LTC2995/2994

Hi Saravana.
I believe you mean the LTC2954 and LTC2955 pushbutton controllers.

If you want adjustable on and off timers, the LTC2954 is best for your application.

Saravana21 said:

Is there any way to add debounce time for OFF event. It is preferred to have more debouncing time for OFF than ON.

For power down via the PB# pin, the LTC2954 has 64ms of debounce by default, but additional delay can be added via a capacitor on the PDT pin. See page 9 of the datasheet for the equation to calculate the value of C_PDT , given your desired delay.

-Aaron

Dear Aaron,

Thanks for your response.

We are planning to use LTC2955 due to our requirements. If I add an external RC debouncing on PB pin, will there be any impact in LTC2955 operation? I have attached a reference circuit for having around 600ms of debouncing for ON and OFF.

Please check and confirm.

Regards,

Saravana

Hi Saravana,
I built your circuit in LTSpice (sim file attached).
2955-1 EZ debounce 9-3-21.asc

As is, it doesn't work.
It's because the PB# pin doesn't discharge to 0V when the switch is pressed.
This is because there is an internal pull-up to an LDO that needs to be factored in.

When considering the internal pull-up, the cap on the PB# should discharge to ~820mV.
Coincidentally, the typical threshold for the PB# pin is 800mV.
There isn't enough margin; PB# must go below 600mV when the switch is pressed (the minimum threshold possible).

When I lowered the 200k resistor to 100k, I was able to get the EN pin to change state.

I will leave the sim file with you to determine what the right RC values are.

-Aaron

Hi Aaron,

Thanks for your quick response.

I understand the concern stated by you.

I have requested my IT colleague for installation of LT Spice in my machine.

I will check and get back for any further information.

Regards,
Saravana

Hi Aaron,

I have an additional question regarding KILL input. As per my understanding from datasheet, always KILL input is related to PB input and INT output pins. Is it possible to control the KILL input by uP or any other peripheral independent on PB input pin? 

Regards,

Saravana

Hi Saravana,
Yes, that is possible.

From the datasheet’s pin description: “A low at KILL# releases the enable output (EN/EN). This pin can be driven by the microprocessor or it can be used as a voltage monitor input.”

KILL# is independent of PB# and INT#.

Be aware of the KILL blanking time, where the KILL# input is ignored for 512ms after EN goes high.

-Aaron

Hi Aaron,

Thanks for quick response. That was helpful.

We have a query regarding figure shown for "Power First Applied to LTC2955-1" (Ref: Figure-7 in datasheet).

We are not using ON pin and hence grounded in design. So we would like to understand the "Power First Applied to LTC2955-1" without consideration of ON pin relation. A sequence with only VIN, PB Pin and EN Pin is what we are looking for. Or we would like to know the status of EN pin during first power application. Please check and let us know your inputs.

Regards,

Saravana

Saravana21 said:

we would like to know the status of EN pin during first power application.

EN starts off LOW.

Saravana21 said:

we would like to understand the "Power First Applied to LTC2955-1" without consideration of ON pin relation. A sequence with only VIN, PB Pin and EN Pin is what we are looking for.

I ran a simulation to show the timing diagram of these pins. See the pic below.

At t=1sec, IN steps from 0V to 6V

Note that there is a 1s lockout time (typ) after VIN reaches UVLO of 1.15V. PB# inputs are ignored in this window, so I wait 1.5s to be safe.

At t=2.5s, PB# steps from 1V to 0V. (Falling edge)

At t=2.5313s (~32ms later), EN goes high.

Note that there is a 512ms KILL blanking time after EN goes HIGH, where PB# inputs are ignored.

At t=4s, PB# steps from 1V to 0V. (Falling edge)

At t=4.1103s (~110ms later), EN goes low. This is close to the calculated 116ms of debounce off time.

Here is the sim file:
2955-1 EZ 9-22-21.asc

-Aaron

Saravana21 said:

we would like to know the status of EN pin during first power application.

EN starts off LOW.

Saravana21 said:

we would like to understand the "Power First Applied to LTC2955-1" without consideration of ON pin relation. A sequence with only VIN, PB Pin and EN Pin is what we are looking for.

I ran a simulation to show the timing diagram of these pins. See the pic below.

At t=1sec, IN steps from 0V to 6V

Note that there is a 1s lockout time (typ) after VIN reaches UVLO of 1.15V. PB# inputs are ignored in this window, so I wait 1.5s to be safe.

At t=2.5s, PB# steps from 1V to 0V. (Falling edge)

At t=2.5313s (~32ms later), EN goes high.

Note that there is a 512ms KILL blanking time after EN goes HIGH, where PB# inputs are ignored.

At t=4s, PB# steps from 1V to 0V. (Falling edge)

At t=4.1103s (~110ms later), EN goes low. This is close to the calculated 116ms of debounce off time.

Here is the sim file:
2955-1 EZ 9-22-21.asc

-Aaron

Hi Aaron,

Thanks for quick response.

After seeing the waveform I got a doubt regarding TMR capacitor. Actually it will add extra debounce only for force KILL option rigth in case of software fails to control the KILL input? Otherwise the normal debounce time will be 64ms and shutdown will happen through INT and KILL pins. Is this understanding correct?

Regards,

Saravana

Yes, I believe your statement is correct.

I encourage you to download LTspice and play around with your circuit.
This will let you check if the part works as you expect.

Have a good day,
-Aaron