Magic Pie II cuts out

[Sean]

Have a Magic Pie 2 with a 36V/16AH battery that is having problems. It
is about 18 months old.

It was garaged for the winter.  This spring I charged the battery. The
throttle indicators are all lit when the power is on. The power cut out as
soon as the throttle was turned under load. If key is turned on and off the power
comes back. Turn throttle & cuts off.

On the advice of my dealer, I disconnected the charger from the wall and from the battery. Let it sit for 20 minutes. Connected the battery to the charger and then plugged
the charger into the wall. Let it fully charge.

The battery charged to 41.7V  (read at battery output with key on) .

I could hold the bike and press throttle without cut out (3-4 times). 
It seemed like the problem was solved.

Had to go away for 5 days and when I returned to use the trike it cut
out first time I used the throttle. Voltage at battery was 41.5V w/ key on.

If I lift the back wheel off the ground I can engage the motor and it
will run. If I put it under any kind of load, it will cut off.

I decided to leave the unit on to drain the battery and re-charge. I am part way discharged after several days. Voltage is 39.5 V. I decided to try another test - I put the volt meter on the battery and measured the voltage under load. As soon as I engage the throttle the voltage starts dropping. It cuts out at some point below 27 V - its tough to catch it exactly because the multi-meter has a lag. The green and yellow throttle lights go off when the throttle is engaged.

Key on and off & it is back on.

For as long as I can remember the first two throttle lights have always gone off whenever I engage the motor, but I only noticed this behavior after the motor was already about 8 months old. Is this normal?

The dealer wants me to disassemble everything and ship it back - they think it is the controller. I think it is more likely the battery. Do you have any trouble shooting suggestions that I can try to determine if the problem is the battery or the controller?

Also, I have been interested for some time in converting to using a 48V
battery. What would I need to purchase to do make this conversion?

1. 48V battery
2. Do I need a new throttle?
3. Programming Cable - do I need to reprogram the controller?

If I order a 48V battery first would that be a good way to find out if
my problem is the battery or the hub/controller?

Finally, are there any settings for the hub motor that would allow me to
trade off between a 48V battery and a 36V battery if it turns out my 36V
battery is ok?

Thanks,

Sean

[Bikemad]

Hi Sean andto the forum.

From your very detailed post I would say that you definitely have a battery problem. If the voltage is dropping as low as 27V when the throttle is engaged you either have some faulty cells, a very poor power plug contact or faulty key switch contacts.


Regarding the 48V battery, there's no need to change the programming of the controller. The existing throttle will work, but the battery gauge will not read correctly and the throttle body will also tend to get a little bit warm.

It's possible to put a suitable resister in line with the power supply wire that feeds the voltage gauge to make it more functional on 48V. Check out this post for further details.

Alan
 

[dofbikes]

Hi Sean.

I've had identical problem and it was a nightmare to troubleshoot it.
I've had very bad news from GoldenMotor on faulty cells on my GM 48V12Ah battery.
I've bought it directly to them (August 2011, supposedly LFP but realy LMN) and they do not assure any kind of warranty, they simply answered that now they've better batts and the best they can do is offer a discount on a new purchase.
Consider to buy real batteries or SLA's. Take a look into suppliers of Giant, KTM, GRACE e-bikes and others and you'll rapidly find who they are and were can you buy good batteries.

Regards,

Rui

[Leslie]

STOP.  Do not discharge the pack any further.



Leave your charger on the battery for as long as you can.  Like a few days.

Find your battery chemistry.

If Lifpo4 read on.  If not let us know.  We will try to save your battery.

Leave your charger on the battery for as long as you can and see if this improves your situation, if so, get a single cell charger, disconnect the BMS and take each cell up to 3.9v-4v and leave for 5 mins.  Do not allow any cell to stay high at 4v, at least use a 5ohm 5 watt resistor to take the surface charge down to 3.65v, take it down a bit at a time, 800ma should be ok down the sense wire in longish energy burns.  Just take your time.

You can get a 4 volt single cell charger from hobby king.

It can take days to balance a very unbalanced battery otherwize. 

You may just be in time. 

Lithium self discharge at a rate of 5-10% per month some cells more and some less, they go out of balance.  So 5mths of storage can do half of your pack.

@ 5 Some cells will be discharged @ 50% and some at 25%, balancing would be a nightmare.  Even try to use a single cell charger and try to charge each cell indifvidually.  Any low cells, leave try take up to 3.9v-4v for 5-10 mins at a time.  Also there is parasitic currnet the BMS takes, some are very low, some not so low.  Over 7mths storage one would be risking pack failure.

You must un plug the battery to the pack during storage, not just turn off the key. And Like a car left in storage, it is good to run the bike for 5-10 mins and recharge the pack at least once a month.

[Leslie]

Question!

BTW when you leave the charger on, do you see it switch on and off say over an hour or two?. 


If so leave th pack on the charger until this behavior stops. If  your charger is not switching off and on after full charge then your charger and BMS co-operation may not be up to the task. You need to get a wall timer to swtich it on and off for you..

What should happen is when high cells go high and the full volatge reaches full charge, but some cells are low and some cells are higher than normal.  The charger should switch off long and down the current, enough for the BMS to burn the high cells down, thus initializing the charger to engage in full again.  If this on and off situation is not occuring, then it is believed the charger is keeping high cells high, never allowing another full charge in to get the low cell up.

Or your cells are bye bye's.

Make sure the 220v 110v selector is in the right position.

[Sean]

Thanks everyone for the good answers. I'll first try to save the battery.

The battery was stored completely disconnected from the bike with key off in a partially heated garage (50-60 degrees).  It was fully charged when last used in November & re-charged in mid to late March.  So it went about 4 months in storage.

The cells I have are supposed to be LiFePo4 but I've also seen in the forums that LiMn batteries were often substituted and that LiFePo4 batteries are supposed to have a supplemental label, and mine doesn't. I'll check at lunch and see and post back.

My charger does not have a 110/220 selector - it appears to be auto-sensing since it is rated for both input voltages.

When I checked the battery last night it was down to 36V. I'll stop the discharge as soon as I get home.

I've never seen the charger cycle on and off, but I've never checked on a regular basis. It's hard enough for me to distinguish between the green and amber on my charger - the colors aren't very distinct from each other.

What cycle time do you recommend (example: 12 h on, 12 h off).

Thanks,

Sean

[Leslie]

I hate telling people to do things, and them find out it helps little.  I do feel so bad for your problem,  I have one person have success with this fix and another only a small improvement.

Recharge the battery to full and leave stand for a while. 

Then set the timer to switch off after 30 mins and on after an hour rest.  If you notice it only stays on full charge for a small time when you switch it back on, this means we are not getting the high cells down enough extend the rest to 2 hours..

What you want to see after you rest the battery from the charge  is bulk charge going in for some time.  Even 15 minutes is awesome. If it goes on for 15 mins then switches off automatically  Then you can make it go on for 20 mins and off for an hour..


Let me explain why this can work.  If the cells are way out of balance. What can happen is a few cells can reach full charge while 1 or two can still be undercharged.  If a few full cell reaches 3.8v LFP or higher than 4.2v LiMn the the charger reaches max volts and stops all the other cells getting a charge, in order to protect the high cells being damaged.

This can indeed happen with self discharge over a long period and a runt cell.  The battery can still work fine with a runt cell, but a runt cell can lose voltage a lot faster than the other cells over a long time..  Moisture on the BMS can also plague a pack to unbalance during long term storages.

When a high cell reaches 3.65v LiFePo4 the BMS also switches a resistor across the cells terminals, to keep it from over charging.  Some charges will not allow the cell to drop from this voltage if they stay plugged in, and the charge can provide enough hold voltage to keep them high.  When you switch the charger off, this resistor over all the high cells brings the voltage down over them  much faster thus taking the whole pack voltage down to not reading a full charge again.  When you switch the charger back on the charger hopefully will go back into full charge mode and give any low cells another decent hit of current, because the resistor has not switched over the low cells, they are not discharged by the BMS, so they start to regain capacity every time the timer turns the charger back on...

So we do this on-off timer thing, over and over, until the charger no longer goes to bulk mode charge rate after the rest period, and this is the indication that you've come to your end.  Try the pack out if you get 30+ kms down the road, then we won here.

[SydneyCommuter]

If its 16ah it's the older Nima one. If you take the end cover off you'll see the cells are cylindrical. The lipoe is square cells.  There are plenty of posts in here on them.
If when you depress the throttle the lights go out, then it's the battery management cutting it. If any LEDs stay lit then the prob is your controller.
In colder climates moisture seams the leading cause of these types of issues. One time I had moisture in my battery (rain related) similar symptoms I'd get lights but first bit of load and the bms cut it. After pulling everything down thinking the water had fried something I couldn't see any obviously fritzed components on the bms board so reassembled it all and all good no issue since.
So my advice open the case and make sure it's all dry first. If that doesn't get it then you need to hunt down any dud cells, there are posts in here on that from memory.
Cheers

[Sean]

Is there a way to tell for sure from the label or do I have to open it up.

Anyone know what actual chemistries were used by GM was for the 36V 16 AH batteries? Were there no 36V16AH LiFePo4 batteries ever produced?

LiFePo4
LiMn2O4
LiNiMnCoO2

[Sean]

Here is the info from the label:

36V16AH

Model: LFP-3616S

Serial #?: 361620100638

Also, plugged the unit back in and it is definitely charging as the light is strong red. I'll have to look closely at that light through this charging cycle because I don't think it was immediately red when I charged it the second time this spring (when I tried putting the battery to charger before charger in wall to balance the cells) - I would have called it more amberish at that time.

What should a normal battery charge cycle look like? Just steady red to steady green over 4-6 hours? Is there an "amber" indicator or is it just that maybe my green isn't very green? I can't say that I ever paid much attention to the lights until now since everything just worked.

If these are one of the LiMn variants, what should I do differently to try to revive the battery?

Right now I'm trying a regular charge cycle until I can decide if the battery charger is cycling on and off after it is done. If it isn't I plan to put it on a timer.

Thanks,

Sean

[Bikemad]

Hi Sean,

Unfortunately your battery pack will definitely have the LiMn₂O₄ cells, because LiFePO₄ rack packs were not available from GM when you purchased yours.

Is there a way to tell for sure from the label or do I have to open it up?

The LiFePO₄ packs have the suffix LT after the item code as in this picture of Gary's pack:



So the 36V LiFePO₄ version of the rack pack should be labelled LFP-3612S-LT.

I have my doubts as to whether your pack is going to become usable again, but you certainly have nothing to lose by trying.  If your pack is charging up to ~42V and still drops to 27V under load, I would say there is clearly a problem with the majority of the cells, and I don't think the cell damage will be reversible to any great extent.

I genuinely hope you can prove me wrong, but my gut feeling is that it's unlikely to happen.

Alan
 

[Sean]

Alan,

I really appreciate all the help and the fast responses.

A bit disappointing to find out for certain it is LiMn chemistry. I actually went over this in great detail at the time I bought my electrified bike and so did my dealer because I made it clear I wanted LiFePo4 based on my research. Fortunately I had that detailed out on my order/receipt and have a good email trail so I may be able to get some sort of discount on a replacement.

I'll consider myself lucky if I can get the battery to the point where it can be a backup I can keep at work that will take me 1 mile home if I get caught with a low battery. Last night it had dropped to 29.0V and recharged to 41.0V - 1V lower than it did last time - probably not a good sign. I'm a bit hesitant to take it very far because my house is at the top of a hill and it is bear to pedal up if the assist is dead.

I'm also going to use the info from your other thread to put a potentiometer between my gauge and new battery so maybe I can adjust the gauge get better feedback about a low battery.  Right now I only know it is low when I have about 1/2 mile to go. Doesn't do me much good if it happens on the way in to work!

These forums are a great resource - thanks again!

-Sean

[Sean]

No doubt about it - the battery no longer has enough juice to work under a load.

Does anyone know the expected life (years, not charge cycles) of the current GM LiFePo4 batteries?

I'm wondering if SLA would be more suitable to my profile (I only ride about 4 miles a day w/assist - maybe 50-80 charges a year).

Thanks,

Sean

[Bikemad]

Sean, if you're happy to do a bit of DIY, you could modify your failed pack to accommodate two of these lovely LiPo packs:


Click for more details.

The two packs shown in the above picture would cost $111.96 plus shipping, and you already have a battery case, charger and a suitable BMS to build a lightweight low capacity pack that would be perfect for your short trips.

Two of the above 5 cell packs would give you the correct number of cells for your 36V battery, but obviously it would only have 5Ah capacity instead of your original 16Ah.

The rated discharge of those LiPo cells is 25C continuous (125 Amps), which is well in excess of what a single Magic Pie requires.
As the charge rate is 5C (25Amps), you can still use regen without any fear of exceeding the recommended charging current.

According to the information on the Hobby King site, these pack weigh 0.643kg each (1.286kg for the pair) which I reckon is ~2.4Kg less than the combined weight of the 40 LiMn cells that are currently inside the pack.

I appreciate that the the thought of substantially modifying the existing pack might sound a bit daunting to you, but as the pack is already unusable, you don't have much to lose, and in my opinion it would be a much better option than using Sealed Lead Acid batteries.

I am not trying to imply that you should do this, it is just another option that you may want to consider.
You will have to decide for yourself whether you feel competent enough to carry out the conversion before you make any decisions, but if you decide that you would like to have a go at it, I would be happy to assist you with further suggestions regarding the installation of the cells and the modification of the wiring etc.

Alan
 

[Sean]

Alan,

First my batteries are definitely dead! No amount of charge on off brought them back.

I like your suggestion and If I could get more AH I'd consider it - but it seems to be cutting it close. Here is my concern. Maybe I don't understand AH correctly - but I see AH as a relative  indicator of how long will the battery last. So all things being equal if a 16 AH battery lasts for 16 miles on the flat, one would expect a 5 AH battery to last about 5 miles.

I have a few big hills and found that I could safely ride for 3 days (2 round trips a day) before I risked being stranded. I just got in the habit of recharging every Wed and Saturday. With a 5 AH battery I would expect to have just under 1/3 the range of the 16 AH battery. That means I will have to recharge my batteries every day and as they age or if I lose a cell or two I probably won't be able to make my two round trips anymore.

I am trying to decide if it is worth trying a GM battery again. I am hesitant:

As I noted in an earlier post I had explicitly inquired about the chemistry when I purchased and been told that I was getting LiFePo4 batteries and I made sure it was noted on my invoice. There was no reason to suspect otherwise when it arrived since the battery does not indicate chemistry and the model actually starts with "LFP-" which sure sounds like it should be LiFePo4. My understanding is the dealer also confirmed this with his distributor - GM USA. LiMn batteries only have half the lifespan of LiPo batteries, so I'm not surprised that mine are now unusable. My use pattern requires long life based on time, not a long life based on the number of charge cycles. This is why I insisted on LiFePo4.

Given the circumstances, my dealer was working last week to contact GM to get some kind of a discount on a new battery and just heard back today:

"I heard from them this morning.
 
Basically, they are telling me that  GS USA were (not honest) and they are not covering anything done through them."

The real word used for "(not honest)" was pretty harsh, but basically I'm being told that GM China is leaving its customers out to dry because they made a bad choice when they selected their US distributor. The name of the distributor was "Golden Motors USA" - hard to have that name without the support of "Golden Motors".

Very disappointing position for GM to take. My dealer is top notch and is willing to give me about $100 off on a new GM LiFePo4 battery but I'm not sure it is worth the risk. I find it amazing that GM would disown its US customers who bought from their chosen distributor and so I wonder what would happen if I buy again and have a problem.