Switching 12v with micro controller

[Makaram]

Hey,
A discussion on a separate thread has convinced me to post a separate question regarding how to switch 12v (or more) from an MCU.

https://www.bristolwatch.com/ele/transistor_drivers.htm was one resource I have found but am getting a lot of mixed messages from the different sources.

Any input would be most appreciated
Schematics would be great!

Thanks

 

[heydonms]

That page seems to cover the most common options. If you can tell use exactly what you're trying to do then maybe someone could offer some advice as to which option would suit you best.

 

[Makaram]

Ive been told that I cant straight switch with mosfet?

 

[jpanhalt]

Ive been told that I cant straight switch with mosfet?

Of course you can. That is a very common, if not most common use for mosfets.

Your other thread got a bit confusing. I am glad you started this thread here. But please be more specific about what you want to switch, and why you don't think yo can use a mosfet.

John

 

[Makaram]

View attachment 67817

J1 in this picture is pulled high when mcu1 is high. (Opening car doors)
j2-1 and j2-2 are disconnected from 12v when mcu2 goes low. (Locking car doors)

I have achieve what q1 is achieving using a relay.

I think its just signal stuff, the problem is simply in the voltage difference

 

[jpanhalt]

I can't tell from your schematic what you are trying to do.

1) J2-2 is always low. It is connected to ground.
2) J2-1 is always high. It is connected to +12V
3) J1-1 is trying to use an N-channel Mosfet to source current. It can be done with an appropriate driver, but an MCU is not that driver. You must drive the gate to 12V + Vgs.

Please take care in presenting your schematics. As a suggestion, please just work on getting one switching mosfet to work. Then expand to the more complex circuit.

When using an N-channel mosfet for a switch, its source should be connected directly to ground. The load should be connected to the drain and V+. The gate needs to be logic level. That is, fully turned on by ≤ +5V .

John

 

[audioguru]

The gate needs to be logic level. That is, fully turned on by ≤ +5V .

But in the other thread the MCU has only a 3.3V supply so even a logic-level Mosfet will not work.

The OP must learn that an ordinary Mosfet turns on when its gate to source voltage is 10V so using a Mosfet as a source follower will not work. The Mosfet also cannot be turned on with a 3.3v MCU.
A normal transistor can be turned on and off by the MCU and its collector can turn on a Mosfet and a gate to source resistor can turn off the Mosfet.

 

[jpanhalt]

But in the other thread the MCU has only a 3.3V supply so even a logic-level Mosfet will not work.

Actually, many newer mosfets are spec'd at <3.0V Vgs. The Si1400DL, for example, is spec'd for 1.3A@2.5 V. That is just one of over 200 low-gate-voltage mosfets at DigiKey. With all the lithium-battery powered devices, I am a bit surprised someone feels you still need 5V Vgs to turn on.

John

 

[audioguru]

Actually, many newer mosfets are spec'd at <3.0V Vgs. The Si1400DL, for example, is spec'd for 1.3A@2.5 V.

That is its "typical" rating without a guarantee. Try one, oops it doesn't work. Try another one, oops it also doesn't work. If you try 10 of them then a couple of them might work. Why gamble?

Its maximum on-resistance is 0.235 ohms when it is cool with a Vgs of 2.5V but the on-resistance increases when it is hot. 1.3A squared x 0.235 ohms= 0.4W which will make the tiny little thing very hot and it will get hotter and hotter.

It is guaranteed to conduct at least 2A (its minimum rating) when its gate-source voltage is 4.5V and when it is cooled somehow.

 

[jpanhalt]

That is its "typical" rating without a guarantee. Try one, oops it doesn't work. Try another one, oops it also doesn't work. If you try 10 of them then a couple of them might work. Why gamble?

That is great information that I was not aware of. Perhaps you should flesh it out with a little more data and post it here on ETO as a sticky. An awful lot of manufacturers must me making loads of money off the scam of selling the defective chips.

John

 

[ronv]

That is its "typical" rating without a guarantee. Try one, oops it doesn't work. Try another one, oops it also doesn't work. If you try 10 of them then a couple of them might work. Why gamble?

Actually you will be much more likely to find one that is better than the spec than one that is at or out of spec. If there is anything fabs hate it's poor yield.

Anyway. Take a look at this circuit and see how it would work for your door locks.

 

[audioguru]

That is great information that I was not aware of. Perhaps you should flesh it out with a little more data and post it here on ETO as a sticky. An awful lot of manufacturers must me making loads of money off the scam of selling the defective chips.

They advertise "typical" ratings. Some are better and others are worse. Design so that your circuit works with a weak but passable transistor then it will work with any of them.

The chips are not defective. They are simply not all exactly the same. Some are not as good as the best ones and they sell the weaker ones and the better ones. You get whatever is available when you buy one.

 

[jpanhalt]

So we can agree that mosfets, both N and P channel, are available that will work at 3.3V, as described in posts #6 thru #8.

I agree with ronv (post #11). We should concentrate of the circuit. Maybe a transistor driver is needed, but that seems to get us back to the other thread.

I am not sure a transistor driver is needed, because I am not really sure what the OP wants out of this thread. PIC's and other MCU's can certainly drive small mosfets directly, particularly when the switching speed is not critical.

Given a 12V supply to the latch and a 3.3V mosfet, the simplest solution for a switch might be an N-channel mosfet (rated at ≤ 3.3V Vgs) sinking current from the door latch.

John