Magic pie ll with liFeP04 48 volts1000 watts & charger HP8204L4

[Gilles1945]

Hi everybody I need help,
I have a Magic Pie ll with lifeP04 48 volts and 1000 watts with a Li-on battery charge model HP8204L4 with twist throttle that was bought in April-2011 never have a problem.

My problem is the battery is charge to 55.6 volts charger output 58 volts drive 8 klm everything drop boom. Check my battery voltage 52 volts wait 5 minutes restart goes 50 feets than shot down again check battery voltage 51 volts.
Can somebody help me on this please
Thanks Gilles

[Bikemad]

Hi Gilles,

If the battery has not been used for some time, it could be that the battery cells are unbalanced and some of them are not being fully charged, resulting in reduced battery capacity. Some regular charging may improve the capacity.

I could also be due to one or more of the battery cells being faulty and the voltage of one cell (or possibly more) is dropping under load and causing the battery's BMS to cut the power to prevent the lowest voltage cell from being overdischarged.

You would need to dismantle the battery and check the individual cell voltages and then put the battery under load in order to expose any weak or faulty cells.

Alan
 

[Gilles1945]

Hey thank you for your quick reply.
I have check with a digital meter all 16 batteries. one is at 243 volts 2 other are 329 volts all the other batteries are at 3.33 volts
Do I have to charge each battery individually if so what charger o be use for one battery and what charger do I need to be on each battery.

Thanks Gilles

[Bikemad]

The GM 48V10Ah LiFePO₄ battery contains 16 separate cells, and charging individual cells requires a special charger suitable for LiFePO₄ cells which must not exceed 3.65V per cell, so it would be much safer to allow the BMS to balance the cells itself.

If you have a CE marked charger, you simply charge it fully until the green LED comes on continuously and then leave it on charge (for several hours if necessary) to allow the cells to balance. The green LED will usually keep going on and off while the battery is balancing, and should eventually stay green continually to indicate that the battery is fully charged, and the cells should then be nicely balanced.

Unfortunately, the UL version (as required in North America) are not permitted to continue supplying current after the battery has been fully charged, and (in my opinion) if you have one of these UL chargers, it may be very difficult to fully balance the cells if they are too far out of balance, as I am pretty sure that the GM batteries' BMS will only balance the cells while the charger is still supplying current.

Most BMS utilise top end balancing, which simply bleeds off excess current from the fully charged cells to prevent their maximum safe voltage from being exceeded.
This allows the lower voltage cells to continue charging at the same rate as the bleed current until the cells are all nicely balanced.

As bleed balancing can theoretically only take place while the battery charger is still supplying power, I think it is very unlikely that any noticeable balancing will occur after the charger has been switched off (i.e. after the UL version of the charger has shut itself down).

In my opinion, the more often you charge a battery, the better balanced the cells will be, but if a battery is kept in a fully charged state, its life expectancy will be less when compared to an identical battery that is always kept at a lower state of charge.

If you use a battery for 3 or 4 short trips before recharging it, it will only have one chance to balance, but if you charge it before each of the trips, it is 3 or 4 times more likely to be properly balanced.

I now try to fully charge my batteries before every trip for better (and quicker) balancing, and by doing this (and then using them straight after the charge process) it ensures that they are not left in a fully charged state for long periods of time. 

Alan
 

[Bikemad]

I have check with a digital meter all 16 batteries. one is at 2.43 volts 2 other are 3.29 volts all the other batteries are at 3.33 volts

You obviously have one very low cell (2.43V) which is 0.86~0.90V lower than the rest (all other cells are within 0.04V of each other).
This could be due to a faulty cell, or it could be the result of poor balancing, higher self discharge on the suspect cell, or possibly uneven current draw from the BMS connections while the battery is unused.

Mark the suspect cell and then fully charge the entire battery as you normally would, and then check all the voltages again.
If the voltage of the suspect cell has quickly risen to around the same voltage as the other good cells after a full charge, I would say that the suspect cell is faulty and will need to be replaced.

However, if the suspect cell is still much lower than the rest (>0.50V difference) it may be due to a slight imbalance that has progressively become worse over time, and it may be possible to revive it by using one of the following methods:

1) If you know anyone who is into R/C electric cars, boats or aircrafts etc. they may have a charger which is capable of charging a single LiFePO₄ cell to 3.65V (a LiPo only charger will go up to 4.2V per cell which is too high).

2) If you know someone with an adjustable bench top power supply that can be set to 3.65V maximum and ~2 Amps current, you could use that to charge the low cell on its own to gradually bring it back up to the same voltage as the other cells.

3) If you have a local model shop, it might be worth paying them a visit and asking if they could charge up the suspect cell (they should also be able to measure how many Ahs it takes at the same time).

4) You purchase a single cell LiFePO4 charger and cut the connector plug off and connect the wires directly to the cell terminals and then charge the suspect cell until it holds the same voltage as the other cells (even after the charger has been turned off for a minute or two).

If you are unable (or don't want to to wait) to try any of the above suggestions, there is one other (more involved) method that you could try:

5) You would have to carefully remove the copper connecting bars from the suspect cell to completely isolate it from the two adjacent cells (or just one adjacent cell and the output wire if the suspect cell in an end one).

Be very careful not to drop an metal nuts or washers, or touch any of the adjacent terminals with a metal spanner or short circuit anything with the copper connecting plates etc.
 
Then make up a couple of leads (ideally one red and one black) that could be attached to the suspect cell.
Connect a 12V 60W car headlamp bulb in series with one of the wires and then connect the ends of the two wires to the respective terminals on each of the good cells in turn for approx 5 minutes on each cell (making sure that you connect the correct polarity on the alternating cell terminals).
The process would be much easier with some suitable crocodile clips that would clip directly onto the cell terminals.

I suggest that you mark all of the positive terminals with a red marker pen to avoid any confusion before you start. 

As the maximum transfer current will only be around 0.25A, it could take many hours, but you should eventually be able to raise the voltage of the low cell until it is pretty much the same as the others (which will all have dropped by a small amount by then).

The aim is to get all of the cells at the same voltage and then refit the copper links etc. and charge the battery again until it is fully charged.

At this point, hopefully, all of the cells will still be at a very similar voltage level, and your available battery capacity should have increased dramatically.

Unfortunately, you won't know for sure until you actually test it afterwards.   

Alan
 

[Gilles1945]

Thanks Alan for your reply I will do the testing during the weekend and will make the result.

Your time is really appreciated.
Gilles

[Gilles1945]

Another question Alan if you don't mind how do I know if the cell is 10 Ah or 12Ah or 16Ah to be shore??
And where can I order a single cell for 48 volts LiFeP04 replacement battery???

[Bikemad]

The cell capacity is often marked on the side of the cells, so you may be able to see it on one of the end cells:



The 12Ah cells are 18mm thick, whereas the 10Ah cells are only 16mm thick.
I don't know for sure, but I think that the 12Ah cells were only used in the 36V packs.

The 48V 15Ah battery packs are finished in blue heat shrink tubing:



But the 48V 10Ah, 36V 12Ah and 24V 20Ah packs are all housed within a rigid aluminium extrusion like this:



Gary (GM Canada) has both the 10Ah LiFePO₄ cells and the 12Ah LiFePO₄ cells on his website, but it doesn't say how many are in stock.
I think that your battery will have the 10Ah cells.

Alan
 

[Gilles1945]

Thanks again Bikemad,

Here is my result check voltage.

3.35-3.35-3.35-3.30-3.50-3.35-3.23-3.48-3.35-3.35-3.32-3.33-3.49-3.43-3.27-3.34-,

Is that considered normal those difference,  every time I re-charge I get different number

Now in what way can I check the Amps value

[Bikemad]

The cells appear to be much better balanced than they were before, as the maximum variation between the highest and lowest cell voltages is now only 0.27V, whereas before it was 0.90V!

If my calculations are correct, the low cell has increased from 2.43 to 3.25, and this 0.8V increase accounts for more than half of the 1.49V that the total battery voltage has increased by.

If the original problem was caused by the cells being out of balance (rather than a faulty or weak cell) it should improve even more with regular charging.

Charge it fully until the charger fan cuts out and the green LED comes on, whilst monitoring the voltage at the output socket.
Make a note of the battery voltage reading as soon as the charger cuts out.

Switch the charger off for about five minutes and make a note of the resting voltage before repeating the above process again as many times as is needed.

Hopefully you will find that each charging cycle will gradually increase the resting voltage as the lower voltage cells should eventually rise to match the higher cells in the pack.

If you check the resting voltage of the cells with the lowest and highest voltage readings, you will hopefully see the difference between the two decrease with each charge cycle.

When the difference stops decreasing, your cells will probably be as well balanced as the BMS can manage to get them.

If you now make a note of the total battery voltage after it has rested for at least five minutes after the charger cut out. It can probably be used in the future to give you a rough indication of how well balanced the cells are without having to dismantle the battery to physically measure them.

If the resting voltage becomes a lot lower than the noted resting voltage, several full charge cycles (with minimal discharge) may be required to bring the resting voltage back up again.

Reassemble the pack and test it out on your bike to see if the cutting out problem has been cured.
Hopefully, having more evenly balanced cells will make a big difference to your range, but if it still cuts out after riding for a relatively short distance, I think the suspect cell will need to be replaced.

My GM 48V 10Ah LiFePO₄ pack came with the CE charger which has an output of 58.4V and my pack has a resting voltage of ~ 57.4V 5 minutes after a full charge. If I switch my charger off for 5 minutes at the end of a full charge, it only takes ~4 seconds for the charger fan to cut out again and the green light come on after I switch the charger on again.

If I leave my charger switched on and connected to the battery at the end of the charge, the battery voltage will remain at a steady 58.36V which allows the BMS to continue to balance the cells if necessary.

The easiest way to measure the maximum current under load and the actual battery capacity (Ahs) is to wire a simple watt meter into your battery supply lead:



This can be used to measure and/or record the following while the battery is still powered on:
Actual battery voltage, Minimum battery voltage, Actual current (Amps), Maximum current (Amps), Actual power consumption (Watts), Maximum power consumption (Watts), Consumed battery capacity (Ah) and Energy consumption (Wh).

Unfortunately, all data is lost when the battery is turned off at the keyswitch (or if the BMS cuts the battery power off completely).

Alan
 

[Gilles1945]

Me again Bikemad,

Back to the drawing board. I assemble everything went to take a ride and same problem again can't do more than 8 Km. At start my voltage was 56 volts and at shutdown my voltage is 52.5 volts
Now I was looking for buying an Power Analyzer and the one you referred to I will be receiving it in a month.
I look in different place to buy they all come from China and receiving again is in 1 month.  Do you know a place in Canada or in the state that they can ship closer and get it ind a week or so.  I do not want to loose my summer bicking.
I appreciate your help  Thanks Gilles

[Bikemad]

It sounds like it's a dodgy cell, as it only needs one weak cell to cause the BMS to cut out under load.

Run the battery until it cuts out and then dismantle it and recheck the cell voltages.

If the same suspect cell is more than 0.50V lower than all of the others, then it will hopefully be just the one cell that needs to be replaced.
The replacement cell should be at the same voltage as the rest of the cells before you install it, so it may be necessary to run the pack lower, or charge it up slightly until all the voltages are very similar.

If you want to perform a load test on the cells, a pair of 12v halogen headlamp bulbs wired in parallel could be used to place a load on a group of 4 cells at a time for testing the voltage drop under load.
If you connect the high and low beam terminals together on both bulbs, they should draw a combined current of about 17A across 4 cells in series.
The bulbs will get very hot while testing, so make sure you place them somewhere safe where they can't cause burn damage.

Make a note of all of the cell voltages before the load test and then measure the individual cell voltages (one at a time) while the load is connected across a group of four cells (1-4, 5-8, 9-12, or 13-16).
Move the load to the next group of four cells and measure those four cells individually.
Repeat until all 16 cells have been measured under load and then compare the voltage drop between the resting voltage and loaded voltage to see if any more cells are dropping significantly under load.

If you only have one cell that dropped excessively and the other 15 cells only dropped a small amount (and are still very similar) then you will only need to replace the single weak cell. 

I did find a watt meter in Canada, but it is almost 6 times the price of the one I previously suggested.

I would stick with the cheap one from China and concentrate on replacing the faulty cell (or cells) while you are waiting for it to arrive. 

Alan