Speak EV - Electric Car Forums banner
1 - 3 of 3 Posts

·
Registered
Joined
·
75 Posts
Discussion Starter · #1 ·
I own a 2012 C-Zero that now has about 29000km on the clock and I made a repair to the battery nearly 2 years ago in which there was a failure of the LTC6802 chip on the PCB of CMU08 that meant the cell voltages of that module were not being read properly and resulted in errors and breakdown of the car.

That repair was successful and the car has been great since then apart from occasional errors the last few months and eventual total failure again last week.

I made diagnostics with Canion and Diagbox to check out the issue.

Canion showed me that some cells in CMU11 were showing low voltage and as I monitored the voltages in real time, the values of those cells were jumping all around the place from around 3.85V to 3.95V. The image below shows the worst case:

138795


This suggested to me a failure of the LTC6802 chip on CMU11 as the chances of two adjacent cells failing at the same time is very low.

I followed this up with diagnostics with Diagbox and saw even more extreme variation of the voltages of those cells, albeit different cells to the onces observed in Canion, but still in the same module:

138796


It is worth noting that the system is measuring a total voltage across the module of 32.505V and that the sum of the individual cell voltages is somewhat different to that at 5.56V less. It really tells me that the problem lies in the measurement of the individual cell voltages and that the cells are operating normally.

So I had total confidence that the LTC6802 chip on CMU11 needed replacing. I already had two of these chips that I bought in case I needed to make this repair.

I put the car up on blocks and disconnected the 12V battery, removed the seats and took out the service plug. I removed both the covers on the underside of the car and opened the HV cable connection ports to check the voltage was 0V which it was. Then I disconnected all the cables and used the threaded bar method to lower the battery. I managed all this in around two and a half hours. Opening the battery was quick with my impact driver and then I looked for CMU11 using the following diagram (thanks Kiev):

138797


I actually already had the CMU numbers written on the modules from my last repair but wanted to be sure so I put the battery on the right hand side of the car and jacked it up so that I could plug in the comms leads to the car and make Diagbox diagnostics on it to check that when I disconnected CMU11, that did indeed correspond to the non-communicating CMU in Diagbox:

138799


Next I removed CMU11 and took out the main PCB. I used Chip Quik low melting point solder alloy and flux to remove the 44 lead SSOP package of the LTC6802 chip and then cleaned up the area before soldering the new one in place taking great care of the orientation and alignment with the pads. Note there is a conformal coating on the PCB that is a bit annoying, but it peels off OK and is not too bad.

I then put the PCB back in the CMU11 module and returned that to the battery pack. The pack was still connected to the car so I performed Diagbox diagnostics and saw correct voltage measurement in CMU11:

138801


Now there is only a 60mV difference which is a reasonable margin or error.

So, confident of success, I put the lid back on the battery, battery back under car and raised using the threaded bar method. Then I basically reassembled / reconnected everything I had previously disassembled. Seats back in, error codes cleared with Diagbox and car back on the road working well again.

I don't know why my car has now had two of these failures. My last intervention had nothing to do with CMU11 so I think just bad luck. Maybe a dodgy batch of LTC6802 chips were used in the assembly. At least I can now repair it for €30 (just the cost of one of these chips). Most of the repair is the labour of disassembling and reassembling the car although the diagnostics are an important part of the repair without which I wouldn't know what to do. Total repair time was two half days.
 

·
Registered
Joined
·
705 Posts
Nice one, i've done this repair twice on 2 different cars but having 2 fail in the same car is a first. The chips are basically just weak and fail after a while, you've got 12 of them in each car so more likelihood of failure I guess.

Nice work !
 

·
Registered
Joined
·
5,589 Posts
I own a 2012 C-Zero that now has about 29000km on the clock and I made a repair to the battery nearly 2 years ago in which there was a failure of the LTC6802 chip on the PCB of CMU08 that meant the cell voltages of that module were not being read properly and resulted in errors and breakdown of the car.

That repair was successful and the car has been great since then apart from occasional errors the last few months and eventual total failure again last week.

I made diagnostics with Canion and Diagbox to check out the issue.

Canion showed me that some cells in CMU11 were showing low voltage and as I monitored the voltages in real time, the values of those cells were jumping all around the place from around 3.85V to 3.95V. The image below shows the worst case:

View attachment 138795

This suggested to me a failure of the LTC6802 chip on CMU11 as the chances of two adjacent cells failing at the same time is very low.

I followed this up with diagnostics with Diagbox and saw even more extreme variation of the voltages of those cells, albeit different cells to the onces observed in Canion, but still in the same module:

View attachment 138796

It is worth noting that the system is measuring a total voltage across the module of 32.505V and that the sum of the individual cell voltages is somewhat different to that at 5.56V less. It really tells me that the problem lies in the measurement of the individual cell voltages and that the cells are operating normally.

So I had total confidence that the LTC6802 chip on CMU11 needed replacing. I already had two of these chips that I bought in case I needed to make this repair.

I put the car up on blocks and disconnected the 12V battery, removed the seats and took out the service plug. I removed both the covers on the underside of the car and opened the HV cable connection ports to check the voltage was 0V which it was. Then I disconnected all the cables and used the threaded bar method to lower the battery. I managed all this in around two and a half hours. Opening the battery was quick with my impact driver and then I looked for CMU11 using the following diagram (thanks Kiev):

View attachment 138797

I actually already had the CMU numbers written on the modules from my last repair but wanted to be sure so I put the battery on the right hand side of the car and jacked it up so that I could plug in the comms leads to the car and make Diagbox diagnostics on it to check that when I disconnected CMU11, that did indeed correspond to the non-communicating CMU in Diagbox:

View attachment 138799

Next I removed CMU11 and took out the main PCB. I used Chip Quik low melting point solder alloy and flux to remove the 44 lead SSOP package of the LTC6802 chip and then cleaned up the area before soldering the new one in place taking great care of the orientation and alignment with the pads. Note there is a conformal coating on the PCB that is a bit annoying, but it peels off OK and is not too bad.

I then put the PCB back in the CMU11 module and returned that to the battery pack. The pack was still connected to the car so I performed Diagbox diagnostics and saw correct voltage measurement in CMU11:

View attachment 138801

Now there is only a 60mV difference which is a reasonable margin or error.

So, confident of success, I put the lid back on the battery, battery back under car and raised using the threaded bar method. Then I basically reassembled / reconnected everything I had previously disassembled. Seats back in, error codes cleared with Diagbox and car back on the road working well again.

I don't know why my car has now had two of these failures. My last intervention had nothing to do with CMU11 so I think just bad luck. Maybe a dodgy batch of LTC6802 chips were used in the assembly. At least I can now repair it for €30 (just the cost of one of these chips). Most of the repair is the labour of disassembling and reassembling the car although the diagnostics are an important part of the repair without which I wouldn't know what to do. Total repair time was two half days.
Fantastic piece of work😁
 
1 - 3 of 3 Posts
Top