actually a simple and robust circuit out of China. I'm impressed, esp if it's at a good price point

@@pauls5745 They only cost about 12AUD for 10 incl shipping. Pretty damn cheap.

If the microcontroller is flashable it would be a trivial matter to reprogram it to do that even if you had to start from scratch after erasing it.(Assuming it has copy protect enabled) Actually, it would probably be easier to erase it, rather than trying to decode the assembler listing. Depends on what type of language they used to write the source code. You'd also have to rewire the red LEDs to be controlled by a spare pin on the MCU.

@@kimchristensen2175 In my experience, the sorts of chips used for these sorts of applications are usually mask ROM or OTP, not reprogrammable. Even if it was, though, you'd need to know all of the details of the instruction set and chip architecture, which are quite possibly very nonstandard. It's probably not feasible without getting assistance from the company who makes them (who likely aren't interested in talking to anybody who's going to be ordering less than a few million of them at a time). It also might well not even be a microcontroller at all.. It could well be just a circuit similar to the other comparator design, just with everything prepackaged into a convenient chip...

@@foogod4237 Could be. Looks like the pinout is close to one of the 12F or 10F series pics. I guess it could be swapped for one of those, but that's probably not worth the effort.

Can this be modified to work on even higher voltage batteries?

@@data9309 I'm sure you could. You'd have to add a higher voltage rated pre-regulator ahead of the existing regulator or replace it with one rated for a higher voltage. And you'd have to modify the resistive divider to scale it properly to the higher voltage.

Totally agree, and Clive's soothing voiceover is just made for sipping whisky at the same time, but this can lead to missing the finer points of 'sschematic' analysis.

I was just thinking the same thing - An A to D and a bar graph LED for £ 1. 5 , from Amzon UK and 1 resistor sounds useful

@@kit-geoffbullough8788 If you don't mind waiting they're only about 50p from AliExpress.

Well, looking at 8-cell, with it at 33.6V full charge its at 2.1V on the ADC input. So they picked the 2V reference, and it will already be saturated before it even hits full charge.

@@clynesnowtail1257 Mega, thanks for doing some maths for me, apparently I was too lazy at the time? 😂 I guess same goes for the single cell as well, right? 4.2V into a 10k / 10k divider would give 2.1V into the ADC, and same thing happens again.

Well, the issue with not using the full range is that you lose accuracy, and given that you'd want to be looking for something a bit over 4.2 V (single cell) for "saturated," you're then looking at covering 50% more range. I have always wondered how accurate these ADCs are; would using only two thirds of the full range for the "non-overloaded" levels make the accuracy significantly worse? Perhaps not, in an application like this....

@@Curt_Sampson Not sure I quite follow what you're saying. In the case of this module, the full range of the ADC is (probably) 0V - 2.048V (using an internal reference, called an FVR, inside the microcontroller). By using the FVR, you no longer have to worry about the supply voltage of the microcontroller affecting the ADC measurements. For a fully charged cell at 4.2V, being fed into a resistor divider made of two 10k resistors (the 1S setting on this module) you get 2.1V fed into the ADC. 2.1V > 2.048V so percentage wise, you're attempting to use 102% of the range. Due to how the ADCs are physically built in the silicon, this will be 'constrained' to 100% of the value. Thus, even if the cell was sitting at say 5V (grossly overcharged), the ADC would see 2.5V, but the digital conversion would still read the same as if the cell was actually at 4.2V (providing a 2.1V input) since, in both cases the ADC is saturated. Regarding accuracy, most micros of this type I'd expect to have at least an 8-bit ADC. 2.048V / 256 steps = 8mV/step. Scaling this value UP by the ratio of the resistor divider gives (for the 1S divider) 16mV resolution across the cell. Presumably that value is pretty much constant in terms of a 'per cell' error, assuming all cells are at least roughly balanced. EDIT: ADC saturation doesn't necessarily equate to device damage.

@@adzib1823 I think your 2.048V hypothesis is flawed. The natural output voltage of a band gap reference is about 1.22V. Any other reference voltage is likely to demand individual device trimming. Many cheap microcontrollers with ADCs, for example the ATtiny5/10, tie the internal reference input to Vcc and let the designer worry or not about a clean supply rail. PIC12F609 isn't much fancier, offering a 1.1V band gap and a 4-bit digipot between the supply rails, the latter borrowed from the comparator module. That's plenty when you have a regulated supply rail and don't need great accuracy, or you are using the common idiom of a pot between the rails as a physical control input.

My birthday as well. Happy birthday.

Happy birthday to you both.

@@bigclivedotcom Thank you Clive 😊

Thank you. It's greatly appreciated.

@@bigclivedotcom no worries. It is a pleasure to be able to support someone as cool as you. Your videos have taught me a lot.

You should be able to change the LED resistors without affecting the calibration.

There are many applications that use a resistor across a capacitor. It still has a "storage tank" effect to give a stable voltage.

@@bigclivedotcom I sense a but, there...

I've actually had one sitting here for years. But never really tested it properly.

It could be done entirely in software.

It will give an intermediate voltage threshold.

If used for brief tests on a balanced system you could tap it off just a few cells to get an indication of the entire pack's charge status.

No transistor needed. It could probably be done with the spare pin on the microcontroller.

@@bigclivedotcom Yes but you would have to re-program for that, vs a couple of discrete components...

I did make a video about the duracell battery indicator.

Theoretically, yes.

@@bigclivedotcom I have a 7 bay charger but it has an LCD display showing percentage . I have itchy fingers to get the screwdrivers out now 😂

They probably exist.

It would pose a slight current draw continuously.

Yes.

If it's that type then shorting the 7S link should work.

There are many cheap microcontrollers like Padauk chips in mass produced products.

@@bigclivedotcom thanks. As I mentioned I have a big project coming along and cheap microcontrollers might come in handy.