Do you have a dead Ezee battery?

[Jimbob]

Hi there - a Newbie to the forum here.


I've got an Ezee electric bike, powered by a 36V 14Ah Li Ion battery.
It was great, then the battery just stopped working at all, then it started working again, then it stopped, and I took the battery to bits.
I used my LiPo balance charger (for models) to do a balanced charge & discharge & the cells are fine - they have plenty of power.
BUT as soon as I try and draw any juice from the whole pack, the voltage drops from 42V to zero.
I'm 95% sure that the BMS (Battery Management System) is to blame. I can't find any dry joints / loose wiring, so....

Does anyone have an old dead Ezee battery from which I can steal the BMS?
Or do you know where I can get a replacement?

I'm fine with a soldering iron, so I'm happy swapping it over.
I'm also VERY aware that I've got 0.5KwH of power & that needs to be treated with extreme caution.

Thanks,

James

 

[Deleted member 4366]

Why don't you by-pass the BMS? The controller should still have voltage control, so will protect the battery from over-discharge. You can still charge through the BMS, which should keep the it balanced and prevent over-charging cells.

The older Ezee batteries have the block-shaped Phylion cells, which have no capacity when they're worn out, so as soon as you put a load on them, they sag down to LVC and trip the BMS. If you then check them with a voltmeter, they bounce back to 40v or more, which gives the impression that they're healthy. Check that because there may be nothing wrong with your BMS.

 

[Jimbob]

d8veh,

Wow - thanks - that's really given me some insight - I get it - I''l give it a go at the weekend.

I always watch the lights on the display anyway & re-charge before it's sitting on red.

Thanks again,

James

 

[Jimbob]

Hi d8veh,

That's wierd - I posted that diagram yesterday with a whole load of text, but the text has vanished......

Lets tray again:

This is a drawing of the wiring of my BMC:



I want to bypass it for powering the bike, but leave it working for balanced charging as you suggested.

As far as I can see there are three potentially valid ways to achieve this:

1. disconnect both thick blacks - controller -ve and batt -ve, and connect them together off the board
2. disconnect the controller -ve from the left of the board and connect it to the right of the board with the batt -ve
3. disconnect the batt -ve on the right and connect it to the left of the board with the controller -ve.

I feel that option 1 has to be a non-starter, as a heavy duty connection from the BMC to the battery is needed (not just the balancing wire).

Because of the position of the connection from the charger socket to the board (on the right) my initial assumption was that option 3 is the correct approach. I am however surprised that the charger is connected there.

Having just had a more detailed look at the PCB, option 3 is my favoured approach, as the GND rail for the board comes from that connection.
I would then need to move the charger connection to the batt -ve (and the electrolytic capacitor).

The thing that's throwing me slightly is that the charger is currently connected to the controller (output) side - that means that the charging current has to flow backwards through the cut off circuit.

Can you (or anyone else see any issues here? I'm aware that this battery has a whole lot of stored energy & I need to proceed with extreme caution....

My soldering iron & DVM are at the ready!

Cheers,

James

 

[Deleted member 4366]

Before anything, you should find some way of connecting your meter to the cell output wires. I solder thin wires onto the connections and wrap the other ends around the probes, or you can use crocodile clips etc.

Connect the base-plate to your bike. You'll need to support the rest of the battery somehow - perhaps bungee it in position. Lift the wheel off the ground and spin the motor with the throttle, then apply the brake and observe the voltage sag. If the BMS cuts without significant sag, it's a BMS fault. if the voltage sags more than 4v first before it cuts, it's a cell fault. Which cells do you have: The thin pouch cells or the thick block type?

 

[Jimbob]

Got it.

The pack ref. is AE906313PM1H2E (37V 14Ah 518Wh EA)

pic of cells:


I can do a full test on the bike later, but I've done a quick bench test now...

Batt voltage under no load:
41.1V
voltage whilst running small DC fan drawing 2.9A directly:
40.2V (and on disconnecting comes back up to 41.1V)


running the fan through the BMC, it gives about one rev - the tiniest blip, and stops.
the Batt voltage doesn't even flicker away from 41.1V

by my sums that makes the test motor around 120W. Is that a big enough test load? I think it should be?

(I really should have got as far as this test without help - sorry for being so slow)

Thanks,

James

 

[Deleted member 4366]

So it's looking like the BMS is shutting off power. It could still be due to a duff cell, so measure the voltage between the cell-pack negative and each of the pins on the multipin connector to the BMS, and then to the battery positive. You should have 10, 11 or 12 results, but, if 11 two wil be the same, if 12 2 pairs will be the same. Post those results. If you can, measure them while the Fan's running. Hopefully you can probe the connections on the pcb.

 

[Jimbob]

Here goes - voltages whilst under load:

40.2
36.1
32.1
28.0
23.9
19.9
15.9
11.9
7.8
3.8
0.1

(the 0.1 was the voltage drop over the wire whilst under load)

so the individual cells are:
4.1
4.0
4.1
4.1
4.0
4.0
4.0
4.0
4.1
4.0
3.7

So one cell is a good bit worse than the others.

I repeated the same under no load:
41
36.8
32.7
28.6
24.5
20.4
16.2
12.2
8.1
4
0

and again, the diffs:

4.2
4.1
4.1
4.1
4.1
4.2
4.0
4.1
4.1
4.0

So the sags under load are:

0.1
0.1
0
0
0.1
0.2
0
0
0.1
0.3

Meaning that one cell is a good bit worse than all the others, but it's not below my understanding of the discharge safety limit, which I thought was 3.6V.

thoughts?

J

 

[flecc]

The BMS is often powered by one end cell and that can be a bit lower as a result. The discharge lower limit is usually at 3.1 or 3.2 volts, so that 3.7 is ok.

D8veh has a technique for charging the low cell, something I've never tried but hopefully he'll advise later.
.

 

[Deleted member 4366]

At 4v unloaded, I don't think charging is going to help. It seems to be a lot weaker under load, but, as it didn't reach 3.1v, there's no reason why the BMS should've tripped; however, it's got a nasty smell of impending doom!

I have four suggestions:

1. Have a good look at the shunt/s where they're soldered to the PCB. Are any of the joints substandard or cracked. Resolder them just to make sure.

2 If that doesn't work add solder to about 30% of the shunt to see if anything improves

3 If you measured those voltages on the PCB, this probably won't work, but worth a try anyway. Disconnect and reconnect the multipin connector a few times to clean up the contacts.

4 Bypass the BMS discharge. Unsolder the output negative, and re-solder it to the black one from the cells, so that both are fixed to the B- position. Leave the charge negative where it is.

Having read your first post again, my bet is on a bad or broken joint somewhere.

 

[Jimbob]

those voltages were measured on the PCB - it has nice test points by the balancing plug.

I'm not quite sure what you mean by the shunt(s) - are they the 4 large power MOSFETs that are doing the switching? pics of both sides of the board, in case they help:




Given the intermittent behaviour I've experienced, I've checked & re-checked everywhere for dry joints, and have done the same just now.

I'll add the extra solder when I'm clear where I'm adding it, beyond that, the only thing I can think of is to disconnect the board entirely and bake it to re-flow it, but that feels risky given the fact it's working fine currently as a charger.

James

 

[Deleted member 4366]

It has solid state shunts. Two on each side marked R01F. Check the joints, but I don't think that the problem is there because there's four of them. The FETs are switched on/off by the gate transistors that switch about 12v to the gate leg of the FET, which is the left one as you look at the front with the legs down. Measure between any gate leg and the battery minus to see if it has the 12v, and whether it goes off when you put it under load.

 

[Jimbob]

In the pic above, the batt -ve is on the left.
Under no load, so switched on I guess,
The gates on the pair of FETs there are at 0V.
The gates on the pair on the right are at 5V.
the output is at 0V.

If I then connect a load & "trip" it,
the output goes up to 40V, then slowly decreases
the gate for the FETs on the left stay at 0V
the gates for the right pair go up to 0.4V less than the output

Something here is clearly not right, but I'm having difficulty tracing back to find the driving transistors for the FETs.
I might look further tomorrow, but I'm getting close to just bypassing that part of the BMC.

J

 

[Deleted member 4366]

It's usually at that point I go for the bypass.

You can use any BMS, which will allow 20A or more discharge that'll fit in the case. You don't have to us an Ezee one.

This one looks OK:
http://www.elifebike.com/peng/iview.asp?KeyID=dtpic-2014-3C-MR54.027KF

 

[Jimbob]

HI d8veh,

I've bypassed the switch side & the bike works perfectly - loads of torque - it all looks great.

I'm a little torn about whether to actually replace the BMC or use it as is, just ensuring I recharge before the voltage gets too low (using the indicator LEDs on the handlebars).

The current BMC handles the charging fine, and there's a risk the new one might not/ I might mess up the re-wire.

Anyway, do you have a paypal account so I can buy you a beer to say thanks for all the help?

Cheers,

James

 

[Deleted member 4366]

You should be OK like that as long as you keep an eye on your panel lights. The only risk is that that low cell reaches the bottom sooner than the others, so it gets over discharged. Don't take it any lower than you have to.

I do everything for love, not money . I get my satisfaction from hearing that you got going again.