Effect of Charger wire length and an optimum charge strategy

[Samson]

Over the last several weeks I have built my first Ebike. It is a low power set up using a 250 Watt geared Hub motor and 14S 10 Ahr Headways with a Cell man BMS. I keep the bike in the garage but I have the charger inside the house so I can monitor the charging cycle easily. To do this I ran a 2 conductor #14 AWG cable from a wall box termination at the charger to the bike XLR connector. The length is 35 feet total run. I calculated the voltage drop over this run based on the wire resistance of 35 feet x 2 conductors = 70 feet and the bikes Schottky protection diode (reverse polarity protection) forward voltage drop to be as follows with various current draw:

Charge current to produce 3.65 volts per cell taking the bike diode drop and wire resistance into account:

Charge: 10 Amps 14 cells x 3.65 V +.50 V diode drop + 2.29 V wire drop at 10 amps = 53.9 VDC
Charge: 7 Amps 14 cells x 3.65 V +.47 V diode drop + 1.60 V wire drop at 7 amps = 53.2 VDC
Charge: 5 Amps 14 cells x 3.65 V +.45 V diode drop + 1.14 V wire drop at 5 amps = 52.7 VDC
Charge: 1 Amps 14 cells x 3.65 V +.40 V diode drop + .229 V wire drop at 1 amps = 51.7 VDC
Charge: 100 mA 14 cells x 3.65 V +.34 V diode drop + .0229 V wire drop at 100 mA = 51.1 VDC
edit: corrected 100 mA Calculation

The BMS temperature is monitored back in the house with a digital thermometer probe on the BMS balance resistors so I can infer what the BMS is doing.
From the above table at 5 Amps charge I need to terminate at 52.7 volts and then drop to 51.1 V for the rest of the charge cycle to allow the BMS to balance the pack for 1-2 hours or so. If I leave the charger at 52.7 V after the current drops off the BMS gets very hot (75 deg C) and cycles on and off with HVC. Dropping the charge voltage to 51.1 Volts prevents this and the BMS runs at about 12 deg C over ambient.

Does this charge strategy make sense or do you let the BMS run hot?

 

[dak664]

I wouldn't let it get hot. A battery charger would limit the battery voltage to a safe value but in your case the wire resistance interferes. A 4-wire connection with a small-gauge pair dedicated to sensing the voltage would eliminate the resistance, assuming your power supply has a remote sense input. Or an interesting project would be to add a negative series resistance that cancels the wire resistance.

 

[Samson]

I wouldn't let it get hot. A battery charger would limit the battery voltage to a safe value but in your case the wire resistance interferes. A 4-wire connection with a small-gauge pair dedicated to sensing the voltage would eliminate the resistance, assuming your power supply has a remote sense input. Or an interesting project would be to add a negative series resistance that cancels the wire resistance.

The four wire sens line approach is the best way technically but none of the power supplies I currently use to charge are set up with sens inputs and I would have to add the additional wiring and a different connector at the bike end. I have been thinking of building a charger from scratch and could add this feature I suppose. It may be easier though to sense the end of charge at 52.7V by the current roll off then drop to a lower voltage to balance. Seems to me a better solution would be for a properly designed BMS to look after the charge termination function at the bike end where it can monitor the battery voltage precisely and any old power supply could provide the bulk current.

 

[dnmun]

why put a diode in the current path? can't you use a normal polarized plug so it is always connected properly? never heard of this before.

the BMS gets hot normally. that is what it does to balance the pack. not familiar with the BMS that you use but they all have to shunt current around the cells that fill up and that produces heat when the current flows through the shunt resistor. no way around that.

why do you have to be concerned about the voltage drop out to the garage either? when the pack is balanced, there is only 60-100mA of current so there is no real delta voltage through the charging leads, and you would always wanna have more voltage applied than the 51.1V of the fully charged and balanced to 3.65V each pack.

 

[Samson]

why put a diode in the current path? can't you use a normal polarized plug so it is always connected properly? never heard of this before.





the BMS gets hot normally. that is what it does to balance the pack. not familiar with the BMS that you use but they all have to shunt current around the cells that fill up and that produces heat when the current flows through the shunt resistor. no way around that.



why do you have to be concerned about the voltage drop out to the garage either? when the pack is balanced, there is only 60-100mA of current so there is no real delta voltage through the charging leads, and you would always wanna have more voltage applied than the 51.1V of the fully charged and balanced to 3.65V each pack.

The diode prevents me from getting polarity wrong when I connect one of several chargers in the house. It also keeps the 50V off of the pins at the bike for safety.

The BMS gets much hotter if I use it with the charger at 52.7V as I mentioned.

The current stays high at end of charge do to the higher set voltage initially, reducing the voltage allows a relatively cool, low current balance. At 51.4 V the current drops to 60-70 mA the cells balance then the current drops to 0 as it should.