overcurrent protection circuit

[Jony130]

Hi all.
What do you think about this simply overcurrent protection circuit (pros and cons) ?
Or some improvement tips.
Also I think that in this circuit we don't need D1 diode.

 

[kubeek]

You definitely do need D1 as it will prevent T1 and LED2 from overvoltage when the relay turns off.
Any way, the biggest problem with this circuit is that there needs to be at least 0.7V drop across R5 for the protection to kick in which may adversely affect the circuit that is powered from it.

 

[Jony130]

You definitely do need D1 as it will prevent T1 and LED2 from overvoltage when the relay turns off.

If I understand how relay coil behave during turns off time. The inductor " kick" voltage cannot be greater than -Vcc.
Because relay coil is connect to a BJT emitter. So we don't have a open circuit situation here.

 

[Norator]

How does the I-squared-T curve of this circuit compare with the curve of the device to be protected and how does it compare with the curves of existing fuses & breakers?
It doesn't seem to me that you need the coil diode but there may be hidden failure modes. I guess the cost of the diode is the cost of the lump sum premium for 'lifetime circuit insurance.'

 

[kubeek]

If I understand how relay coil behave during turns off time. The inductor " kick" voltage cannot be greater than -Vcc.

How did you come to that conclusion? The shorter the turn off time, the higher the kickback voltage.
Assuming the transistor turns off instantly, then the voltage on the coil will rise either to infinity or untill something else gives way. It could be the inter-winding isolation of the coil, but much more likely it will be the breakdown voltage of the transistor plus reverse breakdown voltage of the diode, so about 55V. Then whatever current was flowing through the relay will flow through the diode and the transistor, but the transistor will have 50V drop across it and the diode will have about 5V across it, which could destroy them both.

 

[Jony130]

How did you come to that conclusion?

After T2 and T3 "SCR" is trigger, the voltage at T1 base drops to almost 0V. So the relay coil tries to keep the current to flow in the same direction as before.
The inductor " kick" voltage at T1 emitter goes to negative potential. And when this voltage reach -0.7V the BJT goes into conduction mode again. T1 will be ON until relay coil "discharge" their all energy.

 

[MikeMl]

How much current flows through the relay coil? Is LED2 capable of withstanding that much current?

Sure looks like LED2 is vulnerable to getting zapped by the inductive kick without D1.

Will the relay pull-in reliably with only ~9V across it?

There is a huge current spike through T1 when it turns on.

What resets it?

I dont like it because it wastes a lot of power forever waiting for something to happen.

 

[Jony130]

How much current flows through the relay coil? Is LED2 capable of withstanding that much current?

The relay coil resistance is 390Ω

Sure looks like LED2 is vulnerable to getting zapped by the inductive kick without D1.

Can you explain how is is passable. Can you point where I make a error in analysis show in post 6?

Will the relay pull-in reliably with only ~9V across it?

Yes

There is a huge current spike through T1 when it turns on.

Inductive load and a current spike ?

What resets it?

Switch-off power supply.

I dont like it because it wastes a lot of power forever waiting for something to happen.

Can you show a better solution for 10A surge current (car battery charger).

 

[MikeMl]

...

Can you show a better solution for 10A surge current (car battery charger).

Any time you connect a DC power supply (which has an open-circuit voltage more than ~13V) to a partially discharged car battery, the initial charging current is limited only by the power supply; the battery is happy to accept a charging current in excess of 50A which in all likelyhood will overheat the supply.

Putting your relay circuit between the supply and the battery means that it will always trip when the supply is first turned on. You cannot ever charge a battery with it...

What you need instead is a power supply which is current-limited as well as being voltage regulated. The inital current limit is to protect the supply from the battery; the voltage regulation is to protect the battery from being overcharged by the supply after it has reached full charge state.

Read up on the proper algorithm for charging lead-acid batteries. It was recently covered here and here

 

[audioguru]

Why does anybody need a car battery charger? The car charges its battery every day.
Oh, maybe when the car driver is in jail for a few months or more?

 

[Jony130]

Well, in my country once a year we have a winter. Also we must drive our cars with the lights on all year round.
And sometimes we simply forget to turn off the lights after we turn-off the car.

 

[KeepItSimpleStupid]

Jony130:

Way back when, like in 1982, I wanted a unobtrusive "lights on Buzzer" of some sort. So, I said what might be simple?

I chose "Turn on buzzer when driver's door is open"

All I needed was a piezo buzzer and a diode. The driver's door contact was already isolated with a diode and went to ground when the dome light was on. So, I connected the +12 going to the parking lights to the driver's door switch through a diode.

Bingo: When driver's door is open, the buzzer sounds. Used the system for 17 years and had to replace the buzzer once. The car interior is harsh for a buzzer.

My current car has the: Lights go out when the door opens. Lights automatically turn on. You can, have the lights on and then stop, but not get out of the car and have your car die if you sit in the car too long. That is a rare event, so I just have to remember to open the door.

The "auto on" lights are tied to the emergency brake in both cars. The lights will not come on if the emergency brake is on.

The other car has a number of "modes" for the way the lights operate. Not all I like, but having them go off after 5 minutes after the door is opened is OK except when you want to scare away a deer that stopped in front of you.

I do recommend that you change the dome light to a LED based variety as well. The dome and the license plate light seem to be the bulbs that go first.

 

[audioguru]

Well, in my country once a year we have a winter. Also we must drive our cars with the lights on all year round.
And sometimes we simply forget to turn off the lights after we turn-off the car.

Most of our cars have automatic lights.
When it gets dark (maybe even from dark clouds at noon or in a parking garage) then the lights turn on. When it is bright outside then the lights turn off.

Our cars have "daytime running lights" which for most cars is automatic and they are dimmed (except Jeep and Chrysler cars that blind you with their very bright high beams used for running lights in the daytime).

 

[Norator]

"Depending on prevailing regulations and vehicle equipment, the daytime running light function may be implemented by functionally specific lamps, by operating the low-beam headlamps or fog lamps at full or reduced intensity, by operating the high-beam headlamps at reduced intensity, or by steady-burning operation of the front turn signals. "

If you petition the Canadian DOT it may decide that the Jeep and Chrysler vehicles are in violation of their rules.

 

[audioguru]

[MODNOTE]Deleted Off Topic[/MODNOTE]

 

[Norator]

[MODNOTE]Deleted Off Topic[/MODNOTE]

 

[audioguru]

[MODNOTE]Deleted Off Topic.[/MODNOTE]

 

[Norator]

[MODNOTE]Deleted Off Topic[/MODNOTE]