Lipo Rider Plus - battery slow discharge beyond 3V cut-off

Products & Technology
1

Hi, I have tried to post this in one of the older threads, but not have an answer yet.

If anyone is using Lipo Rider Plus or similar with a 5V boost output on usb-A or board pins: I have the following:

I’m using the Lipo Rider Plus with relatively small Li Ion/LiPO batteries (250-350 mAh) and Arduino Nano Matter connected to a 5V booster output on type A connector in my projects. I have also a a physical (latching) switch to turn the power on/off to Arduino after 5V output. All batteries I have, include integrated PCM/BMS modules (they all do) providing all sorts of protections. However, after a while batteries stop charging. On the board only 4th green LED blinks with 3 solid LEDs which hints on almost fully charged battery.

However, after testing the batteries - they all have discharged below 3V cut-off, when the circuit is left to rest for couple of days to a week and some even discharge to 0V completely, without any load - the switch is off. When this happens the battery “refuses to be charged” - when connected to usb-C external charger again, Lipo Rider board applies 4.2V at the battery pins but no/trickle current is flowing for charging. Which is not surprising as battery is now in over-discharged state and PCM stops it from charging.

My question, how is this possible (discharging below 3V safe cut-out) with no load connected (switch is off on Arduino and even if nothing is soldered to the output)? Battery protection circuit should stop discharge at 3V. How it is possible for Lipo Rider to slowly discharge it over time to below 3V? I tried many different batteries and I have 20 different Lipo Rider boards - they all do the same.

PS. My guess it is a 3V-5V booster has something to do with that as our internal toggle board switch for 5V is always on.

2

Since I don’t have a Lipo Rider Plus, I checked the circuit diagram and the ETA9740 datasheet. The datasheet lacks detailed operational descriptions, so the following is my speculation.
Even if no load is connected to USB-A, if the switch is ON, it generates 5V and supplies about 3mA to the 5V output enable indicator on the board. However, when the battery voltage drops below 3.2V, the BOOST mode should stop, and no current should be supplied from the battery. The phenomenon of the battery fully discharging is very strange.

What is the function of the built-in over-discharge protection circuit in the battery?

3

thanks. I’ve spent some time on this since then. So, if anyone has similar, here is some information.

Firstly, yes 5V buck-boost converter self-draw even without load is significant to discharge the permanently connected battery below cut-off within a week or so. We had tried to overcome this by re-soldering the external power switch to the 5V ON switch contacts on the board itself. While onboard switch stays at 3V the external enables the 5V output. It isolates boost more in off condition. It solved 95% of cases of failing batteries, but not all.

There is also an issue with “refusing to charge” batteries - no recovery mode. ETA9740 sees protection latch on battery PCM, i.e. open circuit, mis-recognises it as battery full at 4.2VDC and no actual charging occurs even over long time. When charger is disconnected, the battery reads still below 3V.

  1. There is still self-draw from Lipo Rider Plus ETA9740 (below 1mA range) with 3V always active and some other bits which could cause over-discharge below 3.0-2.8V with time if battery is permanently connected. There is of course also a self-discharge in the LiPO battery as well, before PCM barrier even.
  2. The ETA9740 itself has no hysteresis 3.0V cut-off and 3.0V enable for battery charging, so with PCM cut-off on battery also at 3.0V it can cause oscillations and instability.
  3. Small batteries like 350mA with max charging current of 380mA are not a good match for Lipo Rider with max 2A charging/discharging. I’m not sure how it handles the battery current with PCM together, but it might cause too harsh charging and no-recovery possibly. Some AI chats suggest resoldering ETA9740 R1 ISET resistor on pin 5 from 56kOhm to higher (e.g. 560kOhm) to limit charging current to match with PCM on battery, which might help with softer current and recovery of over-discharged batteries. This is not documented in the ETA9740, but I might try that, as there are not many alternatives to Lipo as I need also 5V out supply from a battery source.
  4. I’ve noticed that charging Lipo+Battery via USB port on PCs rather a wall plug charger or similar can provide more gentle charging and helping to reset the latched battery. Which points to possible issues of matching high current input on USB-C with high charging current limit of Lipo Rider with low max charging current of Li battery we need to use.
4

This is merely an update to this problem. So, if anyone is planning to use Lipo Rider plus on a gadget which needs bidirectional simultaneous USB-C charging and boosted supply (5v out) with a permanently connected small Li/LIPO battery of 200-400 mAh.

Apart of always on 5V boost draw issues which can be solved by adding an external switch to 3V/5V switch on the board, there are issues with charging conditions which cannot be tailored to specific application/battery capacity in out of the box configuration.

Lipo Rider Plus uses ETA9740 which has max charging current 3A/2A depending on PIN5 ISET R1 value, which is hard set on the board. Schematics show R1 between PIN5 to AGND of 56kOhm which would imply 3A not 2A (while 2A would need R1 = 91kOhm), but this is not the matter here. ETA9740 has also 200mA recovery current when Vbat is measured < 2.8V and 200 mA cutt-off charging current when battery is full. All this makes it a poor match for small capacity LiPO batteries and especially badly interfaces with PCM on these batteries.

Small LiPO batteries of 200-500 mAh would typically need 1C max, or 0.2C standard charging current, 0.1C pre-conditioning current when Vbattery is low (20-50 mAh). Some PCMs on these batteries will have 3.0-3.1 Vcut-off (discharged state) and expect to have a small preconditioning current to gently raise battery voltage above 3.1V or so when normal charging can resume. Board will try to put 200 mAh in this case as this is still above 2.8V, as there is no means to regulate the max charging and pre-conditioning currents or cut-off voltages on the board itself.

However, the main issue, once battery voltage drops below cut-off the PCM on battery effectively disconnects the battery from the circuit, here LiPO rider sees 4.1-4.2V on the battery pins when connected to USB-C charger and assumes it is fully/nearly charged (3 LEDs solid and 4th flashing). However, no noticeable charging happening, you disconnect the charged and battery still reads 3.0 V and no current in output pins. So, we have many “field users” to return the gadgets saying they no longer can charge them via USB-C port. Some who charge the battery regularly do not see any issues, others allow battery to discharge almost always end up with this state.

The only way to recover here is to remove the battery and charge it in a different conditions/board, replace with another battery or connect a second “healthy” battery in parallel, which would trigger PCM to open this battery for charging too. None of batterie we had, were actually damaged/faulty, almost all can be brough to a very good healthy state, once out of “protected mode”.

Messing up with R1 resistor (like cutting it off from PIN5 by breaking the track, fiddly) and adding another resistor in series with R1 to limit max current from 2A to 0.9-1C of battery is not really helpful. As it not addressing recovery mode.

The only solution we end up with as an additional circuit between LiPO rider and out battery which would limit max charging current to 0.9C when Vbatt > Vbatt low, and to 0.1C precondition when Vbatt < Vbatt low. And setting Vbatt low = Vcut-off (PCM) + 0.2V or so. Not tested this yet.

Ideally, I would like to have universal charging board with configurable values of these

Battery Pre-Condition Voltage 2.8 V

Pre-Condition Charge Current 200 mA

Fast Charge Current Riset=56K 3 A Riset=91K 2 A

Charge Termination Current 200 mA