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

[klaus53123]

HelloSailor,

thanks for your quick answer. What I tried was measure the cell Voltage with my Mulimeter. Couldn't measure any significant voltage drop by measuring the cell Voltage directly on the poles (+/-) of Cell 3 under a 120A load. I guess I should see any reducing voltages there if they occur. Or am I wrong?

Thanks for your help
Klaus

[hellosailor]

" I guess I should see any reducing voltages there if they occur."

Your right. With your voltmeter across cell 3, you should see any changes in the voltage it supplies under load, even as the load changes. I'm sure there is some effect caused by the meter itself acting as a tiny parallel load, but doubt it is significant in this picture.


And of course, you could repeat the measurements on the other cells, to see how a "healthy" one behaves in the same conditions.


If Cell3 crumps out and the others don't...Cell3 has an internal problem.

[evm1024]

Originally Posted by klaus53123

HelloSailor,

thanks for your quick answer. What I tried was measure the cell Voltage with my Mulimeter. Couldn't measure any significant voltage drop by measuring the cell Voltage directly on the poles (+/-) of Cell 3 under a 120A load. I guess I should see any reducing voltages there if they occur. Or am I wrong?

Thanks for your help
Klaus

Let's see if I understand what this means. My little diagram is the connections to cell #3.

LiFePO4 represents the internals of cell #3.
C represents the actual connections (poles) on cell #3
A represents the hardware that connects to cell #3

A---C---LiFePO4---C---A

So if I understand you correctly under a 120 A load the voltage measured between C's shows no drop (is the same as the other cells) and if measured between A's shows a voltage drop (2.7 v).

If that is true then you have a resistive connection between A and C on one side or the other (or both).

Regards

[sv.antea]

About the BMV-702. Thank you Hoopla, I got it all mixed. In my Googling I have read so many times about the relay, what it can do and how to program it in differnent manuals and pamflets, but no mention of the connections or how much current it can handle either. I am going to install the BM with my TPPL AGMs for the rest of their life (so I'll know the time of their demise), then do a full conversion to LFP.

[senormechanico]

Originally Posted by hellosailor

I'm sure there is some effect caused by the meter itself acting as a tiny parallel load, but doubt it is significant in this picture.

Meters these days have an 11 megohm input impedance (11,000,000 ohms).
There won't be any measurable effect whatsoever.

[hellosailor]

What evm said.
Another way to double-check things would be to physically swap #1 with a different cell (preferably #1, guessing they have nothing in common in terms of hardware connections) and see if the cell, or the position (#3) in the setup has the problem.
If it is just a connection problem, the handling it gets during a cell swap might also clean/secure it enough to get rid of the problem.

[klaus53123]

Looks like I have some kind of connection problem or BMS problem. Reconnected the wiring from the cells to the BMS and didn't have the same failure again, but a new one..... Endless story....

To explain again:

If I measure between plus and minus of cell 3 I do not see any significant voltage drop on that cell. If I measure the pack voltage I also see no significant voltage drop under load. If I measure between C3+ and pack minus (C1-) I also see no significant voltage drop. All on the multimeter. But the BMS sees a voltage drop so failure is either the connection to the BMS or inside the BMS.... At least this is my interpretation....

Any other views?

[hellosailor]

So if you measure the cells (two of them) and you swap the cells in the bank, and measure the cells, you don't measure anything wrong or significantly different about them.
And the BMS doesn't care which cell is attached as "C3", it will call either one of them bad?
That would point to a bad BMS or bad wiring to the BMS, yes. Very specifically.

[Hoopla]

Originally Posted by sv.antea

About the BMV-702. Thank you Hoopla, I got it all mixed. In my Googling I have read so many times about the relay, what it can do and how to program it in differnent manuals and pamflets, but no mention of the connections or how much current it can handle either. I am going to install the BM with my TPPL AGMs for the rest of their life (so I'll know the time of their demise), then do a full conversion to LFP.

Pleasure. The Victron BMV is a good piece of kit, noting the cheaper 600/700 series can also operate a relay. They are the equal of or better the old pioneering Heart Interface, (later purchased by Xantrex) monitors. The additional input on the shunt of the 702 series is to either measure battery voltage (of a second battery), temperature or midpoint voltage. I don’t think much of the second battery input as it doesn’t measure current. That is best done with a second or more BMV’s with the shunts connected in series on the negative bus.

I am a fan of some Victron gear (not their Phoenix charger range) as it is good value for money, well thought out and their quality control is back to where it was. I recently said goodbye to a twenty five year old wound mod-sine wave inverter that apart from weighing a ton never missed a beat and was bullet proof.

They have a thing called Color Control which I think very useful for monitoring and controlling LFP installations that incorporate current day Victron gear. It is an inexpensive, small (120x130 mm) screen Linux based computer that collects data from all Victron equipment and shows it on the display. Besides communicating to Victron equipment, the Color Control communicates through NMEA2000, Ethernet and USB. For example a USB GPS can even be connected to the USB port!!

Connected up to your Victron BMS, BMV’s, solar regulator, inverter, charger etc you can monitor, control and log input and outputs (AC and DC) from each device and even do that remotely via the internet. Very cool and a must have for early adoptees of technology.

[Hoopla]

Corrrection to 2nd Para of my post above. Should read Centaur not Phoenix: "I am a fan of some Victron gear (not their Centaur charger range).......

[senormechanico]

BREAKING NEWS: NO FLIGHT UNTIL 2016

Despite the hard work of the Solar Impulse team to repair the batteries which overheated in the record breaking oceanic flight from Nagoya to Hawaii, the solar powered airplane of Bertrand Piccard and André Borschberg will stay in Hawaii until early spring 2016.

Following the longest and most difficult leg of the round-the-world journey which lasted 5 days and 5 nights (117 hours and 52 minutes), Solar Impulse will undergo maintenance repairs on the batteries due to damages brought about by overheating.

During the first ascent on day one of the flight from Nagoya to Hawaii, the battery temperature increased due to a high climb rate and an over insulation of the gondolas. And while the Mission Team was monitoring this very closely during the flight, there was no way to decrease the temperature for the remaining duration as each daily cycle requires an ascent to 28’000 feet and descent for optimal energy management.

Overall the airplane performed very well during the flight. The damage to the batteries is not a technical failure or a weakness in the technology but rather an evaluation error in terms of the profile of the mission and the cooling design specifications of the batteries. The temperature of the batteries in a quick ascent / descent in tropical climates was not properly anticipated.

Irreversible damage to certain parts of the batteries will require repairs which will take several months. In parallel, the Solar Impulse engineering team will be studying various options for better cooling and heating processes for very long flights.

The University of Hawaii with the support of the Department of Transportation will host the airplane in its hangar at Kalaeloa airport. Post-maintenance check flights will start in 2016 to test the new battery heating and cooling systems. The Round-The-World mission will resume early April from Hawaii to the USA West Coast. From there Solar Impulse will cross the USA to JFK in New York before making the Atlantic crossing to Europe and then returning the point of departure in Abu Dhabi.

Solar Impulse is attempting a historic first of flying around the world only on solar energy. And while Solar Impulse has completed 8 legs, covering nearly half of the journey, setbacks are part of the challenges of a project which is pushing technological boundaries to the limits. Solar Impulse will try to complete the first ever round-the-world solar flight in 2016 and this delay will in no way influence the overall objectives of this pioneering endeavour.



Solar Impulse RTW - 8th Leg from Nagoya to Hawaii

[Hoopla]

I think Solar Impulse is promoting "lithium-polymer" style cells which if I understand it correctly are developed to avoid the voltage sensitivity/lifespan issues of mainstream lithium chemistry, the core to discussions in this thread. Other than that "lithium-polymer" doesn't appear to have any other advantages except the opportunity for very small form factors.

[hellosailor]

Bear in mind that Solar Impulse, with a megabuck support and development teams, was also GROUNDED for extensive battery damage. From overheating, I believe.


So as a demonstration platform, it has shown marvelous potential--but nowhere near the reliability and duration needed for commodity users.

[senormechanico]

I only posted that for general interest.
Their battery is obviously not what we're using.
I just thought it would be worth reading.

[Hoopla]

I found the battery temperature issues attached to the planes ascent / descent in tropical climates as something which was not properly anticipated a bit weird for such a high profile exercise. I hope I haven't forgotten something in my planning.