Pretty interesting project, thanks for sharing! I have some suggestions about BOM.. In my project I also need to measure current and voltage, and I also don't want to calibrate each unit. But I think INA226 is pretty much expensive (around 1$) and I personally had problems with it (not sure if copies or non-working units). So I decided to use a voltage reference instead. e.g. MCP1525 is about 0.2$ while providing 2.5V and 4.096V reference voltage with 1% accuracy. And for the shunt I use an OP AMP MCP602 (also about 0.2$).
It's a very interesting design. It reminds me of the diyBMS by stuartpittaway but by going wireless you eliminate the need for isolation chips. Why did you choose to measure the current as well. Many designs simply use a voltage divider for getting the voltage and don't do anything about the balancing current. (How's the ADC on the ESP32?) With the balancing current, one can probably do more advanced battery health statistics... With the 4 identical groups of 100 ohm resistors, you will be able to adjust the current consumption in 4 steps. Did you consider the use of different values for each group so that like a DAC, different resistors values would create different currents and when adding all of them together, you would be able to have more than 4 steps. (4 bits = 16 steps). This might make it more versatile for different capacity cells. For testing the INA226 current sensor, did you use the circuit you designed? or just a different test circuit. With 16 100ohm resistors in parallel, and 4.6V, the math does not add up to 1.7A.
Thank you for your questions. For the non balancing mode I want the design as low powered as possible (sure there is the esp32) but this is not possible with a voltage divider with resistors in the kOhm range (maybe with a MOhm) but every ESP32 with the ADC then have to be calibrated in some way (don't have done a accuracy measurement, but at first glance every esp32 seams to differ slightly). Want to keep the total BOM below $10 without having a wired solution. Think the current measurements are also useful for keeping track of the charge an discharge cycles of the cells (you can also calculate the internal resistance of the cells). But I am open for other opinions. The 100 Ohm are just written values, want to test a variety of values 1 Ohm to 100 Ohms. You a absolute right, the Output of 1.7A is not correct. Have populated just four 100 Ohm resistors in parallel so the total value should be 25 Ohm (+R DS(ON) of the MOSFET) and the current have to be in the range of 184mA (at 4,6V) not Amps. The sketch need to be updated: // Calibrate INA226. Rshunt = 0.01 ohm, Max excepted current = 4A ina.calibrate(0.01, 4); have to be replaced, in my case I use a 0.1 ohm shunt. // Calibrate INA226. Rshunt = 0.1 ohm, Max excepted current = 4A (maybe more) ina.calibrate(0.1, 4);
Hi, there are several approaches to do this. You can use a AP with a Webserver, BlueFi (ESPBluFi) or ESPNOW and store the credentials in NVS/Flash storage.
![Li-ion Battery Charging Circuit Design - KiCad 5 [QCB 2]](http://i.ytimg.com/vi/GRd9uTwg7r4/mqdefault.jpg)
Good to see you back, hope everything is ok with your eye surgery!
Looks good. Missed your signature diction. By coincidence, found the XC6220 last week as well :-).
Pretty interesting project, thanks for sharing! I have some suggestions about BOM..
In my project I also need to measure current and voltage, and I also don't want to calibrate each unit. But I think INA226 is pretty much expensive (around 1$) and I personally had problems with it (not sure if copies or non-working units). So I decided to use a voltage reference instead. e.g. MCP1525 is about 0.2$ while providing 2.5V and 4.096V reference voltage with 1% accuracy. And for the shunt I use an OP AMP MCP602 (also about 0.2$).
Thanks for sharing your Knowledge and get well for your eyes!
Good project but a huge negative point is that you didn't even demonstrated any physical connection..
Hope your sight is OK now. Best wishes Arthur
Thanks, my sight is now about 90 to 95%. Can check for new glasses in some weeks.
Super byłoby móc kupić to do baterii, pakietów rowerowych na napięcie około 48v a nawet 60 v. Gdyby była taka możliwość proszę o kontakt
It's a very interesting design. It reminds me of the diyBMS by stuartpittaway but by going wireless you eliminate the need for isolation chips. Why did you choose to measure the current as well. Many designs simply use a voltage divider for getting the voltage and don't do anything about the balancing current. (How's the ADC on the ESP32?) With the balancing current, one can probably do more advanced battery health statistics...
With the 4 identical groups of 100 ohm resistors, you will be able to adjust the current consumption in 4 steps. Did you consider the use of different values for each group so that like a DAC, different resistors values would create different currents and when adding all of them together, you would be able to have more than 4 steps. (4 bits = 16 steps). This might make it more versatile for different capacity cells.
For testing the INA226 current sensor, did you use the circuit you designed? or just a different test circuit. With 16 100ohm resistors in parallel, and 4.6V, the math does not add up to 1.7A.
Thank you for your questions.
For the non balancing mode I want the design as low powered as possible (sure there is the esp32) but this is not possible with a voltage divider with resistors in the kOhm range (maybe with a MOhm) but every ESP32 with the ADC then have to be calibrated in some way (don't have done a accuracy measurement, but at first glance every esp32 seams to differ slightly). Want to keep the total BOM below $10 without having a wired solution. Think the current measurements are also useful for keeping track of the charge an discharge cycles of the cells (you can also calculate the internal resistance of the cells). But I am open for other opinions.
The 100 Ohm are just written values, want to test a variety of values 1 Ohm to 100 Ohms.
You a absolute right, the Output of 1.7A is not correct. Have populated just four 100 Ohm resistors in parallel so the total value should be 25 Ohm (+R DS(ON) of the MOSFET) and the current have to be in the range of 184mA (at 4,6V) not Amps.
The sketch need to be updated:
// Calibrate INA226. Rshunt = 0.01 ohm, Max excepted current = 4A
ina.calibrate(0.01, 4);
have to be replaced, in my case I use a 0.1 ohm shunt.
// Calibrate INA226. Rshunt = 0.1 ohm, Max excepted current = 4A (maybe more)
ina.calibrate(0.1, 4);
That's quite a surgery you got there. Get well soon!
Thanks, till now everything is going very well.
Hi, how can I update wifi credentials without updating code? In esp32
Hi, there are several approaches to do this. You can use a AP with a Webserver, BlueFi (ESPBluFi) or ESPNOW and store the credentials in NVS/Flash storage.