TPS92692: Continuation of closed case

Gerry,

1.  We specified that because we can configure the board as a BOOST or BOOST to Battery (Buck Boost).  For the Boost to Battery case we will have the input voltage + the output LED voltage and that's why we specified it that way for the automotive customers.  For your case you can configure it as a BOOST and it will operate with 20V input with no issue.

2. L3 and L5 are for EMI filtering if you have to pass class 3 automotive specification.  For your application you remove and bypass them.

Thanks Tuan

What about L1?

Gerry,

L1 is used as an input filter for EMI.  You can bypass it also.

Thanks Tuan

I have modified the evaluation board to run at 1.5 A led current by changing the led current sense resistor to 0.115 ohms. This seems to work in terms of measured output power to the leds. I am running with the stock timing resistor of 20 k which should give a switching frequency of 390 kHz. I have put a jumper on J6 to disable the spread spectrum modulation. However the switching frequency is not stable with a variation from around 350 kHz to 440 kHz. The duty cycle is fairly consistent at around 41.8% with a calculated duty cycle from the spreadsheet of 44%. Any ideas why this is happening?

At near maximum output, the peak inductor current is 3.4 A compared to 2.7 from the spreadsheet calculations.

The switching frequency variation is the same with the DM jumper removed (spread spectrum enabled)

Gerry,

It's J8 for the DM that you have to short not J6.  The speadsheet calculates the input voltage to the power stage it does not assume the diode drop on the evaluation board.  

Thanks Tuan

Oops my mistake, screen resolution issue. Will try it again. Nice to know it works.

OK, switching frequency is stable at 387.5 kHz

Now I am going to try increasing the switching frequency and using a smaller inductance

It seems to be working ok at the higher frequency. When you the spreadsheet does not assume the diode drop, do you mean D2 and D3? I have bypassed them and L1 and it doesn't seem to make much difference to the peak inductor current.

I am trying to understand why the peak inductor current is so much higher than the calculated value. This partly the reason our original design did not run at full power.

I have the evaluation board working at 1.5A output current running at a switching frequency of 800 kHz. The analog dimming input works fine. However, both the internal and external pwm dimming setups create flickering in the led output. The external pwm is running at 1 kHz. Any ideas about why this is happening?

Gerry,

If you are changing things and it's not working then you have to trouble shoot/debug and figure things out thing now.   

Thanks Tuan

Yes of course but perhaps you could give me a idea of where to start looking.