Getting more life out of a lithium ion bike battery

[galvitron]

I have been measuring the voltage at the charge port. I didn't think of the lower terminals which are much easier to get the voltmeter leads on! I don't have the RAD charger available at the moment (currently at my office) so I ran the test using the Luna charger- I may repeat RAD charger later.

I charged to both 80% and later 100% and after letting the battery rest for 15 minutes, measured the voltage at each location with the key in each position. Here are the results..

80% charge, key in 'off' position:
Luna display=52.8V
Charge port=29.6V
Lower terminals=27.9V

80% charge, key in 'on' position:
Luna display=52.8V
Charge port=41.4V
Lower terminals=51.7V

100% charge, key in 'off' position:
Luna display=54.6V
Charge port=31.2V
Lower terminals=29.4V

100% charge, key in 'on' position:
Luna display 54.6V
Charge port=43.4V
Lower terminals=53.9V

Conclusion: In each case the voltage reading at the charge port is lower which I assume is causing the Luna charger to cycle on and off regardless of key position. I wonder if these voltages are typical, especially for folks who had success charging with the key in the 'on' position.

It does appear there is some other circuitry that causes the voltage to read lower than the actual battery voltage at each location (especially at the charge port).

The battery performance seems fine, but perhaps a load test would give more information. Maybe time to reach out to Luna.

[Veggyhed]

Thanks for putting the time in for testing. I knew it had to be some sort of issue with additional circuitry. However looking at those numbers that's kind of crazy and they're all over the place.

Sent from my Pixel 3a using Tapatalk

[DickB]

The Rad battery charge port is connected to the battery through a semiconductor switch turned on by the Rad BMS. If the BMS is not seeing charge voltage at the charge port, it won't turn on the charge switch. The low voltage that you're seeing is due to leakage current. ("Leakage" in this context is an electrical term and not a description of a defect.) If you put just a bit of resistance across the charge port terminals, you'll see the voltage drop dramatically, because there is insufficient current to support it.

When the key switch is turned on, the BMS turns on the discharge semiconductor switch, which supplies power to the battery connector and the LED battery pack meter.

The key switch does not directly make or break a connection to the battery; the semiconductor switches do that. The key switch signals the BMS.

More information on the Rad BMS:
https://www.radowners.com/index.php?topic=1025.0

[DickB]

I've been doing further reverse engineering of the Rad battery, and I have a correction to make to one of my earlier statements.  The Rad battery charge port is connected to the battery through a semiconductor switch (a MOSFET) as I stated, but I do not believe that the switch is controlled by the BMS chip. (For complete reverse engineering, I need access to the bottom circuit board layer, and for that I would need to disassemble my only Rad battery, which I am not about to do.) It is turned on when a live charger is connected to the charge port, which I suspect is enabled by additional circuitry outside of the BMS chip.

The BMS chip is always powered on. With no activity, it is in a "sleep" mode drawing very little current.  But the BMS is always active.

Notice that the Rad battery charge circuit has protection diodes, which prevent current from flowing out of the battery into the charger.  Luna's concern about inrush current from battery to charger if the battery is plugged in before the charger does not apply to the Rad battery.  I called Luna out on this, when they incorrectly advised a Rad owner, and got this response:
"Okay well that's how our chargers are. I was answering the question based on products we sell."
Apparently, Luna is only guessing how the Rad system works.

(If I were designing the Luna charger and were concerned about inrush current, I would include protection diodes in the charger.)

A very small amount of leakage current can flow in the reverse direction through the diodes and charge MOSFET, which is why you can see a voltage at the charge port. The voltage that you see will depend upon the input impedance of your multimeter. If you place a resistor across the charge port terminals, even a relatively high resistance, you will see the charge port voltage drop because of the tiny reverse current flow. The bottom line is that you cannot use the Rad charge port to read battery voltage with a multimeter unless the MOSFET is active and charge current is flowing.

The power MOSFETs, which supply power to the motor controller port, are only turned on when the key switch is turned on.  This changes the leakage current situation, hence the change in voltage seen at the charge port with your multimeter. Some Rad models support regenerative breaking - in that case, the batteries would be charged through the motor controller port and not through the Rad charge port circuitry.

The Rad charger never disables current to the battery.  The Rad charge LED turns green when current out of the charger drops below about 200 mA, but current still flows.  I have seen 3 mA flowing even after many hours. I suspect this current is necessary to allow the BMS chip to balance the pack, which is why Rad advises leaving the charger connected for 12 hours periodically to balance the pack. If the Luna charger shuts off current to the battery when it indicates charge complete, battery balancing may not be taking place. I would be interested to know if the Luna charger does or does not continue to supply this apparently necessary small current on charge complete.