HELP finding the right MOSFET

Hi guys,

I'm having some trouble choosing the right MOSFET for a project on which I need it to act as a switch.
It would switch loads of 250mA max. at a logic level of 5v. I thought about using an N-CHANNEL MOSFET in an SMD package because the currents that need to be switched are not too high.

One thing worth mentioning is that I need the MOSFET to draw as little power as possible because I'll be powering the board with 4 NiMh 1.2v, and I want to be able to have the system running for at least a year.

The MOSFETs would switch on both a Servomotor and an RC522 RFID reader when they are needed and switch them off when not.

I've chosen 7 MOSFETs, but I'm not exactly sure about which one I should choose.

BSS816NW

Vgs is pretty low, but Vds=0v -which I don't get-

BSL214N

I like this one because it has 2 MOSFETs in the same IC.

BSS306N

BSR202N

Up until here, the listed MOSFETs are pretty much the same, The only thing that changes is the Vds(th) values -by some millivolt- AND the ID value listed in the Vgs.

TSM2314

This one is beefier. Vgs value is still the same more or less, though the Id current listed in the Vgs row is way higher than in the previous ones

These next ones have nearly the same specs, the only different aspect is the previously mentioned Id -one has two times the Id than the other one-

IRLML6244TRPbF

IRLML6246TRPbF

So... After looking at all of them I have a couple of questions regarding this picture:

  • What does Vds=Vgs mean? My appliances also run on 5v.
  • What does Id mean? Does it influence my power draw? And so, Should I pick the MOSFET with the lowest ID?

N-Channel: Probably means that you have to switch GND. Are you sure that's what you want to do? If you can anwer that question, one can talk about the bottom 80% of your post.
MosFET power draw: Once switched, a MosFET does not draw any significant current. The energy needed to switch it is negligible as well in your case.
1 year on 4x NiMH: Some NiMH cells discharge a lot, make sure the ones you use don't. Approx. that would mean an max. average consumption of ~250µA@5V doable, depending on the duty cycle and consumption of the rest of the hardware.

Switching ground of a servo puts negative 5volt on the servo control pin, if the Arduino pin is LOW (default?).
Are you sure that's healthy for the servo?

I would probably switch a servo high-side, with a PNP transistor or p-channel fet.
Never had the need to do so. Maybe someone who has can chime in.
Leo..

Edit: found a relevant post here.
https://forum.arduino.cc/index.php?topic=373905.0

Hi,

The MOSFETs would switch on both a Servomotor and an RC522 RFID reader when they are needed and switch them off when not.

How often will the project be activated?
Do you know how much current a servo needs?
Do you know how much your RFID reader needs.

12 months from NiMH, you need to do some research about them, they have a self discharge property or shelf life.

Tom.... :slight_smile:

alex2000lopez:
One thing worth mentioning is that I need the MOSFET to draw as little power as possible because I'll be powering the board with 4 NiMh 1.2v, and I want to be able to have the system running for at least a year.

Typical NiMh batteries self-discharge 10-15% in the first 24 hours, then 10-15% per month. Use alkalines instead. Rechargables are great performers for short term high current applications (motors, RC cars), alkalines and other disposables shine in long term, low current applications (remote controls, alarm clocks, powering Arduinos for a year or more).

  • What does Vds=Vgs mean? My appliances also run on 5v.
  • What does Id mean? Does it influence my power draw? And so, Should I pick the MOSFET with the lowest ID?

As you failed to include the table header it's a bit of guess work here, but I think that column is about test conditions. So where Vds = Vgs and Id = 30 uA. After all, exact values for Vgs and so vary a little bit with Vds and Id - see the graphs in the data sheet for actual measurements at different currents and voltages.

wvmarle:
Typical NiMh batteries self-discharge 10-15% in the first 24 hours, then 10-15% per month. Use alkalines instead. Rechargables are great performers for short term high current applications (motors, RC cars), alkalines and other disposables shine in long term, low current applications (remote controls, alarm clocks, powering Arduinos for a year or more).

That's rather out of date. Currently available NiMH LSD (Low Self Discharge) types do much better than that. Eneloops are the classic example but there are now many others. Typical discharge rate is around 15% per year. Most still retain 90% charge after 6 months. And they can handle the same high currents as conventional NiMH. Cost a little more than basic NiMH and have slightly lower capacity for the size but well worth it for many applications where alkaline would previously have been preferred.

Steve

couka:
N-Channel: Probably means that you have to switch GND. Are you sure that's what you want to do? If you can anwer that question, one can talk about the bottom 80% of your post.
MosFET power draw: Once switched, a MosFET does not draw any significant current. The energy needed to switch it is negligible as well in your case.
1 year on 4x NiMH: Some NiMH cells discharge a lot, make sure the ones you use don't. Approx. that would mean an max. average consumption of ~250µA@5V doable, depending on the duty cycle and consumption of the rest of the hardware.

Thanks for the reply,

Yeah, after looking online I've realised that a P-Channel is better for switching the servo, though I don't know what type I should use for the RFID chip.

As slipstick said, NiMH batteries have improved over the years. Eneloop promises a 15% self-discharge in a year period.

Wawa:
Switching ground of a servo puts negative 5 volts on the servo control pin if the Arduino pin is LOW (default?).
Are you sure that's healthy for the servo?

I would probably switch a servo high-side, with a PNP transistor or p-channel FET.
Never had the need to do so. Maybe someone who has can chime in.
Leo..

Edit: found a relevant post here.
Switching power for a servo - General Electronics - Arduino Forum

Thanks for the reply.

I'm going to use a P-Channel + NPN config in order to switch the servo. It's what other posts about this topic suggest.

Something like this:

I've decided to use this P-CHANNEL SMD Transistor:
AO3403

Now, I have to decide NPN transistor I should use. I've thought about using an SMD one. I've picked up 2 options. Feel free to recommend others:

MMBT489LT1-D-112212

NSS30101LT1G-D

The only thing left to discuss is the switching of the RC522 RFID module.

  • Should I use a FET?
  • If so, Which one should I use?

Thanks in advance,
Alex

TomGeorge:
Hi,
How often will the project be activated?
Do you know how much current a servo needs?
Do you know how much your RFID reader needs?

12 months from NiMH, you need to do some research about them, they have a self-discharge property or shelf life.

Tom.... :slight_smile:

Thanks for your reply,

The servo that I'm using peaks at around 500mA and the RFID chip peaks around 30mA.

Which type of transistor should I use to power off the RFID reader?

Thanks in advance,
Alex

BTW, I've linked the schematic of the project :slight_smile:

The diagram has no collector resistor.
Use 1k for the collector and 10k for the base.

Are you sure the RFID module can't be put into sleep mode.
If not, high-side switching could be bad there if it's I2C with normally high pins.
Leo..

For the RFID reader you may be able to switch it by using one of your output pins as power source - assuming it draws less than 20 mA.
Come to think of it, it wouldn't work if it's I2C. If connecting the GND line to a pin you switch it off by setting it HIGH and then you may supply too much current to the I2C (maybe a diode would help here) - whenever something tries to pull low the line. If switching by pulling the Vcc line to LOW you may end up with serious leaking through the I2C pull-up resistors.