MOSFET IC for 3.3v Signal

[iwaters]

Hi All,

Relatively new to electronics, software dev background myself, and finding it a bit tricky. I am working on a project whereby I will be using 5 MOSFETs to drive the RGBW channels on an LED strip and a separate one to control a set of dimmable LED lights. Its all working, but the MOSFETs I was using (IRLB8721) need a 5v signal and the board (Wemos D1 Mini) is 3.3v and so it wont go to full brightness. The strip is 24v and the lights are 12v.

I have looked for alternative MOSFETs that take a 3.3v input, but struggling to find any that arent surface mount (which will be ok later when I go to PCB manufacturer but not for now) or any that are in stock. So thinking a dedicated MOSFET IC would work, but I am confused what I need. Something that can handle up to 24v, 3,3v signal input, not surface mount and in stock and shippable to the UK. Any recommendations?

 

[73's de Edd]

Sir iwaters . . . . .


but . . BUT . . . BUT . . . . . . . BUT . . BUT . . BUT ! . . . . . . . this woukld be HARD ware not SOFT ware !
Sticking with your initially boughted-ed-ed and paid for IRLB8721 PWR FET and considering the on board presence of a +5V supply.
Say that you were typically experiencing a mere 20% drive . . .e.g. ~20 ma at that attempted 3.3 logic level , now expect ~100 ma / 100% . . . . . with use of a discrete full swing level totem pole driver.



Complete Da-Da Sheet
https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d462533600a40153566056732591

73's de Edd . . . . .



When your Plan "A" fails you . . . . . . remember . . . . . . there are still 25 more letters in the alphabet.

.

 

[Audioguru]

The graphs on a datasheet are for a part that has "typical" specs. But many parts have minimum or maximum specs and you do not know which when you buy them.

 

[Harald Kapp]

with use of a discrete full swing level totem pole driver.

Edd, your circuit will make matters worse: With a base voltage of 3.3 V at the 2N2222 NPN transistor the emitter will go to only ~ 2.7 V (you have to subtract V_BE = 0.6 V).
A working solution could be this well known circuit:

struggling to find any that arent surface mount

You can use them in a through hole scenario by soldering stubs of wire to the SMD pins and then use these stubs e.g. in a breadboard. Or buy adapter PCBs with an SMD footprint for the transistor and holes for header pins for the leads.

 

[iwaters]

So in that circuit above I'm splicing the 5v supply into the signal output from the transistor and feeding that to the MOSFET?

 

[Harald Kapp]

"out" is the gate drive for the MOSFET.

 

[iwaters]

"out" is the gate drive for the MOSFET.

Thanks. I'm confused how the 5v supply is fed in though. 3.3v supply and signal go to transistor, out goes to the MOSFET. But where do I put the 5v supply?

 

[Harald Kapp]

But where do I put the 5v supply?

It is shown in the schematic as source V1.

 

[iwaters]

It is shown in the schematic as source V1.

I see that. It goes through a resistor into the out from the transistor. Is that correct? I connect 5v to the out that feeds the MOSFET

 

[Harald Kapp]

Yes, as shown in my schematic. Check the waveforms in the diagram above the schematic.

 

[iwaters]

Yes, as shown in my schematic. Check the waveforms in the diagram above the schematic.

Cool, thanks for your help. That is working perfectly now.

One question though just out of interest. The duty cycle on the pins was 0 to 1024 on the Wemos board. With the circuit above its now 255 for 100% duty cycle. Doesnt matter, 255 sets the LEDs at full brightness. I was just interested why its different?

 

[Harald Kapp]

I have no explanation for this behavior. Are you sure the software is exactly the same? You haven't changed the pwm range by using analogWriteRange()?
Did you measure the duty cycle of the output signal and compare it to the input signal? Or is this just your visual impression?
Are you aware that the luminous intensity of an LED is not a linear function of the duty cycle of the controlling pwm signal due to the non linear sensitivity of the human eye? See e.g. here or here.

 

[iwaters]

I have no explanation for this behavior. Are you sure the software is exactly the same? You haven't changed the pwm range by using analogWriteRange()?
Did you measure the duty cycle of the output signal and compare it to the input signal? Or is this just your visual impression?
Are you aware that the luminous intensity of an LED is not a linear function of the duty cycle of the controlling pwm signal due to the non linear sensitivity of the human eye? See e.g. here or here.

Code is unchanged and identical. The Wemos boards as I understood had a Dury cycle value of 0 to 1024. The Arduino was 0 to 255. I noticed it because after wiring up the circuit you suggested I set it to 100, it was quite bright. I then set it to 300 and it was very bright, so then changed it to 1000 and at that point nothing changed. So then I realised that 255 was full brightness. But it works, at 255 with your circuit the led is at full brightness. Previously even at 1024 it was dim.

So all is good in terms if it doing what it needs to do

 

[Harald Kapp]

The Wemos boards as I understood had a Dury cycle value of 0 to 1024. The Arduino was 0 to 255.

You didn't mention that you switched from WEMOS to Arduino. That explains everything.

 

[iwaters]

You didn't mention that you switched from WEMOS to Arduino. That explains everything.

I didn't swap to an Arudino. Was just mentioning that for reference. Still using the Wemos.

 

[73's de Edd]

The circuit that Harald gave you puts out a normally . . . HIGH signal and pulls to ground . . . . LOW for an incoming pulse duration..
Unless changing the circuit for phase inversion . . . . . . that's the way it will be.
But you found that a software change could correct for that, if you want to compensate in that manner instead.

 

[Harald Kapp]

@73's de Edd : MY circuit is in phase, no inversion, see the signal diagram above the circuit.