Hi,
I'm putting together a simple circuit that flashes an LED using input from a sensor. This circuit works, and I've successfully built it and used an LED strip instead of a single LED. The problem is that if the LED strip contains too many LED's, or the device is run for an extended period of time, it burns up the chip. I've tried putting an transistor into the circuit to power the LED's, but can't get it to work. Can anyone offer any assistance on adding a mosfet to the circuit? I'm thinking that the mosfet would control the power going to the LED's, and prevent too much current from frying the chip.

I understand the basics of circuitry, but I'm a biologist not an electronics engineer.

Thanks,

 

Use an N-MOSFET.
Connect the source to common, the gate to the LTC1050 output, and the drain to the LEDs.
This will reverse the polarity of the signal, so you will also need to interchange pins 2 and 3 on the LTC1050.

 

Something like this?

Also, how do I calculate the size of MOSFET to use?

Thanks,

Use an N-MOSFET.
Connect the source to common, the gate to the LTC1050 output, and the drain to the LEDs.
This will reverse the polarity of the signal, so you will also need to interchange pins 2 and 3 on the LTC1050.

 

Something like this?

Source grounded, drain to LED cathode, put the current limiting resistor back.

Also, how do I calculate the size of MOSFET to use?

Pick one with an appropriate threshold voltage, max current and drain-source voltage won't be an issue at 12V.

 

Thank you.

Does this do the trick?

I appreciate you putting up with the questions this ol' scientist has.

Source grounded, drain to LED cathode, put the current limiting resistor back.
Pick one with an appropriate threshold voltage, max current and drain-source voltage won't be an issue at 12V.

 

Also, how do I calculate the size of MOSFET to use?

Pick one with at least twice the current rating of the LED current.
With a 1k series the resistor the LED current is <12mA, so just about any N-MOSFET should work.
A 2N7000 is good up to about 100mA (200mA max).

 

Does this do the trick?

Nope.

In your circuit, there is no power to the LED.

Your schematics would be easier to read if you drew them more conventionally:

Using package outlines typically obscures functionality.

I substituted a different opamp so pinout is different. The MOSFET is just the first N channel device I clicked on.

 

Thanks dl324,
Apologies for the schematics, I'm using EAGLE software to draw them and still learning how to do it.

You help is greatly appreciated.

 

Back to your first schematic, at 10-12 mA sink current the chip should be ok. R3 should limit the LED current no matter how many there are. Are you using a DIP or surface mount part?

ak

 

Thanks AK,
Problem is I will be running up to ~0.4 A spikes, with most of the time no load. DIP part, surface mount later. I have ran it successfully at this amperage, but the poor ltc1050 chip can't handle it and eventually burns out.

 

I'm using EAGLE software to draw them and still learning how to do it.

I use EAGLE too. I use an older version most of the time because the only thing I like about the newer versions is being able to have selective overbar in text. They made too many cosmetic changes and I've invested a lot of time with the old version and don't care to get used to the newer versions.

Most of us know color coding schematics is a waste of time. I print to PDF and post clips from them.

 

TProblem is I will be running up to ~0.4 A spikes,

Not with the schematic in post #1.

ak

 

Not sure what you mean here, as I did run the circuit with a string of LED's which took ~0.3 to 0.4 A to operate.

Not with the schematic in post #1.

ak

 

With a 12 V source, how did you get 0.4 A through a 1K resistor?

ak

 

I don't know, I just know that it worked until the chip toasted.

With a 12 V source, how did you get 0.4 A through a 1K resistor?

ak

 

That may be a clue.

ak

 

Not actually a 1k resistor?

 

Sure looked like a 1k to me.

Not actually a 1k resistor?

 

Did you measure it?

 

Apparently he must have bypassed the 1K resistor when connecting the strip.
SG