Low pulse on relay close

[Yoshimura]

Hi there,

I want to remove start my camper engine when battery get too low. For this, I have a battery monitor that can close a relay when battery get too low and a remote starter with an "external trigger" input. However, the trigger input needs a low pulse. So I tough of this small circuit that generates a low pulse on power-on.

So on power-on, C1 is discharged and Q1 conducts until C1 has charged mainly through R1. I added R3 so C1 can discharge when power is removed (so it's ready for the next cycle).

Is my assumption that the Remote Starter input has a pull-up resistor correct?
Any suggestion for a proper part for Q1?
How can I compute proper value for R2?

When relay close, it will stay closed for 30 minutes and won't close again before another 30 minutes.

Many thank!!!!

 

[AnalogKid]

So far, ok.

Can you measure the current out of the remote starter input pin to GND? You probably can do this by putting a DMM in current mode, and letting it be the low-resistance path to GND. You need to know the peak current the transistor is going to handle before you can select the transistor.

I would increase R2 to keep the peak current into the base below 15 mA, and preferably at 5 mA.

What is the minimum low-input time the remote starter needs to operate correctly?

If R3 is connected directly to GND, it will not affect the C1 charging time calculations. With this change, consider R1 = R3 = 10K; R2 = 1K.

ak

 

[Yoshimura]

First, thanks for the DMM in current mode suggestion!

So I tried using my Fluke but it lowest range is 6A and the display only flickered between 0 mA and 1 mA.

1 mA would make sense or should I validate with a shunt?

Ok, I tried with a 1K shunt resistor and measured 350 mV. 350 uA make sense?

There is no specification for the pulse width but I assume something around 100 ms would be good.

 

[Yoshimura]

I'm now wondering... Just connecting the DMM between the input and the ground triggered the remote starter. Could it be simply the bounces between the probe and the wire when connected it? However, the trigger can be configured to be 1, 2 or 3 low pulses (currently set to one). So bounces and thus multiple pulses should not have trigger it. Hum...

 

[AnalogKid]

In a 12 V system, 350 uA implies a 33K pullup resistor (or equiv). Seems kinda high to me for a relatively high-noise environment.

Your current schematic calcs out to a 1.5 ms pulse. That's probably too short for reliable operation. Consider increasing C1 to 100 uF.

ak

 

[Yoshimura]

If R3 is connected directly to GND, it will not affect the C1 charging time calculations. With this change, consider R1 = R3 = 10K; R2 = 1K.

I simulated the circuit with R3 connected to GND and when the relay open, base voltage goes to -6V which I think make sense. However, this is not good for Q1, right?

 

[Yoshimura]

So here's my final circuit. I set R3 so transistor gain is 10. With current values, pulse width is 160 ms.

 

[AnalogKid]

Looks good to me.