LM5170: IOUT changes depending on DIR when no current is flowing

Hi Linas,

Thank you for using LM5170.

Could you please double check if the voltage is directly measured between CSA and CSB? It also helps if you are able to sense the current through the Rcs. It converts to 0.76A based on your measurement which is non-trivial.

Thanks,

Yinsong

Hi, thank you for the quick reply.

Yes, the voltage is measured directly connecting a voltmeter on the sides of Rcs (R173). Here is the schematic if that helps:

We are using this chip to charge the 12V battery, and the accurate current sensing is important for us.

Hi Linas,

Sorry for the misunderstanding. I mean the voltage on the two pins of CSA and CSB not the Rcs. Just want to make sure it's the same on the board.

And where does the PM_CHRG_CUR connect to?

Thanks,

Yinsong

Hi Yinsong,

The voltage on CSA and CSB is in the range of 0 to 14 V. It goes to 12V battery, so the battery can be connected / disconnected / charging.

PM_CHRG_CUR goes to the ADC pin of microcontroller.

Hi Linas,

Thank you but I was talking about measuring the differential voltage between CSA and CSB directly at the pins of the IC instead of at the Rcs.

Have you measured the IOUT voltage directly instead of reading from the microcontroller after ADC?

Thanks,

Yinsong

Hi Yinsong,

I just measured directly on pins 35-36, differential voltage is 0.00 mV. I have four prototype PCBs, and all of them are acting the same.

Yes, please see the picture below. First channel is IOUT, second channel is CSA pin (battery is disconnected, EN=Low, Dir=High, SET=0).

Hi Linas,

Thank you for the information. First let me try on the EVM and I'll let you know.

Thanks,

Yinsong

Hi Linas,

Thank you for the patience. I do not see this issue on the EVM. Usually there can be more errors in boost mode (DIR low) than in buck mode (DIR high). Could you confirm that the issue happens in buck rather than boost?
Also, at low current/light load condition, please look it in the absolute error value rather than the relative error. And as the current gets higher the error will get smaller.

Thanks,

Yinsong

Hi Yinsong,

I confirm that the issue is in buck mode (DIR high).

I agree that going into high currents the absolute error gets smaller. But in Buck mode, my current range is 1A max. And the battery charged condition is in 5mA - 50mA range.

Hi Linas,

Thank you for the information.

Are you saying that the normal operation of the battery load in buck mode is 5mA to 50mA load current on the buck output and 1A max?

Thanks,

Yinsong

Hi Yinsong,

Yes, it takes several hours to charge the battery (buck mode), but once it is charged, it will consume very little in float charge mode.

I just did another test: cut the track that connects LM chip IOUT pin with microcontroller ADC pin, in case there was some current leakage. Unfortunately LM IOUT is still behaving in the same way.

Hi Linas,

I see. Thank you for the information.

I would suggest to increase the Rcs value so that your current sense signal can be high enough to overcome the error.

Thanks,

Yinsong