Calibrating a chinese charger with external components

[Laughwvst]

Hi!

I came across an interesting idea where I have multiple chinese chargers (Imax B6 clone), which are overcharging individual cells to about 4,3 volts, while they still read 4,2 volts on the screen. I have done very extensive online research and concluded that it is not worth it trying to reprogram the MCU, because it is simply too much work to even correctly identify it. There is no software calibration.

Then I came across an online post, where one guy had a problem with his charger and the problem was, that he was using a cable with voltage drop of 0,2 volts, so the charger was assuming the cells are full, when in reality they were only at 4,00 volts.

I suppose this is some sort of a beginner problem, because of course I could just attach huge resistors in front of every cable to drop the voltage, but that would firstly emit too much heat and secondly wouldn't be that neat of a solution.

Let's say that one would want to create an "external calibration kit" to these bad chinese chargers so that no one would even have to change anything inside. Would there be any reasonable solutions to create such a circuit that would actually drop the voltage that is seen by the charger, while still not emitting too much heat and would still charge the battery pack correctly?

- Beni

[gcewing]

It sounds like you want some kind of buck converter, but I don't know about the feasibility of using one to drop the voltage by just a fraction of a volt.

Stepping back a bit, it seems like the correct solution to this problem is to not use cheap Chinese chargers. There are other reasons for avoiding them too, such as safety. See Big Clive's channel for some thoughts on that.

[magic]

Then I came across an online post, where one guy had a problem with his charger and the problem was, that he was using a cable with voltage drop of 0,2 volts, so the charger was assuming the cells are full, when in reality they were only at 4,00 volts.

I suppose this is some sort of a beginner problem, because of course I could just attach huge resistors in front of every cable to drop the voltage, but that would firstly emit too much heat and secondly wouldn't be that neat of a solution.

Highly dubious.
Charging current decreases to nearly zero at the end of the charge. Therefore voltage drop across lossy cable or resistor also decreases to almost zero and the cell will go up to 4.3V as usual.

If you want to fix it, you will have to reverse engineer the control circuitry.

And there's the question if it's worth it. Li-ion chargers are aplenty.

edit
I mean, OK, you could perhaps build some circuit which drops exactly 0.1V regardless of current using a series FET and some opamp, but again, is it really worth it?

[Vovk_Z]

I think TS can buy chinese 5 digit voltmeter LED panel meter. They are accurate enough (better then 0.05% precision). So he can calibrate not only charger but his DMM too.
For example, This one.

[tunk]

If you don't mind undercharging, then you could use a ~0.2V Schottky diode in series.

[Rick Law]

Hi!

I came across an interesting idea where I have multiple chinese chargers (Imax B6 clone), which are overcharging individual cells to about 4,3 volts, while they still read 4,2 volts on the screen. I have done very extensive online research and concluded that it is not worth it trying to reprogram the MCU, because it is simply too much work to even correctly identify it. There is no software calibration.
...

How did you determined that it was your charger's being 0.1V under?  It could be whatever you use to measure your charged cell is measuring 0.1V over thus giving you the perception that is was charged to 0.1V over.

I assume you used a DMM to measure and you got 4.3V?  DMM giving you +-0.1V is not uncommon, so you have to make sure that your DMM have that level of accuracy and it is in spec.  Before you do anything to fix the problem, you best make sure you are fixing the right guy.

What make/model is your DMM?  How old is it?  Was it kept right and have reasonably fresh battery...   May be you need a new DMM instead.

[james_s]

A much better solution IMHO is scrap it and buy a charger that works properly. LiPo batteries can burn your house down, several people on the RC forum I'm on have had house, garage and vehicle fires caused by defective or improperly used chargers, it's just not worth it.

[Laughwvst]

I am completely aware of the dangers lithium batteries can cause and also that it is much better choice to just use a better charger. They are not even that expensive. I just became interested in Chinese chargers and this sort of became a little project of mine, or investigation at least. I would never charge batteries unattended with these cheap chargers (Actually not with any other charger either.) and I am always charging them in a small room basically made out of concrete and have fire equipment nearby. You guys don't have to worry about my safety, although it is really nice you are thinking safety first. 

The meter I use mainly is a UNI-T 139C and it has been measuring quite accurately when comparing to the bench meter at my school, so I wouldn't consider the way of measurement a problem here. I have some other cheaper meters also around the house and two of them showed exactly the same voltage. I measured the whole battery pack voltage and divided it by the amount of cells and also confirmed the measurement by measuring the individual cell voltages through the balance lead. They add up correctly and result in all of the cells being 4,28 - 4,3 volts.

I also know this is not worth it, it is a 15€ charger, so of course it is not. This is solely a project and training for me to understand if there are any ways to simply affect the cell voltage the charger is seeing and I am sure there might even sometimes be faults that lead to this.

The Schottky diode approach sounds quite interesting and I am definitely going to try that as soon as I get myself some of those.

I also found from the charger settings that it is able to charge either Lipos, Li-ions and LiFePO4 batteries and by those it means that the nominal voltage is either 3,7V (Lipo), 3,6V (Li-Ion) or 3,3V (LiFe) per cell. Do these batteries have otherwise exactly the same charging characteristics except the cell voltage? I could just for example drop it to LiFe charge and see how much under it actually charges the lithium ion cells (which I only need to charge) if the cells are charged exactly the same way but with a lower voltage. If the charge curves are not exactly the same, then I am not going to use that method because the goal is also to have it charge Li-Ions exactly correctly. At least in theory, keeping in mind that it is a 15€ Chinese charger. 

I would always rather undercharge my batteries to maybe 4,1 volts and I definitely wouldn't mind that. It is always better for the cells.

- Beni

[Rick Law]

...
I just became interested in Chinese chargers and this sort of became a little project of mine, or investigation at least.
...

If you don't mind undercharging, then you could use a ~0.2V Schottky diode in series.

If you are going to go with a quick and dirty simple voltage drop, I would jump on Tunk's bang wagon - Tunk's recommendation absolutely works.  In fact, when I use my CCCV ("constant" current / "constant" voltage) power supply to charge, I always connect it via a Schottky diode.  I use my CCCV to charge as often as I do the regular charger.  The regular charger is a family-charger so often it is already busy.

Since my CCCV psu is adjustable, the Schottky is not for voltage dropping - it is to prevent back-flow.  I am not consistent in my "done charging" process (habit).  Sometimes, I turn the psu off first before removing the battery.  Current may flow from the battery back to the psu.  Since I am not sure how well my psu like current being stuffed back into it, I use the Schottky diode to prevent that.

If you are going to experiment (educational), a quick and dirty voltage drop will not educate much.  I am not sure that is a good place to start.  A more accessible (there by less accident-prone) open-board adjustable CCCV PSU may be a better place to start.

EDIT: added the missing word "adjustable" in the last line.  Non-adjustable CCCV wont be fitting here.

[amyk]

If you're willing to spend the effort to make something I would rather spend the effort to reverse-engineer the circuitry and figure out what it's using for a voltage reference instead.

[Laughwvst]

I think how I will continue is that I'll get a hold of a 6S pack (the highest amount of cells that can be connected to the charger) and see how it will balance the individual cells. Currently I only have 3S and 2S packs, so I am not able to test all of the different circuits that are balancing the cells. I want to find out that is it actually seeing all the balance voltages wrong, or is there maybe a fault in some single component. (I suppose some single resistor messes all up)

Dave actually did a teardown of this Turnigy Accucell 6, which is basically very similar to these Imax B6 clones I have and even showed the schematic of it. However, when comparing my charger to the original Imax b6, it seems that where it uses a multiplexer chip to see the balance voltages, this charger doesn't have that kind of dedicated chips anywhere. All there really is on the board are some transistors, resistors and caps and then the microcontroller, so actually this might even be quite easy to reverse engineer as long as I don't have to know anything about the microcontroller itself. It has no labels and people have been discussing about probably this same mystery chip in some forums from 2016 and basically they have had no success with communicating with it or understanding all of the pins.

There seems to be 6 similar voltage divider looking setups right next to microcontroller, which are labeled "2C1, 2C2, 2C3, 2C4, 2C5 and 2C6" so that might be a good place to start. There also seems to be a trend where actually the main output of the charger is out of calibration, but the balance ports override that port, so I have to test how this feature works on this charger before I am going to have any better guesses.

Thanks for helping guys. I'm going to give an update if I understand something or if I don't. 

- Beni

[Rick Law]

I think how I will continue is that I'll get a hold of a 6S pack (the highest amount of cells that can be connected to the charger) and see how it will balance the individual cells. Currently I only have 3S and 2S packs, so I am not able to test all of the different circuits that are balancing the cells. I want to find out that is it actually seeing all the balance voltages wrong, or is there maybe a fault in some single component. (I suppose some single resistor messes all up)
...
...
- Beni

Oh, goodness, you are brave.  I wouldn't mess around with a 6S.  Particularly if you are experimenting.  Dangerous as heck.

Well, your life, your house...

I resign from this thread.  I hate to see Darwin at work.