Hi guys,

I am new to the forum and this is my first post. I am fascinated by the circuits but I am a noob when it comes to designing them so I ask for your precious help.

I am looking for a circuit to use in my car. I added an Android head unit and a rear view mirror with camera function to my old car. I want protect the newly added gadgets by using an on-delay and off-delay relay circuit because every time I turn the key they get on and off instantly. Most of the circuit designs are for timers but I don't need a timer. Circuit will operate according to a line energized by turning the key. As you may already know cars tend to have "dirty" electricity varying between 12 and 14,5 volts. I believe the main load will be on the relay so the circuit does not necessarily need to handle much current. Any help would be greatly appreciated.

 

How long of a delay do you need for both ON and OFF?
In this circuit the value of C1 will determine the delay times.

 

Should add a diode across the relay coil to protect the MOSFET from the inductive transient.

 

Should add a diode across the relay coil to protect the MOSFET from the inductive transient.

Q1 shows an internal protection diode.

 

Q1 shows an internal protection diode.

that does not protect against the inductive spike, which is a positive voltage at the MOSFET drain.

 

From the way the cathode is drawn in the schematic, one might suspect that the drain-to-source diode might breakdown in reverse. Looking at the data sheet, there is no indication that this is the case.

Schematic from post #1.



Schematic from On Semi data sheet.

 

If you zoom in on the BS170 image, you will see that the diode is show as a zener. This is a perfectly adequate way of protecting the transistor, if you don't mind a small increase in EMI because the zener voltage is greater than the clamping voltage of a simple rectifier across the coil. An added benefit, according to a P&B app note, is that the relay turns off faster with this clamp method, increasing contact life. Also, it creates a smaller perturbance on the Vcc rail.

The zener voltage often is not specified on the datasheet. When it is, it often is only a minimum value that is greater than the transistor's max Vds rating. I get that there is a difference between an actual zener diode and a simple rectifier with a minimum reverse breakdown voltage. Whatever - it makes a good-enough clamp in a circuit such as this.

"Body diodes may be utilized as freewheeling diodes for inductive loads "
https://en.wikipedia.org/wiki/Power_MOSFET

ak

 

Should add a diode across the relay coil to protect the MOSFET from the inductive transient.

Shouldn't be a transient as the voltage decays across the relay coil.

 

From the way the cathode is drawn in the schematic, one might suspect that the drain-to-source diode might breakdown in reverse. Looking at the data sheet, there is no indication that this is the case.

Schematic from post #1.
View attachment 327960


Schematic from On Semi data sheet.
View attachment 327959

That mosfet symbol is just for schematic purposes. It's actually from the 2N7000 but I use it generically.

 

that does not protect against the inductive spike, which is a positive voltage at the MOSFET drain.

Sorry, my mistake. Yes, you need a diode across the inductive load.

 

If you zoom in on the BS170 image, you will see that the diode is show as a zener.

Not in the ONsemi data sheet I looked at.

Shouldn't be a transient as the voltage decays across the relay coil.

??
The inductance will generate a positive voltage spike at the MOSFET drain when the MOSFET turns off (LTspice sim below).
Why do you think it won't?

 

The inductance will generate a positive voltage spike at the MOSFET drain when the MOSFET turns off (LTspice sim below).
Why do you think it won't?

View attachment 327962

Because it does not shut off as you show in the sim. As C1 "slowly" discharges the voltage across the mosfet will "slowly" increase.
I don't see a transient.

 

Because it does not shut off as you show in the sim. As C1 "slowly" discharges the voltage across the mosfet will "slowly" increase.
I don't see a transient.

Point taken.

 

Because it does not shut off as you show in the sim. As C1 "slowly" discharges the voltage across the mosfet will "slowly" increase.
I don't see a transient.

Fair enough.
For reasonable values of delay and relay coil inductance, the turn-off time is slow enough to minimize or eliminate the transient.

 

Not in the ONsemi data sheet I looked at.

It might be that small signal FETs don't have one. Been too long . . .

I checked Motorola, IR, and Harris. For power MOSFETs, all show a zener in the drawing. The diode has the same continuous and pulsed current ratings as the FET. I'm sure I've seen a minimum breakdown voltage spec somewhere, but of course I can't find it now.

ak

 

Thank you very much for your answers it feels like i am home right from my first post.

How long of a delay do you need for both ON and OFF?
In this circuit the value of C1 will determine the delay times.
View attachment 327935

I am not quite sure but at least a 10 seconds delay would be enough. I like to play with components (which generally ends bad in my trials ) and I can change the value according to my need. You gave me something to build on. Many thanks!

 

For 10 seconds start with C1 at 10uf and see how that works.

 

For 10 seconds start with C1 at 10uf and see how that works.

I will gather the components first thing on Monday and let you know of the results. By the way, is it possible to use BC547 instead of the mosfet and regular 1n4007 instead of 1n4148? I have everything but the mosfet and the diode.

 

BC547 won't work as designed, needs a mosfet.
1n4007 is OK

 

is it possible to use BC547 instead of the mosfet

A MOSFET is preferred.

Also a BC547 is only rated for 100mA, which is not enough to drive an automotive 12V relay.

Do you have any other transistors?