Post Go back to editing

LTC3300-1 inconsistent cell currents

Hi,  I am looking for some guidance to help me diagnose LTC3300-1 operational differences across PCB's.

As part of a startup company project I am designing an Energy Management System for off-grid mobile applications (boats, recreational vehicle, ...).  the system is based on LTC6810-1 with cell balancing using LTC3300-1.  The master system is designed to monitor / balance 4 cells in series - slaves can be added for each additional packs of four cells in series (LiFePo4).  Four prototypes have been built.

All is working great, except that the cell charge and discharge average current (nominal value of 2A) show significant variances between cells and boards: from 0.36A to 3.50A

The balance board layout has been copied from the DC2100B design with the same components. Only circuits for cells 3, 4, 5 and 6 are connected - cells 1 and 2 are grounded.

Transformer turns ratio = 1:2
R-sense_pri = 0.008 ohm
R-sense_sec = 0.016 ohm
R-tonp = 22kOhm
R-tons = 43kOhm

Here are the cell average discharge and charge current for the 4 boards:

Discharge cell
board # 1 2 3 4
1 2.75 2.27 2.52 2.41
2 2.55 2.30 3.50 2.27
3 2.46 2.21 3.40 2.29
4 0.36 2.23 3.25 2.30
Charge cell
board # 1 2 3 4
1 3.00 1.80 1.23 1.40
2 2.95 1.82 1.45 1.60
3 3.02 1.83 1.43 1.54
4 1.90 1.82 1.52 1.54

Issue 1: discharge current for cell #3 on boards 2, 3, 4 and cell#1/board1 is too high.  That might correspond to the max on-time set by R-tonp?
discharge current for cell#3 / board#1 is OK, so design must be OK.  It seems that the LTC3300-1 does not control the mosfet based on peak current measure / 50mv at R-sense_pri - defective LTC3300-1?

Issue 2: discharge current for cell#1 / board#4 is close to zero.  discharge of cell#1 for boards 1, 2, 3 is OK so board design is OK.  What could cause that?

Issue 3: charge current for cell#1 / boards # 1, 2, 3 is too high.  Same problem as issue 1, but on the secondary side?

Issue 4: charge current for cells #3 and 4 is a bit low, but I can live with it.

On the scope, voltage on sense resistors is quite noisy (ringing). But overall it is working OK so I guess it does not interfere with proper operation of LTC3300-1.

Considering the number of issues I would be surprised if they are related to defective components (except maybe for issue 3?) 

Any guidance to help me better diagnose the origin of these issues would be very much appreciated?  I can forward by email schematics and PCB layout if that can help.  Any additional measures I can do to help diagnose? 

Thank you

Phil