Cell pack hybrid scooter no initialisation!

21221 Views 94 Replies 10 Participants Last post by djurica86, 1 mo ago

Josefmifsud1974

Discussion Starter · Sep 29, 2017

5

Hello friends I hope someone can help me in my dilemma.
I changed the cells in my hybrid scooter battery pack and when im trying to turn on from mosfet switch it remains blinking red.
The battery pack is 10 cells 3.7v at 31Ah.Has and integrated bms and 9 cells have a board between terminals.I removed the pcb chips and installed on the new cells.The bms has a can bus system also which communicates with the main ecu.
But im stuck the voltage is there but the bms is not giving the signal to swith on the battery and turn the led to green.
The battery is installed in the Hybrid scooter MP3 made by Piaggio.
I have more documentation if needed.
Hope I will have some feedback because my hopes are diminishing

See less See more

1 - 20 of 95 Posts

Solarsail

· #2 · Sep 29, 2017 (Edited)

Is this LiFePO4 and what is the configuration XsYp? With so little information provided, it is difficult to figure what the issues may be. The boards between terminals can be balancing boards, if this is a 10s, then there should be 10 of them (not 9).

What is a "hybrid scooter"? Does it also have a gas engine? Did you say that you removed the chips on the PCB? Why would you do that? Or is that an inaccurate description? Is the blinking LED on the BMS inside the pack, or is it controlled by the ECU?

I would think that the BMS is not getting the right signals from the balancers - due to a disconnect or a bad board, it cannot see the voltage across a cell, or some other issue with one of the 10 balancers. Looks like you already are suspecting a balancer and testing it, and there may be signs of fire damage on it?

Josefmifsud1974

Discussion Starter · #3 · Sep 29, 2017

thankyou for your reply and feedback i very appreciate your time and help and i will try give more infornation as you request and as possible to my knowledge.
Yest the MP3 Piaggio scooter hybrid has both eletric and gas engine which work in tandem.some time ago it just stopped and the battery died , from there the engine start but stays idle.
the battery pack is sealed so i gave it a try to regenerate the pack.
So i opened the pack and found that all cells were at 0v,there are 10 cells in side the pack KOKAM brandded 3.7v 31AH.
So i managed to find 10 cells of the same capcity to replace them which are 3.7v 30Ah UNITEK.
So i dismatelled the original cells removed the pcb boards from the original cells and installed them on the new cells so all is back to how it was.
Yes there are 9 pcb boards with a chip and the last board is just taking 3.7v from the last cell and supplying the terminal part unit which has a mosfet switch,in fact connecting just the last cell it start blinking red led which states that the battery is disarmed.
my frien technician tested one board and its not fire what looks on the board but there was black silicone between the boards and the bar which connects the batteries in series.My fiend tried to look the software of the chip but its locked so if a chip is gone we cannot overwrite a new one.
The termials on the pcb board one of them shows voltage the other i dont know.

See less See more

Josefmifsud1974

Discussion Starter · #4 · Sep 29, 2017

4

Some more data and photos.
The old cells are the wide ones,the pcb boards i removed from the old and installed on the new.I charged the new cells one by one to 4.23v.The third photo is the Kuadro VMS ecu which charges and controls all the electronics in the scooter.

See less See more

Josefmifsud1974

Discussion Starter · #5 · Sep 29, 2017

This it the hybrid system overview whith some technical data i managed to retrieve from friends has some usefull information for who is more profesional than me and knows more into electronics.

Josefmifsud1974

Discussion Starter · #6 · Sep 29, 2017

Josefmifsud1974 said:
Some more data and photos.
The old cells are the wide ones,the pcb boards i removed from the old and installed on the new.I charged the new cells one by one to 4.23v.The third photo is the Kuadro VMS ecu which charges and controls all the electronics in the scooter.

The red button its the intialising switch and next to it there is an led the first photo which shows the terminals,To initialise the battery you switch ignition on and push for a second the button which should turn to green and starts to comunicate with the main vms kuadro.

Solarsail

· #7 · Sep 29, 2017 (Edited)

You are not answering my basic questions. Please read my comment once more and answer ALL questions. Are the 10 cells in parallel or series? I assume series. Are they LiFePO4 with max 3.7V or are they Lico with nominal 3.7V?

If the mosfet switch is in the battery pack and part of the BMS protection, then just do away with all the balancing and protection, and get your scooter going. Then buy a $10 balancer board and use that to charge the pack. So -

Remove all boards and BMS and mosfet switch in the pack.
Connect all 10 cells in series, and charge them externally with a charger.
Put a high amperage diode with heat sink in series with pack so the scooter generator will not charge it. (Or put a voltmeter and continuously check the voltage to make sure it does not over charge.)
Get your scooter to work.
Then get a 10S protection balance board, with the proper current rating, and install that, and remove the diode.

If the ECU expects an OK from the BMS, the above will not work. Maybe you can make the BMS happy without it being connected to the balancer boards, so it tells the ECU all is fine?

The BMS receives a signal from the balancer board that the voltage has not exceeded the threshold (4.2V for li-ion) or the voltage is not below a threshold (3.0V for li-ion). If you remove the balancers, then find a way to send this signal to the BMS and make it happy. Sometimes the signal is sent from one balancer to the next in series, until the last balancer sends that to the BMS.

Since there are 9 boards and not 10, I suspect this is a serial balancer and the signal is sent from one board to the next board and then to the BMS. Is the BMS connected directly to all 9 boards (i.e. parallel balancer) or is the BMS connected just to the last board (serial balancer). If it is serial, you will have to find out the protocol of the signal to make the BMS happy.

If you can make the ECU happy and remove all boards and BMS, then use this board, or many other boards on eBay. Make sure the amperage is sufficient for the scooter. What is the max amperage?

http://www.ebay.com/itm/10S-40A-36V...169897?hash=item41af3217a9:g:R~IAAOSwXeJXc3KJ

See less See more

Josefmifsud1974

Discussion Starter · #8 · Sep 29, 2017

10 in series to a total of 37v, lion based cells

Josefmifsud1974

Discussion Starter · #9 · Sep 29, 2017

yes the ecu wait an ok from bms threw Can bus sustem.
and yes the boards are connected one to the other wit 2 wires from one to nest till they connect to the terminals part were there is the mosfest switch.
Is it possible that 4.23v blocks the unit.
I will discharge the cells to 4.12 each and recheck.
So your opinion is that these boards if the values are correct will make it work.
is there a way to read the chip

Josefmifsud1974

Discussion Starter · #10 · Sep 29, 2017

Solarsail said:
You are not answering my basic questions. Please read my comment once more and answer ALL questions. Are the 10 cells in parallel or series? I assume series. Are they LiFePO4 with max 3.7V or are they lithium ion with nominal 3.7V?

If the mosfet switch is in the battery pack and part of the BMS protection, then just do away with all the balancing and protection, and get your scooter going. Then buy a $10 balancer board and use that to charge the pack. So -

Remove all boards and BMS and mosfet switch in the pack.
Connect all 10 cells in series, and charge them externally with a charger.
Put a high amperage diode with heat sink in series with pack so the scooter generator will not charge it. (Or put a voltmeter and continuously check the voltage to make sure it does not over charge.)
Get your scooter to work.
Then get a 10S protection balance board, with the proper current rating, and install that, and remove the diode.

If the ECU expects an OK from the BMS, the above will not work. Maybe you can make the BMS happy without it being connected to the balancer boards, so it tells the ECU all is fine?

The BMS receives a signal from the balancer board that the voltage has not exceeded the threshold (4.2V for li-ion) or the voltage is not below a threshold (3.0V for li-ion). If you remove the balancers, then find a way to send this signal to the BMS and make it happy. Sometimes the signal is sent from one balancer to the next in series, until the last balancer sends that to the BMS.

Since there are 9 boards and not 10, I suspect this is a serial balancer and the signal is sent from one board to the next board and then to the BMS. Is the BMS connected directly to all 9 boards (i.e. parallel balancer) or is the BMS connected just to the last board (serial balancer). If it is serial, you will have to find out the protocol of the signal to make the BMS happy.

If you can make the ECU happy and remove all boards and BMS, then use this board, or many other boards on eBay. Make sure the amperage is sufficient for the scooter. What is the max amperage?

http://www.ebay.com/itm/10S-40A-36V...169897?hash=item41af3217a9:g:R~IAAOSwXeJXc3KJ

The continuous discharge is 60A,on the battery it says 37v 31Ah 100A 1/2Kw

Solarsail

· #11 · Sep 29, 2017 (Edited)

Looks like the balancer boards have an Atmega micro controller. So the signal to the BMS is probably a digital signal. Since the BMS sends a Can bus signal to the ECU, you will have to spoof that.

Sorry, I have no idea about the Atmega board and what makes it happy. Maybe there is a break in the signal path to the BMS? Maybe one cell is overcharged? Maybe one board is not functional. You will have to figure out the signalling from one board to the next and see which board is stopping the signal and work on that board.

Solarsail

· #12 · Sep 29, 2017 (Edited)

4.23V is definitely overcharge and is bad for the cell, and reduces the capacity over time. You should never charge over 4.15V or 4.2V. Reduce all charges to below 4.2V.

The Atmega chip may think there is something wrong with the overcharged cell and refuses to OK it. Check the signalling between the boards and see which board alters the round-robin signal.

Solarsail

· #13 · Sep 29, 2017 (Edited)

Is this hybrid a serial hybrid? Is the gas engine connected to the wheels and motor or just to a generator?

The manual says it is not serial hybrid. What a waste.

And the battery is so small. Only 0.5 kWh.

I think you should spoof the signal to the ECU, get rid of this battery and put proper 18650 cell 1.5 kWh battery of same size instead. Just my opinion.

Solarsail

· #14 · Sep 29, 2017

I would reduce each cell to 3.5V and see if the BMS comes up. This is the range of both Lico and LFP.

Could it be that the old cells are LiFePO4 (2.8V - 3.6V) instead of Lico cells? Since the battery is prismatic, it could be LFP. Have you looked up the chemisty of the Kokam cells?

Josefmifsud1974

Discussion Starter · #15 · Sep 30, 2017

I am attaching KOKam specs
I very aprecitate your support

Josefmifsud1974

Discussion Starter · #16 · Sep 30, 2017

Solarsail said:
Is this hybrid a serial hybrid? Is the gas engine connected to the wheels and motor or just to a generator?

The manual says it is not serial hybrid. What a waste.

And the battery is so small. Only 0.5 kWh.

I think you should spoof the signal to the ECU, get rid of this battery and put proper 18650 cell 1.5 kWh battery of same size instead. Just my opinion.

The engine and electric motor are connected to the wheel as one,during braking it charges the lithium battery and supports the gas engine when needed or vis versa.When the battery shows no com it blocks the engine to only idle and the scooter is stuck cant be driven.

Josefmifsud1974

Discussion Starter · #17 · Sep 30, 2017

Solarsail said:
I would reduce each cell to 3.5V and see if the BMS comes up. This is the range of both Lico and LFP.

Could it be that the old cells are LiFePO4 (2.8V - 3.6V) instead of Lico cells? Since the battery is prismatic, it could be LFP. Have you looked up the chemisty of the Kokam cells?

yes i will reduce the voltage in the cells,at present i noticed that the cell is showing 4.23v between terminals and when i check negative and one of the pins on pcb board there is voltage 3.98v on the other pin i dont know what is transmitting,so which voltage i should look to keep in range so that it can activate the mosfet switch?
Thankyou again

Josefmifsud1974

Discussion Starter · #18 · Sep 30, 2017

Solarsail said:
Is this hybrid a serial hybrid? Is the gas engine connected to the wheels and motor or just to a generator?

The manual says it is not serial hybrid. What a waste.

And the battery is so small. Only 0.5 kWh.

I think you should spoof the signal to the ECU, get rid of this battery and put proper 18650 cell 1.5 kWh battery of same size instead. Just my opinion.

Would be interesting but how to sppof the bms im not that expert to go around it.

Solarsail

· #19 · Sep 30, 2017

It could be that 4.23V is too high and the Atmega thinks this is a bad cell or that the generator or charger are feeding too much voltage, or that the BMS is broken and overcharging the cells. So it sends a fail message to the BMS terminal board, and that refuses to go green light and turn on the Mosfet.

The spec says the Kokam batteries are LiPoly, so they are Lico, and go up to 4.2V. Do not overcharge. Reduces battery life fast.

Number the balancer boards - #9 sends message to BMS. #8 sends to #9, etc. Then look at board #1. Does it receive a message from BMS or anywhere? Then see what message it passes to board #2. Check the message from #2 to #3 - is it same OK message? Then from #3 to #4, etc. Until you find a board that gets OK message but sends a not-OK message to the next board. Call this board #x. Swap #2 board with #x. Is board #x getting OK from #1 but sending not-OK to #3? If yes, then #x board is bad. If #x is good, then look at board #2 that is now in the original place of #x. Does it get OK, but pass not-OK? If so, then something wrong with the cell, or with the connection to the cell or the connection between cells at this location.

You can also easily force the Mosfet to turn on - just connect source to drain, or put bias on gate. But I am afraid that BMS has already sent a not-OK message to ECU and even if you bypass the mosfet, the ECU will refuse to come on.

See less See more

Josefmifsud1974

Discussion Starter · #20 · Sep 30, 2017

So 9 boards have 2 wires conecting them together.
From one board to another till they connect to termal board.
The last cell in the pack at the bottom has two wire which go on a seperat way to the terminals block.when this wire is connected the red led starts to blink so 3.7v from cell 10 .
One terminal of each boards shows voltage the other i dont know.
For which ok signal i should look

1 - 20 of 95 Posts