Expert calibration

[Валери Цветков]

Hello,
i am trying to calibrate my charger because when i charge a 1.2V battery there is almost 100mV difference. For example, the voltmeter shows 1.453V, and the charger 1.545V.
I calibrated pin 0,1 and 2 without problem with 1.5V battery.
I also calibrated the "+" terminal.
But when I connect the battery between the "-" terminal and GND the battery drops to 0.8V, the cables heat up.. it's like I shorted it, and the charger (on the "-" terminal) shows 0.257V.
Am I misunderstanding how to calibrate the + and - terminals or is there a problem with the charger?

[greg...@gmail.com]

I think the Microcontroller has 8 bit A/D, so voltage is only detected in relatively big increments. A better microcontroller, with more A/D resolution (like 12 bit) would give finer increments

[diggit]

valeri:

Hi, sorry, I completely missed this thread.

Did you follow this procedure?

Which charger do you have?

Can you take a photo of the PCB? There are chargers with simplified balancer circuit and full balancer circuit. These require spefiy FW and it is easy to mix them up.

Which FW exactly you flashed?

Nonetheless, absolute voltage on battery terminals should not matter for Nixx cells, as these are usually stopped using -dv/dt ("delta peak") principle.

Battery connecting circuit looks typically like this. Q22 should be open when battery- calibration is done.

greg:

Please check the facts before stating something. Atmega32 has 10 bit ADC, ARM MCUs (Nuvoton 0517,...) typically have 12 ADC. Cheali is using oversampling with noise injection effectively increasing resolution: https://github.com/stawel/cheali-charger/blob/master/docs/artificial_adc_noise.md ADC resolution is not the isse.

Anyway thanks for bringing this thread to my attention ;)