I am using the LT8611 for an MPPT solar charger design, based largely on a design from the following Analog article.
My schematic is attached below.
My issue is with the IMON output from the LT8611.
I am testing using a fixed 15Ohm load instead of a battery for the moment. I am measuring the output current with a current meter in series with the load and at the same time using the microcontroller's ADC to sample the IMON output. IMON net is very noisy, >100mV p-p of noise but I am overcoming this by sampling 64 times and taking the average. Also tried adding an RC circuit in series, frequency cutoff of ~2kHz, which didn't help too much. I am varying the output voltage and thus the output current using a digital potentiometer in the feedback path controlled using I2C.
I repeated a test 3 times over an output current range of 190mA to 660mA (my load would fry if I went much higher) and got the results in the excel attached below.
The datasheet suggests a perfectly linear relationship between the output current and IMON reading when the SYNC pin is pulled high, which it is in this test. As you can see in the excel graph there is a significant offset error as well as some very non-linear behaviour such as when increasing output current results in a reducing IMON value around the 450-650mA range.
To perform an MPPT algorithm effectively I need the IMON value to behave like the datasheet suggests it does.
Any ideas? Am i going to have to use an external instrumentation opamp?
Any help is greatly appreciated,
Has the Microcontroller ADC offset been considered in the voltage sampling ?
Has the Imon voltage read by the ADC of the Microcontroller been compared with a voltmeter ?