In this paper, a unique average output current monitor feature of National Semiconductor's LM5117 synchronous buck controller will be introduced and a method to use this feature in order to implement precise current limiting will be shown. An example of a constant current regulator providing output overvoltage protection will also be presented.
The current-monitor output signal in the LM5177 is a welcome new feature. It has a gain of 20x from the CS pins, which feature a Kelvin connection to eliminate errors from ground-pin current. It also appears to include a sampling feature to kill the spike from the mosfet's high Ciss-charging current (although on average we'd expect the Ciss charge-discharge currents to cancel).
But unfortunately the CS input amplifier has a typical offset of 25mV, and 120mV max, and no offset-drift spec (compare this to the respectable 1.5% voltage error-amplifier spec).
This large offset would seriously-degrade implementing any lab-style adjustable precision-current-limiting application, e.g. of the type we are used to seeing on bench power supplies. One would be forced to add their own precision amplifier to the sense resistor, and at that point the advantage accruing to the LM5117 from its unique monitor feature would be lost.
1 comment
write a commentWinfield Hill Posted Jul 6, 2011
The current-monitor output signal in the LM5177 is a welcome new feature. It has a gain of 20x from the CS pins, which feature a Kelvin connection to eliminate errors from ground-pin current. It also appears to include a sampling feature to kill the spike from the mosfet's high Ciss-charging current (although on average we'd expect the Ciss charge-discharge currents to cancel). But unfortunately the CS input amplifier has a typical offset of 25mV, and 120mV max, and no offset-drift spec (compare this to the respectable 1.5% voltage error-amplifier spec). This large offset would seriously-degrade implementing any lab-style adjustable precision-current-limiting application, e.g. of the type we are used to seeing on bench power supplies. One would be forced to add their own precision amplifier to the sense resistor, and at that point the advantage accruing to the LM5117 from its unique monitor feature would be lost.
reply