48V12Ah Battery problem

[Leslie]

I tested mine with a simple 12V inverter and then simply plugged in a lamp. If one battery can output 160 watts, without dropping the voltage below 10.5V, I gave it a pass. It's a pretty rudimentary test, but it certainly proved that my batteries (except one) were able to output enough power.

Giving one of my 12ah SLA 15v for a while today and the other day seems to be OK.  I leave it on a solar panel to day and regulated it at 15v for as long as I thought it could be missing capacity.  About 10-20 minutes each time.

I think if you get any sulfination you can sacrifice cycle life for some return capacity but don't count on getting away with it too much,  Do some short runs and overcharge it for a little bit each time and see if things improve for that battery..

[GM Brazil]

You can see any problesm if you put a load on the pack like 4X12v 50 watt halogen bulbs in series.  Wait for a 5 minutes to strip any surface charge then test the voltages of each block to see if one cell block is suffering.

A bad conection to one or more Pcell in a block will show a faster voltage drop between the pins.

I'll try this!

Do I measure with or without load?

[Leslie]

You can see any problesm if you put a load on the pack like 4X12v 50 watt halogen bulbs in series.  Wait for a 5 minutes to strip any surface charge then test the voltages of each block to see if one cell block is suffering.

A bad conection to one or more Pcell in a block will show a faster voltage drop between the pins.

I'll try this!

Do I measure with or without load?

Ohh I came from hardcores thread.

Yes you test while under load as this will spot light out any bad cells blocks.

When a single celll goes out in a block the resistance increases because 1: The connection between the plates is damaged, and because there is more resistance ion one cell it increases the resistance over the whole parallel block.  Its like having 2 series resistors, three resistors in parallel, replacing one with a larger value in a series block.  Put an input and a load on it then measure the v drop.  The parallel block you put the larger resistor in will have the larger drop.

But with batteries its the opposite.  The voltage actually drops over the terminals.   When voltages vary too much between two parallel cells the difference actually makes a potential difference so one battery begins to discharge into another causing the voltage over that parallel block to drop faster and the peukerts effect effect should be larger on the suspicious block.  SO if you can brew up some decent load to draw good amps your test can be over much faster.

Just keep the lights on and keep testing the cell blocks until the runt show its ugly head.

Any thing with 2.5 ohm resistance and give up 1000 watts will do very close to 20 amps.  That's a real load test.

With a decent load you might just be able to pick up some amps between a damaged cell and a normal ones in a parallel block positive terminals. Look for some current flow when one cell discharges into another.  Maybe a load meter will provide enough parallel resistance to shunt some through to a damaged cell giving you a result from any normal current shifts between parallel cells.

You can get some peltiers 8 amp cheap on line.  They are good for 15v each so you could go for 4 in series for a 48v pack and cool your beer while you wait to find the bum cell.  I do try to encourage people to play with peltiers. lol

They invented peltiers as heat pumps for cooling, but it turns out they make very good heaters.  And infra red energy amplifiers. They can take the infra red sun off a roof and pump it it your home with more efficiency than they can cool things..  Well that how my theory goes.

[Leslie]

The Nicrome wire I got failed,  Too long, the rain got on it, a kink blew and the parallel resistance did not turn out as good along the strands as intended. Some twists in the wire overheated.

What didnt go wrong with the home made resistor.

SO I am going to turn to my LM338 and some halogen bulbs. Its all a bit more difficult than I first thought, I still have three LM338 now I can maybe squeeze 6 amps out of each.  I have two ready for 5 amps and some of the nichrome wire left..

See what I can make up.

[Leslie]

http://www.alltronics.com/cgi-bin/item/23P011/44/-25-Ohms%2C-125-Watt-Resistor

This maybe an expense alternative, But the Dealers may want to afford some quality resistors. You would need ten of these 250 watt resistors in parallel  10 of these in parallel will do 75v at 33 amps max and 50v @ 20amps nicely. at $8.50 each it isnt the cheapest.

If we going to get the us normal GM users to do load tests we will need something cheaper than these resistors and easier than a circuit of LM338s with bulbs.

[GM Brazil]

I'm using just paralel bulbs and is ok for low load tests.

But to detect the cell I'll really need at least an 0.5C test. I already discarted the option of the controller been the responsable for the cut-off with the low load test. The battery shows 50% of its capacity and the BMS cuts at 47v.

So, should be some cell group that is going low voltage before the others and then the BMS cuts down. there is such a thing? Cut down by the signal of one group or the BMS just do this with the whole pack voltage?

[Leslie]

I'm using just paralel bulbs and is ok for low load tests.

But to detect the cell I'll really need at least an 0.5C test. I already discarted the option of the controller been the responsable for the cut-off with the low load test. The battery shows 50% of its capacity and the BMS cuts at 47v.

So, should be some cell group that is going low voltage before the others and then the BMS cuts down. there is such a thing? Cut down by the signal of one group or the BMS just do this with the whole pack voltage?

If the BMS is cutting at 47v then the last questions remain.  Does it LVC  on a heavy take offs,  Whats its maximum current draw on the take offs.

In most situations you should see a difference.  You may need to be quick to check it when LVC happens,  But why it could be happening maybe a different reason.

If the connections between the cells are not solid this can cause cells to discharge faster. If there is an arcing, this particularly can cause heat and damage cell into its core, and you can pick a bad connection to a cell like this.  LVC may be protecting that cell so it might be still good.  Give the bike a good run and If you can probe the cells with a temperature gauge straight into LVC  .  You might turn with up some  interesting results.

When there is bad LVC at 47v if its your cells there will be some resistance  over the damaged cell and it will cause heat inside, be fast to test the temperature.

Also make sure that the BMS wires are connected properly.  It could be a false positive if one or more the B is intermittent or connection fails in the plug.  A simple cell test shows its connected.  And reseating the BMS B connections male plug into its female counter part may solve any intermittency issues.

A touch of Inox electronics spray in the plugs is not such a bad idea if you have salting or corrosion on the pins.  It can protect any charge on the pins from the salt air in any connections.

[Leslie]

Oh and one good thing the lower current load tests does provide. Is that it inst a heavy load that's causing the problem.  The high voltage 47v LVC does show there is still plenty of capacity in the other cells in your pack, so you need to test the cell voltages just before LVC, while you're testing the capacity wait around till just before LVC and give the cell voltages a test,



From
B- to B1
B1 to B2
and so on,

And write them down while you are doing the load test.  Be careful mot to short the pins and don't take too long,  Any cell block that are much lower just before LVC is the culprit cell block.

[GM Brazil]

Thank you!!

I'll try this weekend, but with the low current system, no free time to go after the 1kW resistor or nicrome wires now;

[dicko]

Hey guys

just wanted to check if you had any luck with your 48v12ah battery pack. mines got the same problem now. Below is my email to GM.

'Ive had my E-Bike up and running for around 6 weeks now and it’s running well, until the last few times ive riden it.
The first 4 weeks I could ride to and from work on one charge, 9km to work, 9km return.
Last week, I could only ride to work and halfway back, and the battery cuts out.
A few days ago, I could only ride to work 9km before needing a recharge.
Nothing has changed, I ride the same way each time, with hard tyres.
I noticed on the site that a 48v12ah battery should do 50km on a full 48v charge, but im not getting anywhere close to this.
I always use regen breaking where possible, and also peddle aswell.
Is there a way to check if any lithium cells are faulty?'

I did check all my cells and they looked ok. there were 2 cells which went up and down in voltage for no apparant reason, very strange.

Thanks for any help.

[GM Brazil]

I changed the battery of my voltimeter, compared its readings with true certified lab oscilloscopes. I'll let the batteries connected with the low power load now and when the BMS shuts down I'll read the voltages of the cells - even with the BMS off is possible to check the cells voltage.

I also checked the lateral soldering of cells (the middle is very "glued" I prefer to not try to open now). The construction is made of very, very good quality parts - only the screws look cheap. Better tools and quality control of screws can solve those ugly and damaged ones.

Then I'll post results. And no answers from TOM since the last e-mail with the first test results.

[Bikemad]

And no answers from TOM since the last e-mail with the first test results.

Bruno, your post with the results was dated 3rd June, are you saying that Tom has ignored your problem for over two months?

Have you sent him any other emails in between?

Alan
 

[GM Brazil]

Yep, and yep. There are more than two months since the results, not considering the first month for the first e-mail and this is my one battery problem. I'm concerned when it became a bigger number of clients.

I will have the equipment to test the battery for my clients, and check what is ok with the warranty and what is not - but if I need to wait like this to every problem my clients face I will have serious problems.

[Leslie]

I am fresh out of excuses as to why TOM doesn't even answer emails to an authorized rep.

Maybe they should put us tech officers on the payroll, lol. I don't care about making money, I care about this community.

People like myself has put a lot into building an Ebike community here.  Not to mention what Alan and the GM authoized dealers input as well.

Alan is the main reason these forums are so damned good.  We have lost the spam, along with other material that is not acceptable since.  And the wealth of information from all our member has been abundant. Overall attention to this site and GM sales are on the rise.

I see if I can get some info from our boss here, I cant promise any reply from GM or any follow up after this post but I will try.

If Bruno cant get a reply what chance have I got?

I will try, but don't expect anything more than what we are seeing.

[Leslie]

Ok I got my reply from GM.

Without copy pasting the PM I will condense it to this post below.

Ahh damn it, I don't like posting private messages, but this is not so bad thing to do in light of the good news it caries..

The information I think that matters to contents in this thread,  spell checked and I quote Golden Motor.  Yuan

hey Les,

Tom is extremely busy with the shipping since there have been quite big orders this summer. . Philip is ultra mega busy with the shareholders and with the new 12000m sq factory where we will build everything ourselves, so that the quality will change drastically as we have already employed 2 QC officers. it would last about 3 months, till December where the factory will be fully functional, making near-perfect Magic Pies and controllers, hence reducing technical problems, hence giving tom some air to breathe,

if you would care to let me explain. we sell about 9000 kits per year, that gives us a pure profit margin of 12%. hence we won't have enough money to find people like tom or quality control officers to operate this business. we would have to find shareholders (philip), spread awareness (me), get huge 5000W motor orders (tom and the rest of the sales team), get more dealers (me and the rest of the sales team)...

If Philip succeeds then there would be no more problems. pray for philip that he gets the investing companies to invest

IF in any circumstance, we do succeed. then you would be looking at a huge scale company. I mean the government has gotten their ass soaked into our pots and they want a piece of GM as long as we prove ourselves worth of huge government financing. we would be granted 4years of tax exemption, 90 000 sq m of operational ready factories, 2500 employees and a huge ass cheque to give us a boost in everything.

Incentive to invest is likely to prevail. Government supported e-vehicle company is on the radar in this day and age.  The stakes are high and the rewards are assertive and attainable. The rewards will be passed down to the end user and GM yet again will improve.  I am confident GM will prevail.

If you look around GM are facing the same problems at present all similar companies are facing. IMO there is not a single company that can successfully manufacture all products we need 100% to the customer satisfaction.  Buy your hub here, you battery there, and controller somewhere else and be lucky you have the balls to make it all compatible.  This to some is not a reality as they do not have the time nor skills to make other brand name products work with each other.

GM have plans to change this trend

IMO we are seeing GM at another point in their life, at no time it seems GM is going to give up, be sure GM will soldier on.  


My thoughts I have put to GM to ease our situational during this transition period are as follows.

I have some put some ideas towards GM to help improve the GM packs reliability out of the box.  Like an BMS access point plug to individually balance  and charge cells, and make it easy to test them.  GM to introduce a low amp single cell charger to connect to the BMS connection for sale to customers and made available for the dealers wishing to ensure the cell balancing is correct before put to consumer use..

Some new strategy of the Dealers may needed to be adopted.  Like individual charging cells and some small cycles be done to the pack before sending.

I we can get the individual cell block access connection to the cell blocks installed this will make it easier for everyone to balance cells individually, the port may be accessed inside the pack,

I cant make promises I will get what I ask, but its not much to ask for IMO and I would install this connection myself to the GM pack.

As the shelf life gets on with the GM packs, balancing may not be optimum when they arrive to the customer.  The access port will allow GM authorized dealers to balance cells properly before they send them making reasonable shelf life on the pack have no impact to the packs long term performances.

I cant say whether my idea will be implemented but I will say there is no reason why the GM packs cant have a plug installed easily by its user until we can get some changes..