> Lawrence Rhodes wrote: > >> The charger chip that senses current and voltage is a nice >> feature. Can these chips be added to a charger like the B&W >> or maybe a variac bad boy.
Lee Hart wrote: > Most of these chips are designed for small low-voltage low- > current battery packs, as used in laptop computers etc. It's > not going to be easy making them work correctly with a large > EV pack. The shunt to measure current is often inside the > chip. This is simply too small for a big pack. On some, you > can use an external shunt, but they still assume relatively > low currents, and so designed for a shunt with a large voltage > drop, like 0.25v at full current. If you tried this charging > at (say) 25 amps, the shunt would dissipate P = 0.25v x 25a > 6.25 watts. This chip uses an exernal shunt. Yes, the voltage drop required is .25V, but that works out nicely to an available 0.01R shunt for 25A - so the 6.25W dissipation is not a problem. > The chips are usually powered from the pack > voltage, which is usually 12v or less. You have to provide > some kind of low voltage supply for the chip. This can be > complicated because some chips sense battery voltage via their > supply voltage. Not this one, the pack voltage sensing in independant from VCC. > Most have on-chip temperature sensors. This works fine for > small batteries, where the chip is mounted right inside the > battery pack. It does not work for EVs, where the chip is far > away from the batteries and so not at the same temperature. The bq2031 uses an external NTC thermistor, on the end of a wire if you like. > Finally, the algorithms designed into these chips are often > very specific, designed for one particular type of battery. In > some cases, I swear neither the chip manufacturer nor their > customers have any clue as to what good charging algorithms > are. For example, they are designed charge fast, without > regard to battery life. Or, they don't worry about charger > efficiency, and use linear regulation. Or they don't provide > delays for state-switching, so they abruptly switch modes due > to momentary glitches in AC line voltage or momentary loads on > the EV's battery (like opening a door so the dome light comes > on, so the DC/DC converter kicks in, loading the propulsion > pack. When you close the door, the sudden drop in propulsion > pack current falsely triggers a mode change from finish to > float). This one doesn't seem too bad (it is designed for lead acid, with the correct voltage/temperature compensation), in fact the Constant Current / Constant Voltage / Float profile looks similar to the Zivan - so, not ideal maybe but useable.. Regards, Evan.
