Turn me on

Gents, I'm trying to kluge together an entry announcing system using a garage door opener sensor, a rotating light and a horn. Most of it works except for the horn. I'm using a MOSFET in series with the horn and the power supply. The problem is the horn won't turn on. I believe the problem is since the horn is purely inductive that the MOSFET can't do the massive inrush current the horn requires. I don't have a 5v relay that will push 10 amps through its contacts. I tried a transistor but must it sill won't work as it's trying to suck all the power throught the base-emitter junction and not the emitter-collector junction.

Hi,
Where do you have your MOSFET switch located? If it's an N-channel MOSFET you need to place it between the load and ground. (or use a proper high side driver arangement).

You also need to make sure that the voltage you're driving the gate with (5V?) is enough to turn to the MOSFET on, "normal" MOSFETS usually needs around 10V between gate and gource and may not fully turn on with 5V. There are logic level MOSFETs available that will work finw with 5V.

/Henrik.

I do indeed have the MOSFET in the low side of the horn circuit. Off the same power supply using it's own MOSFET I've got a rotating beacon pulling about 4 amps using an N Channel MOSFET(NTE 2376) to switch it. Neither are on at the same time and the power supply is capable of driving the horn because if the leads of the horn are touched directly to the power supply the horn makes noise. When I send 5 volts out to the MOSFET it turns the rotating beacon on for the time I have it programmed. When I use this arrangement to power the horn, as soon as 5 volts is put on the gate the horn thunks then after 3 or 4 seconds begins to turn on. After a total of 5 seconds it is fully on but the MOSFET and the 12F683 are extremely hot. In fact, this arrangement has already burned up one PIC. I have a 1k resistor going from the PIC to the gate. If I remove the horn load with a new PIC and use an LED on the IO pin to see if I get the rate I want, 1 second on 1 second off, the LED works fine. I just can't get the horn to come on without burning up something.

I then tried to use a transistor as a silicon switch and that's when I found the current trying to draw through the base.

I know this isn't really an electronics bulletin board but what I am trying to build at least uses a PIC. Thanks for any thoughts or ideas.

I do not see a fly back diode mentioned in your description.
Do you have one?

Hi,
What's the powersupply voltage and what's the current draw of the horn?

I still think it's a Vg-s problem, ie the Gate to Source voltage is too low to fully turn on the MOSFET, why the PIC gets warm I don't know though - perhaps you get a dip in th epower supply voltage and the whole thing starts to oscillate or something (?).

If the powersupply is 12V then I'd try something like in the attached schematic. It has the drawback of inverting the signal but that can be taken care of if it works. Also, if the PIC is powered from the same supply as the horn (thru a regulator of course) make sure you have ample decoupling right at the powersupply pins of PIC.

The intrinsic body diode of the MOSFET is most likely enough in protection in this case.

/Henrik.

Quote:

---

Originally Posted by AvionicsMaster1

I don't have a 5v relay that will push 10 amps through its contacts.

---

Why must it be a 5v relay? If you have a higher voltage available, you could use that with an NPN, right? In fact, at 12 v you'll have a choice to pick from. I have sugar-cube relays that state the contact rating to be 10A @ 28v, but if you dont want to take chances you can always use larger ones.

Regards,

Anand

The reason I used a 5 volt relay is because that's what I had available. I recently bought a 12 relay to try. Thanks for the transistor mosfet suggestion. I haven't had time to try it yet.

I did monitor the power supply while the circuit was operating. I used a fluke meter that could catch peak voltages. During the horn on cycle the power doesn't drop off more that a few hundred millivolts.

I don't have decoupling caps at the chip but I'll try that also.

Thanks for all the ideas.

There are N-Channel mosfets with low threshold level, the so called logic Mosfets.

Like the , BUK9575, 12N06L, IRF3205

Also if anyone interested in P-channel a good one I tested recently is the IRF4905. An excellent P-channel Mosfet.

Ioannis

I thought I'd finished this out but something didn't work.

Thanks to Henrik, using the transistor and 12 volts to turn the MOSFET on does work; most of the time. Periodically it will skip parts of the routine but restart at the right place. I feel this could be due to the power supply and all I need is to find some filter caps.

Thanks for all the help.