I mean a 200 ohm resistor in  serial a standard green LED to emulate a load. 

Hi David,

Could you please confirm the UVLO voltage? And the current for each channel is recommended not to exceed 200mA. If you have only one LED string, I think you can combine channels to drive this string.

When you powering on the device, will VDD have any voltage measured? And have you supplied PWM to the IC?

BR

Patrick

Hi Patrick

Thank you for your prompt reply, the UVLO pin is 3.09V. the Vdd is 0. I am wondering we have burned the driver. in this case is there any way to confirm if the chip is still good or not?

Thank you,

David 

Hi Patrick, 

If the EN pin is low, will the VDD output 6V or 0V?

Thank you,

David

Vdd is 6V after enable pin turns high.

Hi Patrick,

On the demo board, when I test it,  I only linked in one LED chain on LED1.  and jumpers on PIN1 and PIN2 of  JP1, JP3 to JP7.  which means that even the 27K resistor not linked in. the controller still can remove or disable the unconnected channels. 

I assume this will happen on my board, so I only linked in one chain, and left rest of them floating. some of them even the PWMs are floating( there are in chip pull down). 

if these 27K resistor are critical, I need to put them in. 

now assume that we have VDD, EN, and PWM are correct. but there is no switch signal to boost the voltage up.  what is the best strategy to diagnosis from here. 

Thank you,

David

Hi Patrick

see attached waveform, yellow is the fault, pink is IFB4 which has a LED chain connected. 

based on the data sheet, after the enable( falling edge of the fault signal). the IFBx will start a process to detect the linked LEDs. seems in my case it failed at this step. 

the LED chain I have tried on the 

TPS61196 demo board, which works fine. by the way, I have updated the demo board with the resistor to set it to 250mA. I need some help to figure out why the TPS61196 think there is no channel is used( I do have one connected). by the way, how do I know the fault is triggered by which channel. let's say if the fault is triggered by channel 3. and I have a channel 4 connected. will the chip still drive the switch to boost the voltage?

What should I do next?

Thank you,

David

sorry, one correction on the waveform, I double checked the probe, both probes are X10, so the waveform yellow is 1V per grade, and pink is 500mV per grade.  Thank you,

Hi Patrick

I did notice that the voltage on the IFB is negative and the unused IFB have similar waveform. I will catch the waveform and post it here tomorrow. by the way, I will also capture the same waveform on the demo board to make a comparison. 

Here are my questions that may be tricky, 

1. on the demo board, C11, the BOM is 0.33uF with a note (1) say use 1uF - 4.7uF. I didn't pay attention to this note, so I used 0.33uF on my board, is it possible that this wrong capacitor value will cause some strange behavior. 

2. in my design, I put an external resistor 0.75 ohm outside the IFB, and use a high common mode voltage OPamp to measure the current. could you please ask circuit expert that this OPAMP in the design won't introduce any strange behavior. 

Thank you,

David

Hi David,

1. You could capture the waveform of REF to see if this voltage is affected to be unstable.

2. This seems ok to me but I think it is better to disconnect the Opamp and re-test.

BR

Patrick

Hi Patrick, 

Please have a look at this thread, the current sensor chip will sink a 40uA current to it. this could be the reason of the negative voltage. please have a look at this thread. 

https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/1256308/ina241a-common-mode-input-resistance/4755215#4755215

I will remove the sensor and test it again. hopefully this will make the drive start. 

Thank you,

David

Hi David,

Noted, let me know when you have any update.

BR

Patrick

once I removed the INA241, the LED driver works. Thank you for supporting.

Hi Patrick

To make the design works, I made some patch over the original design, try to remove the 40uA drain when 61196 start. could you please help review this schematic to see if this one will work?

Thank you,

David

current_sensor_fix.pdf

Thank you very much,

David

Hi Patrick

Is it possible to know the value of the current sensing resistor inside the chip? 

is there a chance to estimate the current by measure the voltage of the IFBx?

Thank you,

David

Hi David,

I don't think the IFBx voltage could reflect the current accurately because this IC is applying adaptive control.

BR

Patrick