LIPO flat battery charging module

[Veketti]

Dear All,

I've been wondering to step up the powering of self designed battery powered gadgets to use generic LiPo flat battery pack. Instead of designing any charger by myself due to these being dangerous, I've been looking for modules from Ali. Anyone have experiences with these modules? To my understanding it should work so that you can charge the connected battery pack through USB C connector and power your gadget through the OUT + and - and have all the protections for the LiPo. Just need combined boost & buck module if need only 3.3V or boost module if 5V.

Is my understanding correct? Is this good idea? Maybe better module suggestions? Or dangerous and stick with the AA batteries? 

[CountChocula]

The TP4056 is a popular single-cell charging IC used in a whole bunch of different devices. These little modules are usually pretty safe, particularly if you get the ones with a built-in protection circuit (typically using something like a DW01A) chip, which covers both overcharge and overdischarge). Here's from Big Clive on using the modules, and with some more information on do's and dont's.

Adding a dc/dc converter downstream of the battery pack/charger circuit is also pretty standard. You can even find modules that combine the two on a single PCB.


—CC

[ArdWar]

Is my understanding correct?

Yes, pretty much. It works okay as a charger.

Is this good idea?

It depends, DW01A allows you to discharge the battery way too deep for my liking. Seems it was optimized for e-cigs usage. 2.4V is fine if you're drawing significant current from the battery (the lower threshold is designed to compensate battery internal resistance), but a bit too low for low current discharge use.

[Peabody]

Most of the flat lithium polymer cells I've seen have the protection circuit built in.  It's a tiny circuit board at the wiring end of the cell, with usually two little I/Cs.   If yours are like that, you can use the TP4056 module that just has the charger.

[ArdWar]

Most of the flat lithium polymer cells I've seen have the protection circuit built in.  It's a tiny circuit board at the wiring end of the cell, with usually two little I/Cs.   It yours are like that, you can use the TP4056 module that just has the charger.

It... depends™

Most prismatic cells intended for commercial end user do have protection
Most prismatic cells where you buy bulk, or one intended for hobby/maker RC/drones usually don't have anything at all. The PCB is just a crude busbar if it even got one.

[Veketti]

So I ordered some combined LIPO charger and boost converter called LX-LCBST. I was planning to also read the LI-PO cell voltage using voltage divider by MCU. But it seems the Battery negative and output negative are not the same. Measured around 450ohm between them. So my understanding is that the voltage measurement then doesn't work with only the battery positive. What are the possibilities in this case to measure the voltage by MCU? I'm using STM32G070.

[Peabody]

The only difference between battery ground and circuit ground should be the on-resistance of the two mosfets used in the protection circuit.  Does your battery already have protection built in?  If it does, you might want to see if disabling the protection on the module will eliminate the difference you see.

Check out this video:



If he is right about the protection chip on the module being the HM5434, then the on-resistance per the datasheet is 50 milliohms, which is much lower than your 450 ohms.  But if you are measuring with the module powered down, that may explain the high resistance since the mosfets will be off.  But just try doing the short he describes and see if all that resistance goes away.

In any case, what you really want to measure is the voltage difference between the two grounds when the circuit is powered up and running.

[Veketti]

Thanks. Yes, I was measuring without power on the module. There are protection PCB's on the batteries so I soldered the "T" bridge on the module which seemed to bridge the output negative to the battery negative. Problem solved.

[Veketti]

Getting back to this. So I soldered the "T" connection to disable the protection as the flat batteries I have, were advertized to have protection. First full charge and the battery is completely dead. No voltage from the leads. Nothing visually wrong with the battery, not swollen or anything. I'm afraid to replace it with new one, maybe it does the same. I tried with second set and this time had current meter and voltage meter and the voltage rose 4.205V and still ~200mA charging current when I disconnected it to prevent damage.

So the guestion is, is this normal behaviour or something failed to work? Could it be that the protection circuit on battery is that overvoltage disconnects the battery? But how it then can be reset?

Thank you in advance.

[Peabody]

If the protection circuit has triggered for some reason, that should be released automatically when the reason for it has gone away.  But I'm not sure I understand what the problem is.  4.2V and 200mA sounds ok for charging.  Do you have any other way to try charging the dead battery?

[Veketti]

The original Lipo battery that it charged is dead. No voltage on leads.

I tried to say, I built second set and tested the charging. So new battery, new charging module and this time I measured voltage and charging current. I let it charge until the voltage rose up to 4.205V then disconnected as I was afraid it would ruin the new battery as well. How high can I let the voltage rise until the battery gets damaged? Either the protection works and cuts the charging or ends up in another dead battery.

I’ve got imax B6AC charger as well that I could test to charge the dead battery. But maybe it will not do anything as it doesn’t recognize the battery due to open circuit. Maybe try lab power to charge it?

[JustMeHere]

Check out the ESP32 Thing at Sparkfun.

https://www.sparkfun.com/products/13907

It has a battery charging circuit.

Here's the schematic: https://cdn.sparkfun.com/assets/learn_tutorials/5/0/7/esp32-thing-schematic.pdf

[Peabody]

The charger charges at whatever current the resistor sets, and the voltage gradually rises to 4.2V. Ater that it stays at 4.2V, and the charging current gradually drops as the battery reaches full charge.  When the charge current drops below 10% of the original full charge current, charging terminates. So the voltage should never go above 4.2V, plus or minus a percent or two.

You should look at the datasheet of the TP4056 to see how it works.

It sounds like the dead battery is really dead.  Maybe a connection on its little PC board is open.

[Veketti]

Dear All,

Maybe I can continue on this topic as it is about the LX-LCBST charging&boost module. So I utilized it in another device which was exactly the same pcb with STM32G070, but the sensor not being anymore the Winsen ZE07 CO sensor, but this time Winsen MH-Z14A CO2 sensor. Now every time the MH-Z14A takes reading the MCU keeps high pitch whining and in long run heats a bit. If I power the device through lab power, no noise and no heating issues. So it has to have something to do with the LX-LCBST module, but what? I then added electrolytic capacitors to the MH-Z14A power pins and to the LX-LCBST output pins. They were stupid big 180uF and 330uF. Now the noise is almost gone, but still I can hear it. Scoped the 3.3V rail on the pcb and it had high frequency 100mV peak to peak ripple when the sensor took readings (and this was when those caps were present).

According to MH-Z14A datasheet, it draws quite considerable amount of current. 125mA peak and average current 40mA. Compared to less than 20mA what the ZE07 sensor draw that didn’t cause any (noise) issues. LX-LCBST is rated for one amp so it should handle this. Is there a way to mitigate or prevent completely these (noise) issues? And why the MCU is audibly buzzing?

Thank you

[Peabody]

I think the noise is caused by the output ripple of the boost converter, which gets worse with increased output current.  Can you identify what boost converter IC is used in the LX-LCBST?  The typical boost chip is some variant of the MT3608.  But that chip oscillates at a fixed frequency of 1.2MHz, which you would not be able to hear.  Also, the datasheet for the MT3608 shows a typical application using a 4.7uH inductor, which is different from what's used on the LX-LCBST.  Can you use your scope to identify the ripple frequency?

[Veketti]

It has marking B628 so it is HX3608 or LN3608 or similar. So should have switching frequency of 1.2MHz. Inductor has marking 100, so is it then 100uH? Measured it in circuit and got 10.8uH. Measured the input and output capacitors and got around 11uF for both. Datasheet is recommending 22uF. I don't know how accurate it is when measured in circuit.. I'll measure the ripple frequency once I get to the scope.

[Peabody]

Yes, or SDB628.  So it's the generic 1.2MHz boost converter.  The inductor is probably 10uH.

I don't know why you get ripple at a much lower frequency.  If you have caps on both input and output, I don't think that ripple should happen.  So I don't know what's going on.