Hello everyone! I have a few questions about my circuit (no longer am I trying to do 120V), I'll stick to the safer 12V scheme.

I am using this thread and this thread as a reference. I am using the LM555 timer and the IPP110N20N3 G transistor. I am most curios about the transistor side.

1. Do I need anything in-between the 12VDC source and the transistor aside from the LED?
2. Do I need anything after the transistor?

Thanks in advance,
J

 

Good to go. The LED will draw 750mA at 12V.

 

The FET is upside down. LED will always be on.

 

Do I need anything in-between the 12VDC source and the transistor aside from the LED?

Good question. If your 12V source is truly 12V, you should be OK. It's getting close to the maximum the device can survive, but OK. If it's an automotive 12V system that might really be up at 14V or even more, you may have a problem.

If you're planning to switch it at high frequency, you might consider a gate resistor to limit the 555 output current to 200mA. The current required to charge and discharge the gate capacitance can be significant as frequency increases. Of course, adding gate resistance slows switching and causes heat in the MOSFET.

You might also consider a pull-down resistor, 10K say, on the gate to protect against a failure of the 555 that could put the gate at an indeterminate voltage.

 

The FET is upside down. LED will always be on.

Good catch. And he even drew the substrate connection and parasitic diode...

 

The FET is upside down. LED will always be on.

Bernard,

Thanks for your comment. Why is it upside down? Once current is applied to the gate, it will allow current flow from the source to the drain. The diode on the drain line shows current can only go toward the drain. The image from the datasheet is shown below. Does this specific transistor work in a different way?



Thanks,
J

 

If your 12V source is truly 12V, you should be OK....

I think I'll put a little voltage divider before this circuit to bring the 12V down to 11.8 or something. The max voltage they suggest in the datasheet is 11.9 - good catch,

If you're planning to switch it at high frequency, you might consider a gate resistor to limit the 555 output current to 200mA....

Ok now I'm a little confused. The 555 timer datasheet only shows what current the 555 timer can source - but aren't I more concerned about the voltage the output pin (pin 3) can supply? The mosfet gate is tripped by a voltage, right? What voltage can the 555 timer output supply? Does it just supply the voltage from the Vcc pin (pin 8)?

And I would only need to limit the current if the MOSFET required a large current to trip the gate right? The MOSFET data sheet only talks about a gate threshold voltage (Typ 3V, page 2 of data sheet). But I bet MOSFETs gates don't draw large currents, right?

You might also consider a pull-down resistor, 10K say, on the gate to protect against a failure of the 555 that could put the gate at an indeterminate voltage.

Good idea, this would help things from floating around between the 555 and the gate, right?

 

Why is it upside down?

The source should be grounded, not the drain. With the device connected as shown in your schematic, the body diode would always be forward biased and the LED would always be on; regardless of the gate voltage.

 

I think I'll put a little voltage divider before this circuit to bring the 12V down to 11.8 or something. The max voltage they suggest in the datasheet is 11.9 - good catch,

A voltage divider would waste a lot of power. Maybe just a series resistor.

The 555 timer datasheet only shows what current the 555 timer can source - but aren't I more concerned about the voltage the output pin (pin 3) can supply? The mosfet gate is tripped by a voltage, right? What voltage can the 555 timer output supply? Does it just supply the voltage from the Vcc pin (pin 8)?

And I would only need to limit the current if the MOSFET required a large current to trip the gate right? The MOSFET data sheet only talks about a gate threshold voltage (Typ 3V, page 2 of data sheet). But I bet MOSFETs gates don't draw large currents, right?

In the DC realm, you're right. It takes almost no current to change the state of the MOSFET's gate, and all you need to worry about is whether the gate voltage from the 555 is high enough to fully turn on the MOSFET. But that gate is like a small capacitor and it holds a charge. To repeatedly charge and discharge it at high frequency requires moving that small charge back and forth many times and that adds up to an alternating current. I'm guessing here at the numbers, but at anything less than 10kHz it's probably no concern at all. Over 100kHz it likely is a concern. In between, I'd have to do the calculation.

Good idea, this would help things from floating around between the 555 and the gate, right?

Well they're connected, so normally there is no "between". But ICs can fail or have undefined states (for instance when power is first applied), and that resistor makes sure the gate is turned off for sure unless it's turned on by the 555.

And I agree that the MOSFET source should be ground, not the drain. A MOSFET works either way, but as noted the body diode will conduct in the orientation you have shown. It can't shut off the LEDs and might burn out, depending on the LED current and the ratings of that diode.

 

A voltage divider would waste a lot of power. Maybe just a series resistor.

Ok good idea thanks

In the DC realm, you're right. It takes almost no current to change the state of the MOSFET's gate, and all you need to worry about is whether the gate voltage from the 555 is high enough to fully turn on the MOSFET. But that gate is like a small capacitor and it holds a charge. To repeatedly charge and discharge it at high frequency requires moving that small charge back and forth many times and that adds up to an alternating current. I'm guessing here at the numbers, but at anything less than 10kHz it's probably no concern at all. Over 100kHz it likely is a concern. In between, I'd have to do the calculation.

Hmm ok. This will take a little thinking on my part. I am flashing the LED at 40Hz, so nothing super fast. So maybe I'm OK without the series gate resistor. But I will still put in that pull-down resistor. Those never seem to hurt.

And I agree that the MOSFET source should be ground, not the drain. A MOSFET works either way, but as noted the body diode will conduct in the orientation you have shown. It can't shut off the LEDs and might burn out, depending on the LED current and the ratings of that diode.

Ok, this is making more sense now.
I want to select a new MOSFET - I think the one I chose is way overpowered for my application. So far I've narrowed it down on mouser to:
Through hole, single channel, N-channel, >2A drain current, gate source voltage > 10V, gate, and gate threshold voltage > 2V.
Anything else I need to consider (gate charge, drain-source resistance?), or can I choose the cheapest MOSFET?

 

Ok good idea thanks

Hmm ok. This will take a little thinking on my part. I am flashing the LED at 40Hz, so nothing super fast. So maybe I'm OK without the series gate resistor. But I will still put in that pull-down resistor. Those never seem to hurt.


Ok, this is making more sense now.
I want to select a new MOSFET - I think the one I chose is way overpowered for my application. So far I've narrowed it down on mouser to:
Through hole, single channel, N-channel, >2A drain current, gate source voltage > 10V, gate, and gate threshold voltage > 2V.
Anything else I need to consider (gate charge, drain-source resistance?), or can I choose the cheapest MOSFET?

Select ON resistance to be resonantly low. If you have one amp and one ohm of on-resistance, the MOSFET will need to get rid of 1 watt of heat.better MOSFETS are out there for $2 or less so pick a decent one.

 

Try IRF540..