I am using the LTC4015 with plain satisfaction. More then 1000 boards are working and stable.
Solved/clarified all the doubts about this and that.
LTC4015 is working with 12 cells Lead_acid battery and in "hardware" mode (CHEM0 and CHEM1 are opened).
I use the telemetry to monitor the system and everything is fine and as expected.
The +5V for CHEM0:1 and CELLS0:2 are derived from the 24Voutput followed by a LM317 + LT1763-5 (the LTC4015 is a part of a system where the LM317 and LT1763-5 are available for other purposes. So, instead of use Int_Vcc I use a separated +5V to reduce the LTC4015 consumption). This chain works well because in the main project CELLS0:2 are detected rightly.
I am charging the batteries with 3.6A and the Isystem is over 27A. No EMI/RFI (passed with plane votes all the CE tests ... here in Italy).
I have a strange behavior with a new setup. I want now to manage the current charge and I change CHEM0:1 from Z Z to L H respectively.
At the startup of the firmware, after 3 seconds, when I have initialized other i2C chips as usual, I try to access the LTC4015 by a i2c_start(0xD0) and the LTC4015 i2C doesn't reply.
Seems that when I pull CHEM0:1 I block the LTC4015 i2C interface.
If I desolder the jumpers everything works fine. Tried on 5 prototypes. The story is always the same:
if I pull CHEM0 and CHEM1 the LTC4015 becomes inaccessibile by i2C interface.
You should not pull the CELLS pins to a voltage besides INTVCC. I'm not sure if this is the issue or not, but it is not allowed.
The CELLS pins are designed to be strapped directly to INTVCC for the 'H' setting. If you need a pull-up resistor to INTVCC, you can use a 3kOhm pull-up, but no higher.
Perhaps this is affecting the logic somehow, so fix that and the SMBus issue may go away?
let me try.
If the CHEM0 CHEM1 are in a kind of open collector your explanation has a sense.
But if this is true for CHEM0 and CHEM1 why CL0:2, pulled up at +5V are workinh well?only a doubt to stimulate the solution.
anyway I try your suggestion, I mean pull-up or down directly IntVcc
Pulling to 5V will work most of the time. However, there are things coupled within the IC that are expecting the CHEM0/CHEM1 pins (and CELLS pins for that matter) to be referenced to INTVCC. If there is a fluctuation on INTVCC and these pins are pulled to a constant 5V source, there could be unintended operation or even damage to the LTC4015.
Referring CHEM0:1 to IntVcc the LTC4015 is accesible again.
Discovered the matter. The +5V I generate outside the LTC4015 ar eactive with a delay imposed by me of 100mS. When the LTC4015 logic reads the CHEM0:1 the state is for sure not stable.
So LTC4015 complains. This is reasonable.
Now: RSNSB = 9mOhm, so in "hardware mode" the Icharge is 3.6A.
I want to charge the batteries at half the current set by the RSNSB, and I set ONLY 0x1A reg.
I left all the other registers at the default value.
Icharge = (ICHARGE_TARGET + 1) × 1mV/RSNSB
for Icharge 1.6A I calculate ICHARGE_TARGET = 13
MPPT is OFF, Equalize is not active, no NTC installed.
The charging current remains at 3.6A, so my set didn't take effect.
Do other register writes work? Double-check that you are set for a Lead-Acid chemistry. Other chemistries will not be affected by this if JEITA is enabled.