# Why is my 555 PWM circuit limiting the current to my 1W LED load?

#### FranciscoB

Joined Feb 8, 2014
85
Initially I had a resistor in series with the LED to limit the current but I noticed it was not very bright. I decided to experiment and remove the resistor, connecting the LED directly to the 5V. I only measured close to 200mA passing through it, when it should be near 300mA.

The LED is sold as a 1W bead with Vf of 3.6V and 320mA current at this voltage.

#### Attachments

• 1.3 MB Views: 9

#### sagor

Joined Mar 10, 2019
786
An IRF540 needs a higher gate voltage to be fully turned on. Your 5V supply and drive from the 555 may not be enough to turn on the MOSFET fully

#### FranciscoB

Joined Feb 8, 2014
85
An IRF540 needs a higher gate voltage
Higher than?
What I understand from the datasheet, it should be 2V to 4V. With my cheap digital multimeter I'm measuring 3.2V at Pin3 of the 555IC.

#### BobTPH

Joined Jun 5, 2013
6,076
Look at your datasheet again, the threshold voltage is where the MOSFET just begins to conduct. Look at the spec for the on resistance to see what voltage is required to turn it fully on.

Bob

#### FranciscoB

Joined Feb 8, 2014
85

#### BobTPH

Joined Jun 5, 2013
6,076
At 4V, that MOSFET is guaranteed to piass only 250 uA. Thay is what is called the threshold voltage.

The place in the dataset to filnd what voltage fully turns it on is the spec for Rdson. In this case, that is 10V. The curves on page 3 show a typical transistor. Based on the range given for the threshold, those curves might be off by as much as 2V.

Bob

#### KeithWalker

Joined Jul 10, 2017
2,616
If you change the drive circuit to turn the MOSFET fully on, make sure you have the limiting resistor in series with the LED or you will fry it If the 5V supply can deliver enough current.

#### Audioguru again

Joined Oct 21, 2019
5,413
The threshold voltage is when the Mosfet is barely turned on and is almost turned off at 250uA. The IRF540 has 2V to 4V so some of them conduct only 250uA when the Vgs is 4V.
The graphs are for a "typical" one but you get what is available which might be the minimum which needs a Vgs of maybe 8V to conduct 1A.

The output of an old 555 with a 5V supply does not reach anywhere near 5V, Some go as high as 3V. Then some IRF540 Mosfets will do nothing.

Use a IRL540 Mosfet instead, and some of them conduct 250uA when Vgs is 1V and others need 2V. A sensitive one will conduct 10A when the Vgs is 4V so use a current-limiting resistor in series with the LED.

#### FranciscoB

Joined Feb 8, 2014
85
Ok, this is working as it should, now with current-limiting resistor and logic-level p-mosfet. Thank you so much for the tips!

#### Attachments

• 233.6 KB Views: 4

#### dl324

Joined Mar 30, 2015
15,445
working as it should, now with current-limiting resistor and logic-level p-mosfet
You have the source and drain of the MOSFET reversed. As drawn, the parasitic diode would be forward biased and the LED would always be on. The 47 ohm resistor isn't necessary.

#### Audioguru again

Joined Oct 21, 2019
5,413
When wired correctly, some of those P-channel Mosfets will never turn off when driven from an old 555 because the gate voltage will never go high enough. A modern Cmos 555 will be needed.

#### FranciscoB

Joined Feb 8, 2014
85
Im sorry for the mistake it is incorrectly drawn but working nicely.

#### dl324

Joined Mar 30, 2015
15,445
Im sorry for the mistake it is incorrectly drawn but working nicely.
No problem. You said it was working, so it was only a matter of drawing the MOSFET upside down.

If you use a symbol like the AO3401 datasheet uses, it's easier to not get source and drain connected wrong:

The source on a P channel MOSFET is usually connected to the higher voltage and the source on an N channel is connected to the more negative voltage.

When I draw MOSFETs by hand, I never use the dashed line to indicate enhancement mode. Neither did any of the electrical engineers I worked with for decades. It took too long to draw and the only depletion mode MOSFETs we used in our integrated circuits were decoupling caps. Things might have been different with only NMOS and the NMOS loads were depletion mode.

Last edited: