Bli medlem i Norsk elbilforening og støtt driften av Elbilforum. Som medlem får du i tillegg startpakke, medlemsfordeler og gode tips om elbil og lading. Du blir med i et fellesskap som jobber for mindre utslipp fra veitrafikken. Medlemskap koster 565 kroner per år. elbil.no/medlemskap

Enerdel battery fault - repairing...replacing...

Startet av pexic, tirsdag 23. april 2019, klokken 15:56

« forrige - neste »

pexic

Me again

Some more info about RLEC here : http://www.metricmind.com/audi/21-rlec.htm

I've also noticed that I have another RLEC board that is very hot comparing to the rest of them so, for now, I have to buy 2 RLEC boards, If someone knows from where I can buy those please tell me.

Thank you very much in advance !

mohpet

2011 Think 4-seter med Lithium (ex Zebra), ESP, a/c og PTC varme
E-twow Booster 2S sparkesykkel
ex 2003 Kewet 5
ex 2020 Vespa Elettrica L3 (70 km/t)
ex Go Motorboard sparkesykkel
Asker

pexic

Thank you Mohpet!!!

Meanwhile I have a small idea (small for now) :
Let's assume that RLEC board is still partially functional, only some smd transistors are fried (as I saw in all other bad RLECs stories on the net).
What if...it would be possible to fool the RLEC boards by connecting all Analog_to_Digital inputs of MC908GZ32VFAE (all the inputs used for measuring the cells voltage) to one single source of voltage.
Meaning : this CPU measure all 12 cells, there are problems between some cells and respective A/D input of CPU, disconnect all A/D inputs used to measured those12 cells and connect these intputs-together- to one single voltage -let's say 3.5V.
This way that specific RLEC board will report to MLEC that all 12 cells have 3.5V.

Don't know if this will be possible but, using an separate and independent BMS, it will be possible to change the cells type - to use Leaf cells for example, or any other type of cells as long the respective cell pack will have  30...49V.

The big problem of this "hack" is : there are 24 A/D channels needed to measure 12 cells  temp + voltage and this CPU only have 16 pins for A/D inputs... so, most probably the measurements are multiplexed : external circuits connects at high speed different thermistors/cells to the same A/D input pin. If the temperature and voltage to be measured are mixed -> ...fail. If the voltages and temperatures use different input pins -> there is hope :)

We will see, I will start measurement on one good RLED, let's hope for the best.

Last chance -in my case- is to start to learn arduino/Rpi/whatever programming  :))

Have a nice weekend Th!nkers ;) and thanks again Mohpet.

p.s. sorry for my English.

pexic

Hi friends

What I found out by now (see pictures):
Du kan ikke vise dette vedlegget.
All cells voltages measurements reach U1 integrated circuit ( INA 148  +/-200v common-mode voltage difference amplifier ) at pins nr 2 and 3.
At the U1 output is one resistor (R20) then one of the inputs of U2 (pin nr 13).
U2 is an analog multiplexer (MC74HC4051A) and mix also some other values that needs to be measured by the CPU. But all cell voltages are sent by this- pin13- input to his output (pin3) and then to A/D pin 24 input of the CPU (CPU is U6).

So, I removed R20 and I injected a fixed voltage to U2-pin13 input.
This was done by connecting an 100k potentiometer on the back of the board. This was connected to +5V and GND also (after D4 and D3 were removed).
Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.

The result was : RLEC is reporting same voltage for all 12 cells.
RLEC 04 voltages before modification:
Du kan ikke vise dette vedlegget.

RLEC 04 voltages after modification:
Du kan ikke vise dette vedlegget.



All good so far but work is not done on this part, next days I have to connect the first cell directly to U2-pin13 so RLEC report voltages more close to reality.
Also the top balance feature was lost and it must be replaced by another stand alone BMS ( more work but an stand alone BMS is cheap and easy to "install".

Okeey, having no more 0V cell voltage reported I start to check the car behaviour:

Now the car is charging, both batteries in fact ( the 12V battery is also charging in parallel with the Enerdel battery)
No errors are displayed in dash during charging.

Ignition On :  only two red leds are displayed in dash now :
red 12V battery led and
red wrench led.
Car does not start do these errors.

This 12V battery red led -> don't know why is still on... the 12V battery is charging to maximum 13,11V according to CAN (13,4V according to multimeter)


I will attach more test results pictures from Comtool maybe someone will notice other strange data.

Also - if someone will try this ugly RLEC hack described by me please REMOVE all the burned smd components from the board before start those modifications!

Thank you in advance for your help, see you soon.

pexic

Amd here are the test results after hack:
Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.Du kan ikke vise dette vedlegget.

BauDemo

RLEC 03 does not seem to be healthy...
and it could be the reason for the car not wanting to start.
laddplats? -www.uppladdning.nu
nikometer? - www.evmonitor.info

worry

This is really good work pexic!
I don't understand much of what you explain here, but you seem to understand a lot of the logic on the rlec. Keep up, and please tell us about what you find. Thank you.  :+1:
Think lithium

pexic

#22
Hi BauDemo & Worry

RLEC03 has received the same hack as RLEC04 now, car simptoms are the same :(

I have replaced the potentiometer with one resistor to ground (22k...47k) and I have connected the first +Cell to U2 pin13.
Du kan ikke vise dette vedlegget.
Now the RLECs (03 and 04) are reporting that all 12 cells of them have the voltage of the first cell of each one.

For example : if RLEC03 first cell has 3.3V -> RLEC03 will report to MLEC that all his 12 cells have 3.3V.

This way the info/voltages is/are more close to reality.

Next step is to see what the .. is this 12V battery error:)

With ignition=off (key removed) only 25mA are drawn from the 12V battery.
With ignition=on and all accessories OFF there are 4.2A drawn from the 12V battery , is this normal ?
Don't know for sure but it seems wrong for me so now I intend to remove all fuses, one by one, and check where those 4,2A are directed.
I don't think the PCU unit needs so much current but who knows ? if someone can make this test on a good car please tell me the cost :)

On the good side - my Enerdel battery is fully charged now :)

I'll keep you posted, see you soon

BauDemo

The Enerdel battery pack uses approx 3-4A when contactors are closed.

Check the fault codes reported by the PCU via CHAS.
laddplats? -www.uppladdning.nu
nikometer? - www.evmonitor.info

pexic

Hello friends

I need your help again, as usual :)

I use this schematic to make a serial cable in order to use CHAS software.
Du kan ikke vise dette vedlegget.

Please confirm if this info is accurate.
The CHAS software report: no communication
I've tried under WindowsXP on two computers:
1- Laptop with serial to usb chinese adaptor (CH340)
2-Computer with COM port
I always used "COM1".

-> no communication at all - like the PCU is dead ...

I really hope is not dead , how can I check ? Is this above info/picture accurate ?

Thanks in advance for your help

BauDemo

Did you read the documentation? You need to go through the setup button every time you start CHAS.
laddplats? -www.uppladdning.nu
nikometer? - www.evmonitor.info

pexic

#26
 yes :( every time I start the CHAS 2.6 softvare:
- I have to choose  "DefaultVariableData.var" location (which is in program files/chad where I save it the first time after instalation)
-Then I enter setup and select "compatability mode"  The COM number (1) and baud rate (38400) are also selected (same COM and baud rate are also in Computer / device manager / com ports )
I've read many times the CHAS V2.6 user manual doc file, even before the first instalation.
I'm lost, I could not belive that PCU refuse to talk with me, I never told bad things about him to anyone :)
Maybe I need to pre-install some other softvare (like NET.framework for Comtool )  ?
Or maybe I have a defective/old CHAS version, could you please zip and send me your version by mail ?
plastic4pexic@yahoo.com
Thanks in a million times Nikolai !

pexic

Hello, it's me  :)

I find out the problem: Rx and Tx must be inversed, see picture for details:
Du kan ikke vise dette vedlegget.

I still have some not accesible menus :
Du kan ikke vise dette vedlegget.

What I was able to extract:
Du kan ikke vise dette vedlegget.
Du kan ikke vise dette vedlegget.

Now I have to see what these info means...
Have a nice day to all!

BauDemo

Do you mind publishing photo of your cable and annotate it?
What connector do you have in the car? Is it the square with 2x2 pins or the other with the 1x4 pins?
You can always test the GND pin.
The RX and TX pins are trickier but one is always 12V I think.
Unless you tried too much and burned the RS232 driver in the PCU, but I have never seen that happen before.

I am using an FTDI RS232 to USB adapter and works every time.
https://www.ftdichip.com/Products/Cables/USBRS232.htm
Don't remember the model, but it was one of the white box ones without lights.



laddplats? -www.uppladdning.nu
nikometer? - www.evmonitor.info

BauDemo

OK, posted at the same time...
So VcSta is 2 - "Ready"... so that is ok
PrndSta is 0 - "Park"... this is ok if the car is in Park, and it will not drive, but if you have the stick in "D" and the PrndSta is not showing 3 "Drive", then there must be an issue with the gear shifter.
Check this, and post back.
I have seen some reports of microswitches in the gear shifter needed to be changed.
laddplats? -www.uppladdning.nu
nikometer? - www.evmonitor.info

© 2024, Norsk elbilforening   |   Personvern, vilkår og informasjonskapsler (cookies)   |   Organisasjonsnummer: 982 352 428 MVA