My battery monitor shows lower voltage when overkill charge path is opened.

[gtxltd]

I'm not sure why. The easy answer is capacitors inside the inverter/charger but who knows.

 

[gtxltd]

The important take away from this for all the beginners out there (including me) is that you can't disable the charging FETs without some ill effects on discharge. What I mean is there is a considerable amount of heat in the BMS if you have substantial discharge current with charge disabled.

 

[Gazoo]

The important take away from this for all the beginners out there (including me) is that you can't disable the charging FETs without some ill effects on discharge. What I mean is there is a considerable amount of heat in the BMS if you have substantial discharge current with charge disabled.

I thank you for your help and the warning.

Even though the unit has a 2000 watt inverter, initially I would only be using around 600 watts at most in the event of a power outage. It's actually a UPS. The UPS likes to float at 27.07 volts or a little less than 3.4 volts per cell. I am going to leave charging on. I want to see if I have any substantial capacity loss after a few months or so.

But it is nice to know I can turn charging off and discharge to a lower SOC and hold the cells there if I do notice capacity loss. It's also nice to know if using the inverter charging should be left on to go easy on the FET's.

 

[gtxltd]

Yes we should all leave charge enabled in BMS (except in emergency protection event). The charge source should be set up such that we never overcharge.

 

[Deleted member 23531]

Charge enable/disable gives unexpected results.
When I open the charge path my battery monitor shows 25.8 volts which is a surprise.
When I re-enable it the battery monitor shows 26.3 volts same as the android app indicates.

Where is the ground reference for the sensor (the hall effect sensor on the (+) lead)?

What I suspect is happening is the circuit is open and floating, but your ground is not floating "with it".

EDIT: a simple schematic of your wiring would show this, if you can draw one for us.

 

[John Frum]

Where is the ground reference for the sensor (the hall effect sensor on the (+) lead)?

What I suspect is happening is the circuit is open and floating, but your ground is not floating "with it".

EDIT: a simple schematic of your wiring would show this, if you can draw one for us.

key {
    nnnawg|NNN| = fused busbar position where nnn is the wire guage and NNN is the fuse rating
    nnnawg|UUU| = un-fused busbar position where nnn is the wire guage
    <-> = bi-directional current flow
    -> = uni-directional current flow
    <- = uni-directional current flow
    eg = equipement ground
}
system {
    pos {
        002awg|100|<->breaker<->class_t_fuse<->battery.pos
        002awg|100|->inverter.current.pos
        004awg|100|<-monitor.sensor<-converter.pos
        006awg|060|->bp_65.current->fuse_block.upstream.pos
    }
    neg {
        002awg|UUU|<->bms.current<->battery.neg
        002awg|UUU|<-inverter.current.neg
        004awg|UUU|->converter.neg
        006awg|UUU|<-fuse_block.upstream.neg
        006awg|UUU|<->inverter.eg
    }
}
fuse_block {
    upstream {
        pos
        neg
    }
    downstream {
        pos {
            016awg|001|->monitor.pos
            016awg|001|->inverter.signal.pos
        }
        neg {
            016awg|UUU|<-monitor.neg
            016awg|UUU|<-inverter.signal.neg
            016awg|UUU|<-bp_65.ground
        }
    }
}
battery {
    pos { cell_in_series<->... }
    neg { cell_in_series<->... }
}
inverter {
    current {
        pos
        neg
    }
    signal {
        pos
        neg
    }
    eg
}
monitor {
    sensor
    pos
    neg
}
converter {
    pos
    neg
}
bp_65 {
    current {
        in
        out
    }
    ground
    signal { * shorted
        pos
        neg
    }
}
bms {
    current {
        upstream
        downstream
    }
}

 

[Deleted member 23531]

Well, that's one way to share a schematic! I'll try to explain my theory anyway using this.

If I understand your problem earlier in this thread it is that "monitor" gives erroneous pack voltage readings when the charge path of the BMS is opened (meaning you use an app to control the MOSFETs in the BMS in some way). So the issue I think we're talking about is this "monitor":

004awg|100|<-monitor.sensor<-converter.pos
fuse_block {
pos : 016awg|001|->monitor.pos
neg : 016awg|UUU|<-monitor.neg
}
monitor { sensor, pos, neg }

This monitor is sensing the bus bar positive which is connected to the battery positive, but the negative bus bar has been "floated" by the FETs in the BMS, relative to the battery negative, when you disable the charge path. Basically "neg" of the monitor is connected on the far side of the BMS FETs but "monitor" is still connected directly to the battery. So it's sensing an open circuit which has nothing setting the voltage that it's sensing, except capacitance and, ultimately, parasitic or ESD devices in the BMS.

Voltage needs to be measured differentially, and that monitor is sensing V(sensor) - V(neg), but "neg" is no longer connected to the battery (-) even though "sensor" is still connected to the battery (+). So the difference you're seeing is "ground float" or "open ground". It will typically be limited to ~600mV since at that point ESD structures in the circuitry of the BMS will limit further excursion between the grounds, but it really depends on the circuit design of the BMS.