which is also a low resistance for DC. This will dampen any spike before it can build just like a battery. The only way to create a big diode-burning spike is to disconnect an alternator output that is under high load. As long as you have the output connected to something with low resistance like batteries or other windings, nothing bad is gonna happen.

Besides those battery switches that are operated with the alternator working (disconnecting the output) there is one other thing that can ruin the diodes and that is a fuse that blows. If that happens during bulk or early absorption phases you can be sure the diodes are shot.

cheers,
Nick.

[s/v Jedi]

Originally Posted by S&S

Thanks for the explanation Nick! Running a thermocouple for temp control should be equally straightforward.

A BMS does two things:
It provides a resistive bleed to help balance the cells and,
2, it provides interrupts to prevent overdischarging or overcharging based on bank overvoltage, bank undervoltage or cell voltages out of balance.

Okay, I understand no.2 but what is a "resistive bleed"? I have the feeling that it is something like a resistance switched parallel to the cell when it's fully charged but the other cells aren't full yet? But that would also discharge that cell?

cheers,
Nick.

[S&S]

Cool. I'm okay with Wikipedia (as long as it conforms with what I already know)

[S&S]

Originally Posted by s/v Jedi

Okay, I understand no.2 but what is a "resistive bleed"? I have the feeling that it is something like a resistance switched parallel to the cell when it's fully charged but the other cells aren't full yet? But that would also discharge that cell?

cheers,
Nick.

If an individual cell is running at an overvoltage, the cell is shorted through a resistor (R)to drain off the charge (+to- of that cell) As R is high, I is small.

So in short , yes, it discharges that cell (a little).

[jallum]

Okay, then... so the only real annoyance with the MiniBMS at this point is that it doesn't distinguish between over-charge and over-discharge... so, at LVC, it'd open the relay and you wouldn't be able to charge your batteries until it has been reset. Kind of a pain in the ass if you manage to get yourself into this situation, but more than doable. Better would be to differentiate the two states and provide a relay/contactor control for each like the Genasun.

The only other deficiency I can see is (and this is by design, as it's supposed to be simple) that it doesn't give you any kind of info on the current state of the pack.

For my rig, I just sprung for the Elithion BMS... not the cheapest option at ~US$700 for a 2x8 pack and controller, but I like the fact that it can be setup to automatically dump easy-to-parse pack state data out it's serial port once a second and that it tracks SOC, DOD, temps, general pack health (etc, etc.) I cut my teeth doing industrial machine control software and it's going to be fun seeing this data on my iPad come end of April... for now, the simulator will have to do.

[hellosailor]

"their slow start/stop sequence." That's also designed to prevent belts from snapping. If you've ever watched a loose or imperfectly aligned v-belt when you kick in a heavy load (i.e. switch in a flat battery for charging) you can literally see the belt jump, hunched up like an inchworm. And jump right off the pulley in some cases. Or snap.

Gradual loading is just kinder to everything that is attached.

But getting back to new battery chemistry...I've been reading a lot about the many new electric cars due to come on the market, and AFAIK none of them is using this new battery chemistry. Maybe all those big investors have missed the boat. Or, maybe in four five years we'll find out this is yet another big promise from China that doesn't quite pan out. (At least one major company deploying electric cars IS Chinese, and even Warren Buffet has invested in it, but they're also using conventional chemistry.)

[imcbride]

Originally Posted by hellosailor

But getting back to new battery chemistry...I've been reading a lot about the many new electric cars due to come on the market, and AFAIK none of them is using this new battery chemistry.

GM is investing in LFP (most likely using it for the Volt). The USGOV just pumped a bunch of money into A123, a US-based LFP battery producer. Tesla motors is using Li-ion. Their website says "We constantly review new technologies for our vehicles. Right now, Li-ion offers the best performance and value."

I haven't heard any downsides to LFP, through the whole lifecycle. There's even an outfit in the US with funding to recycle them.

I hope I won't be eating my hat any time soon, but I think LFP is here to stay.

[jallum]

LFP is also the safest of the Lithium-Ion chemistries, so far. LFP isn't prone to thermal runaway like LCP (Lithium-Cobalt) or LMP (Lithium-Manganese) chemistries, where overcharging can lead to complete meltdown and ignition -- no fun. It's also not exactly bleeding-edge tech... the chemistry was discovered in 1996.

[electric1]

p.s.: still hoping for a BMS explanation!

ciao!
Nick.

Nick,

BMS has 2 critical functions and additional non-critical / optional functions. First critical function is HVC ( High Voltage Cutoff ) which stops the charging process when any cell in the series connected string reaches predetermined HVC voltage level. HVC level selected by BMS designer is usually still lower than actual max voltage for LFP chemistry, which is done on purpose to prolong battery life. LFP batteries don't like living on the edges, they like living on the flat part of the curve, between 10% and 90% SOC.

Second critical function is LVC ( Low Voltage Cutoff ) which stops the load when battery reaches predetermined LVC level.

Common non-critical option is shunting/balancing , which helps to keep cells in the string balanced. Amount of shunting/balancing is determined by how much current BMS can shunt, which determines how much energy is dissipated into heat by bleeding resistors. Early versions of BMS had lots of shunting capacity which creates safety hazards when battery is in enclosed environment and can't dissipate heat fast enough. MiniBMS was carefully designed to avoid such hazard. Each module has 5 Watt bleeding resistor, but max current is limited such that only 3 Watts are dissipated, so resistor does not get hot at all. Also, each module has temperature controlled PTC fuse, which interrupts the shunting if resistor gets to 80F. In my extensive testing this has never happened, not even close.
In reality, LFP cells tend to stay in balance if they are all located in the same temperature location and not subjected to extreme currents for long time. This is because imbalance is created by varied IR ( Internal Resistance ) of each cell, which is a function of the temperature, load and SOC. As long as these 3 factors are within norm the pack will stay balanced. A small amount of shunting that MiniBMS provides is an insurance policy that pack will stay balanced long term.

Other optional BMS functions are related to data gathering, like Jallum mentioned above, to collect digital historical data and analyse it on a computer. However, these functions usually dramatically increase the cost of BMS system and require more skills to install/configure/use. Over past 2 years in EV world we discovered that many people don't want the complexity and cost of the digital BMS, so there was a strong demand for simple analog BMS with only critical functions, which prompted design of MiniBMS.

Just like anything else in life, there are different products meeting different demands of different people. Some want a Cadillac and some want a Honda. So consider Elithion BMS as a Cadillac and MiniBMS as a Honda.

I believe most of your sailboats have Lead Acid management systems, like Xantrex Link 2000 for example. These systems provide SOC reading and other management functions. So my idea of drop-in LFP replacement with simple BMS assumes that you would continue to use all your existing systems and treat LFP battery the same way as you treat LA battery, with added benefits that LFP provides, such as long life and larger useful capacity at lower weight.

As for sizing LFP cells, its not much different than sizing LA cells. You figure your typical load in Amps and typical time you want to sustain the load between charges and multiply the two to get required AH capacity. Then you derate the capacity to stay within safe SOS and DOD margins. The difference here is that LFP has wider margins and virtually no Puekert effect, so you get more useful capacity from LFP cells.

Where you often derate LA pack to 50% for long life, you would derate LFP pack to 80%, so you get extra 30% of the LFP pack with the same rated size.

Hope this helps.

[electric1]

But getting back to new battery chemistry...I've been reading a lot about the many new electric cars due to come on the market, and AFAIK none of them is using this new battery chemistry.

Unfortunately LFP chemistry does not have the best energy density by either weight or volume amond other Lithium chemistries. However, LFP is the safest and requires minimum management. Its also less expensive since it uses cheap Iron and has very little Lithium in it. These factors make it very attractive for DIY market and stationary power markets. From this perspective LFP is perfect for house banks on the boat, where size and weight is not as critical, but safety and longevity is important.

Lithium packs in production EVs have very complex BMS systems, which control battery cooling ( often liquid cooling like Tesla ) and complex algorithms determining available range. There are many "moving parts" in those packs, which drives their cost up even more. All this is done in cars to use maximum energy density, which is still 10 times less than gasoline, but its getting better and better.