Hi
I'm designing a 12V SLA battery charger based on the LM5118. It's a buck-boost from 10-30V. I used Webench for the main circuit with an output current of 4A.
I added some circuitry to measure output current and do a switch from cycle charge to float charge, but this really sits after the main circuit and simply adjusts the feedback resistance from an open drain comparator.
When I started testing the circuit I discovered that with a 4A output the current limit can be sitting at 5/6A when in buck mode. This was more than I wanted so I changed Rsense from 15 to 20 mohm.
I used a DC electronic load to characterize the prototype for efficiency and current limit. I used typical operating points for example 10,15 20 and 30V and 1,2 and 3A load current.
On average I had 89-93% efficiency and the current limit range was from 2.1A(Vin=30V) to 3.5A(Vin = 15) and 2.4A(Vin = 10V). The output voltage switched based on a presentable output current. All seems fine.
I then made the first connection to a battery. Immediately the charger stated to "sing". (make an audible low frequency sound)
I started probing the circuit with a scope and noted the following,
The input voltage from my bench top power supply takes a dip, about 2.5V.I tried various other power supplies just to see if it was a input power supply issue. It does not appear to be.
I then viewed the input power supply dip and the LO MOSFET gate drive, they are synchronized. I also viewed the output current and the LO MOSFET gate drive. Here I can see a rise in output current up to about 3A. It seems that the LM5118 runs along fine for a "short" period of time and then the PWM becomes erratic (out of control) and eventually shuts down and then restarts.(By synchronized I mean the input voltage dip starts when the output current increases and the PWM stops at the current limit point to prevent failure.
The above behaviour sounds like the behaviour described in the current limit section of the datasheet.
There is mention of "In applications with low output inductance and high input voltage, the switch current may overshoot..."
So the question is why does the circuit appear to behave correctly with a resistive load, but not with a battery which I'm guessing is a low inductance load. It probably looks more like a capacitor.
Can you give me some pointers on how to resolve this.
Your help would be appreciated.
Thanks,
Andre