LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[diugo]

The Junsi CellLog only has a single alarm port, so it can only drive a single relay. But, the same relay can simultaneously drive a shutdown solenoid and sound a klaxon.

One solution, outlined earlier in this thread, is to employ two CellLogs in "parallel"---one to handle the LVD, and the other the HVD.

You can insert a one-shot timer circuit between the LVD relay and its shutdown solenoid to give you the time delay you want.

[witzgall]

Hi all,

I finally got around to wiring up a Cellog8s to my pack. I can get it to recongnize a single cell, but when I wired it to the pack, I get a "Err Volt" after it starts up. Of course the manual says nothing about this.

I have it wires so that the bottom wire is connected to the last negative post, then the next four connected to each group of positive posts (I have a 2P4S configuration). any ideas on what I am doing wrong? I extended the wires, but the extensions work for one cell, you would think it would then work for more???

Chris

[NahanniV]

Originally Posted by diugo

The Junsi CellLog only has a single alarm port, so it can only drive a single relay. But, the same relay can simultaneously drive a shutdown solenoid and sound a klaxon.

One solution, outlined earlier in this thread, is to employ two CellLogs in "parallel"---one to handle the LVD, and the other the HVD.

You can insert a one-shot timer circuit between the LVD relay and its shutdown solenoid to give you the time delay you want.

The CellLog alarm port can only sink 500mA, so it will need some additional components to drive most relays suitable for a boat.

I think it would be easy to split the alarm signal from the CellLog into HVE and LVE based on the total pack voltage: If the voltage is above 13V signal is HVE Below 13V signal is LVE.

I think this is what the HousePowerBMS board is doing. There is only an alarm signal from the cell boards and the HousePowerBMS board must decide if it is HVE or LVE; The MiniBMS Board is simpler in that the ignition signal determines if the alarm from the cell boards is HVE or LVE: If ignition is on expect LVE(driving), Ignition off Expect HVE (charging).

[diugo]

Originally Posted by NahanniV

The CellLog alarm port can only sink 500mA, so it will need some additional components to drive most relays suitable for a boat.

I think it would be easy to split the alarm signal from the CellLog into HVE and LVE based on the total pack voltage: If the voltage is above 13V signal is HVE Below 13V signal is LVE.

The relay is only driving a solenoid---it certainly does not need to draw in excess of half an ampere. Heck, the right transistor or optoisolator will do the job for less than 1mA.

And if a circuit needs to make a decision based on whether pack voltage is above or below 13V, what do you even need the CellLog for? Simply put eight such comparators on a board, and it can handle the LVD and HVD directly.

Another option is a $25-50 programmable microcontroller board with analog inputs and digital outputs, such as an Arduino / Maple / Olimex. The sky is the limit with something like that.

[diugo]

On a quite amusing note---and a largely nautical one at that---how does my young niece pronounce "LiFePO4"?

Life of Po!

[mcarling]

I have a theory about a possible reason for the Balqon clearance sale.

The Winston 60AH cell has a specification weight of 2.3kg +/- 50g.
The old Sinopoly 60AH cell has a specification weight of 1.99kg +/- 100g.
The new Sinopoly 60AH cell has a specification weight of 1.83kg +/- 100g.

Perhaps Winston might soon offer a revised line of lighter weight cells.

[ebaugh]

Originally Posted by diugo

The relay is only driving a solenoid---it certainly does not need to draw in excess of half an ampere. Heck, the right transistor or optoisolator will do the job for less than 1mA.

And if a circuit needs to make a decision based on whether pack voltage is above or below 13V, what do you even need the CellLog for? Simply put eight such comparators on a board, and it can handle the LVD and HVD directly.

Another option is a $25-50 programmable microcontroller board with analog inputs and digital outputs, such as an Arduino / Maple / Olimex. The sky is the limit with something like that.

It takes about 3A surge to close a Tyco EV200, and 170ma to hold it closed if I remember right....

The analog inputs on all the micro controllers I've looked at are referenced to the same ground, it makes it difficult to get the accuracy needed on the voltage readings since you have to use a 16V scale on a 5V input loosing precision. To do it right, you have to add something like an LTC6803 on the front end or at least a few independent ADCs somehow. It's possible, but not simple.

[Dockhead]

Here's a car burnt up by its LiFePo battery bank:

Battery Vehicle Society • View topic - Lifepo4 cells catch fire!

Yikes! Hope these bugs are ironed out soon.

[diugo]

Originally Posted by ebaugh

It takes about 3A surge to close a Tyco EV200, and 170ma to hold it closed if I remember right....

It's funny, how at the same time we are finally embracing exciting new battery technology, some still steadfastly rely on that tired old contactor.

Re voltage accuracy of inexpensive microcontrollers, there are simple software techniques to get better than 10mV reliably---which really is good enough for disconnect purposes, don't you think?

[diugo]

Originally Posted by Dockhead

Here's a car burnt up by its LiFePo battery bank:

Battery Vehicle Society • View topic - Lifepo4 cells catch fire!

Yikes! Hope these bugs are ironed out soon.

This incident was nearly four years ago. Reading the whole thread, it appears the vehicle used the same Hi-Power cells that Lloyd showed---the ones that used the aluminum clamp on the copper foil inside, that completely dissolved due to galvanic corrosion.

This particular bug was quickly ironed out a while ago.

[ebaugh]

Originally Posted by diugo

It's funny, how at the same time we are finally embracing exciting new battery technology, some still steadfastly rely on that tired old contactor.

Re voltage accuracy of inexpensive microcontrollers, there are simple software techniques to get better than 10mV reliably---which really is good enough for disconnect purposes, don't you think?

10 mV is good enough. What technique can improve precision?

[diugo]

Originally Posted by ebaugh

10 mV is good enough. What technique can improve precision?

The most common software technique is simply take the same analog measurement several times in quick succession, toss the high and the low, then average the rest.

[Dockhead]

Originally Posted by diugo

This incident was nearly four years ago. Reading the whole thread, it appears the vehicle used the same Hi-Power cells that Lloyd showed---the ones that used the aluminum clamp on the copper foil inside, that completely dissolved due to galvanic corrosion.

This particular bug was quickly ironed out a while ago.

That's good to know!!! Wow, that would suck if it happened on board a boat!!

[mcarling]

Originally Posted by diugo

This incident was nearly four years ago. Reading the whole thread, it appears the vehicle used the same Hi-Power cells that Lloyd showed---the ones that used the aluminum clamp on the copper foil inside, that completely dissolved due to galvanic corrosion.

This particular bug was quickly ironed out a while ago.

It also seems that it was left charging long after it was fully charged, unattended, at an ambient temperature of -5C (which is warned against), the cells were not banded against expansion, and that the battery bank was immediately adjacent to a petrol tank.

I trust that none of us would make all of those mistakes and hope that just one or two of them would not be enough to start a fire.

[hellosailor]

Bob-
Thanks! I had no idea the PO4 batteries were a different voltage, never heard that mentioned before.

Dockhead-
But your example of of PO4 battery fire, isn't. The guy had PO4 batteries, he had a fire, he's got no evidence that the batteries caused the fire, as opposed to his charger, or his use of the batteries in cold weather that was clearly beyond their rating. That's like saying my refrigerator caused my house to burn down, because it was found in the remains.
I can see that a really vigilant battery make (or their distributor) might want to perform an autopsy on the remains. Or not have anything to do with it, because that might boomerang. But the only way to be sure about "cause" would be some very careful work investigating and the OP of that thread didn't take it to that degree.

Memo, add a "low temperature crowbar shutdown" switch to your charging setup.<G>