Hello,
I've been doing some testing recently with the TPS61220 converter and I've noticed an interesting issue that isn't specifically mentioned in the component's datasheet. I apologize, this post will get a little long due to extra background information. Questions are in bold for convenience. The converter was designed to output 5.5V. My testing included three different operating scenarios that the converter would see in my application:
1) Functionality of the device when VIN and EN are tied together when no voltage is applied to the VOUT pin (the output of the TPS61220 is connected to a storage reservoir that may have a voltage as high as 5.5V).
- The result of this test was what would be typically expected with the TPS61220. The converter began operating when VIN was between 0.6V and 0.7V and worked perfectly, outputting 5.5V to my storage reservoir and load.
2) Functionality of the device when the EN line is grounded and VOUT doesn’t have an externally applied voltage.
- This resulted in the converter not switching, which was expected since the EN line is grounded, and the input voltage passed through to VOUT through the internal pass through diode of the TPS61220.
3) And then my final operating scenario, functionality of the converter when the EN line is grounded and VOUT DOES have an externally applied voltage of at least 5 V from storage reservoir.
- This result is where my question lies. As VIN was slowly increased from 0V to the converter's start-up voltage, the L pin of the TPS61220 IC began to sink a large amount of current, as much as 200 mA. This occurred when VIN was between 0.4V and 0.5V. The output voltage of the converter was not changing, and no current was passing from the VOUT pin to the storage, so I don't believe any switching was occurring yet. However, the sinking of current at startup is definitely not desired.
Here is my guess as to what is happening: Based on the functional block diagram in the datasheet, it appears as if the Gate Driver is allowing the gate voltage of the low side drive NMOS FET to float, which is allowing the current to sink into the L pin, through the LSD NMOS FET, to ground. Does that seem like a possible explanation? Link to datasheet: http://focus.ti.com/lit/ds/symlink/tps61220.pdf See page 4 for the block diagram.
This sinking of current occurs until VIN exceeds ~0.5V, and then the converter begins to operate as expected once VIN exceeds 0.6V, outputting 5.5V as desired. I've read through the datasheet for the TPS61220 trying to get an answer to why this occurs but I haven't found much of anything. Another interesting tidbit, this hardly ever occurs when using a brand new TPS61220 IC, but almost always does after using the converter a handful of times. Could this be related to a burn-in issue? There has been one occurrence where this does happen with a brand new IC, but not typically. I’m not violating any of the recommended maximum ratings with the converter circuit, so I don’t think its anything in particular that I’m doing that is damaging the IC, but I suppose I could be wrong.
Ultimately, is this result something that should be expected when an external voltage is applied to the VOUT pin of the IC and the converter has yet to begin its normal operation? The component’s datasheet doesn’t mention anything specific towards the issue, such as using a series isolation diode between the converter’s output pin and the storage reservoir, but perhaps it's use is inherently implied?
Sorry for the lengthy post. I appreciate any thoughts you might have on the issue.
Scott
