Post Go back to editing

# LTC4412 switching to different input

I am prototyping the LTC4412 and I have a question on how to improve my design.

My circuit is like Figure 2 of the datasheet.

(Sense pin is connected to Out).

The design works and switches to higher input but only if the difference in input voltages is above 0.8V.

In fact, it depends on the load. My tests are in the following table.

 Load Input voltage differenceto switch to higher input 100mA 0.6V 200mA 0.8V 500mA 1.2V 1A 1.6V

Am I doing something wrong?

• Hello,

your circuit looks like our demoboard schematic:

https://www.analog.com/media/en/technical-documentation/eval-board-schematic/DC1635A_REV2_SCH.pdf

Upper realization. How is your 'SOLDERJUMPER_2WAYS' connected/handled?

Markus

• Thank you Markus for the response.

In my prototype, the SOLDERJUMPER does not exist. As I wrote the Sence Pin is connected to Out.

• Hello Marcus,

Do you have any information on what can be wrong with my board?

Thank you

Dimitris

• Hi Dimitris,
What are your operating voltages for IN1 and IN2?
Once IN1>IN2+700mV+20mV, IN1 should carry the load.
This shouldn't vary too much with load current.

-Aaron

• One of my boards is working with 5V and another with 12V.
Both have the same problem.
For example for the 5V project, I have set In2 to 4.6V, and when In1 is below 5.6V it switches to In2.
How did you calculate this 700mV (for the formula IN1>IN2+700mV+20mV)

• Dimitris,
The ~700mV comes from the body diode drop across IN1’s PFET.

I drew out your circuit in LTspice and simulated it:

I hold IN2 at 5V, and ramp IN1 from 0V to 10V, then back to 0V. There is a 1A load on the output.

When IN1 ramps up, we can see that OUT begins to follow when IN1=5.65V.
The 650mV difference between IN1 and IN2 at switchover is because Q1 was off, so current had to pass through the body diode when it was forward biased.

When IN1 ramps down, we can see OUT switches over to IN2 when IN1=5.05V.
The voltage difference at switchover is less than before, because Q1 was on, which minimized the drop across the PFET.

During switchover, the biggest difference you should see between IN1 and IN2 is ~700mV (body diode drop).

Can you get some scope-shots of IN1, IN2, and OUT as you ramp IN1 above and below IN2?
How are you setting the load current?

Sim file is attached. 4412 switchover.asc

-Aaron

• Thank you Aaron for the detailed answer.

Unfortunately I don't have an oscilloscope right now to use.
I am not expert in electronics but I understood a lot from your answer.

I think I am ok with my circuit with a voltage difference of about 700mV.