hello all. i am considering driving a p - channel mosfet with an arduino mini 3.3v (generic). i have included the schematic for reference. my question is; would this circuit cause any problems to the circuitry of the arduino because it will be sinking 4.2v instead of 3.3v? the current is tiny but the voltage is higher. i know that people use mosfet drivers to solve this problem but i thought since the difference is very small it may be ok and i want to keep the circuit as simple as possible.I have tested the circuit and it works fine, i am just wondering whether prolonged use may cause any damage. i know that the atmega chips are the same on the 5v arduino and the 3.3v version so it seems like it shoiuld be ok unless it goes above 5v?
any help will be much appreciated, thanks.
ps. please ignore the non embedded attachment, completely unrelated....
You must not apply a voltage > Vcc + 0.5V to any input or output.
jremington:
You must not apply a voltage > Vcc + 0.5V to any input or output.
why is this since the chips are the same?
With those resistor values the clamping diodes will have no problem taking care of the excess voltage.
That said, you're going to have a problem switching on that MOSFET as you will bring the gate to just -2.1V.
wvmarle:
With those resistor values the clamping diodes will have no problem taking care of the excess voltage.
That said, you're going to have a problem switching on that MOSFET as you will bring the gate to just -2.1V.
the circuit seems to work fine...it was the prospect of damaging the board which i was mainly concerned about.
i just measured Vgs and it is 0.49V. the Vgs threshold on the datasheet is minimum -0.5V.
Does this mean that i am right at the minimum for keeping it off?
Yeah, borderline. Would not bed any money on it.
Why not simply switch it low side with a N-mos? Much much much easier.
septillion:
Yeah, borderline. Would not bed any money on it.
Why not simply switch it low side with a N-mos? Much much much easier.
because im switching an mp3 module which leaks current through its serial port when its low side switched (its a battery powered circuit).
ive just changed the pull up resistor to 470k and its now dropped to 0.39V. would you say that is an acceptable value?
Is there any chance you could specify the relationship between the resistor values and Vgs? Thanks.
You could use an n-channel MOSFET to drive your p-channel MOSFET, as shown in Fig 3 of this application note.
(There is no need for the zener diode - it's there for protecting the gate of the p-channel MOSFET at much higher supply voltages)
At the cost of an extra FET, it will solve all your problems
You need an N-channel to switch your P-channel FET. (Thanks John, who posted while I was typing.)
What you are doing now is running the P-channel in its linear range, which causes it to heat up and you won't get 4.2V on the output. Depending on the current drawn by the MP3 module this may actually work, but it doesn't seem like you have done the calculations to be sure that it works.
Note that "it works when I tested it" isn't the same as "works reliably for the life of the device." The thing may be heating up only a little now, but if you leave it in your car on a hot day and then switch it on, it may overheat. So a year from now, "It blew up for no reason!" is the result.
You have a very simple voltage divider. Work out the two output voltages for yourself: with either 3.3V or 0V on the bottom end and 4.2V on the top end, the voltage difference between Gate and Source can be easily calculated.
You can even remove one of the resistors that was shown in Fig 3.
thanks for the help on this everyone. i didnt want to use an n channel to drive the p channel purely because it increases the complexity and with my original design which used only n channel drivers i could accomplish the same thing with the same amount of mosfets.
MorganS:
You need an N-channel to switch your P-channel FET. (Thanks John, who posted while I was typing.)
What you are doing now is running the P-channel in its linear range, which causes it to heat up and you won't get 4.2V on the output. Depending on the current drawn by the MP3 module this may actually work, but it doesn't seem like you have done the calculations to be sure that it works.
Note that "it works when I tested it" isn't the same as "works reliably for the life of the device." The thing may be heating up only a little now, but if you leave it in your car on a hot day and then switch it on, it may overheat. So a year from now, "It blew up for no reason!" is the result.
the mosfet is able to withstand up to 30V, 4A. this is miles higher than the, probably absolute max 300mA, 4.2v which my circuit will need. if i was using a lot of power your concerns of overheating would certainly be an issue but in my circumstances is it something which realistically would be an issue?
Ideally i dont want to have to use a second mosfet as it increases the complexity of the circuit. of course if its imperative then so be it. thanks again all.
the mosfet is able to withstand up to 30V, 4A.
...only if it is driven correctly and certainly not both at the same time.
MorganS:
...only if it is driven correctly and certainly not both at the same time.
ok, but the heat is due to the current passing through the mosfet correct? so my low current and voltage demands should cause little heat up even if it is running in 'linear range' ?
Did you do the calculation to check this? Do you need help doing the calculation?
If a 4A rated MOSFET heats up noticeably with <300 mA then there's no calculation needed. You're firmly in the linear range, and probably drop 0.5-1V or even more over that MOSFET (measure the voltage between source and drain pins).
It's supposed to have an on resistance RDS(ON) of <85mΩ at VGS =-2.5V, giving you a VSD or <25 mV at 300 mA, and a dissipation of <8 mW. That won't cause any noticeable heating.
If you switch it properly you have a VGS of -4.2V, and the on resistance is even lower.
MorganS:
Did you do the calculation to check this? Do you need help doing the calculation?
aha, i just did them and they came pretty close. very satisfying! this helps me with reducing the gate voltage but will it help me with the question of whether my low current demands will warm the mosfet to a dangerous level?
wvmarle:
If a 4A rated MOSFET heats up noticeably with <300 mA then there's no calculation needed. You're firmly in the linear range, and probably drop 0.5-1V or even more over that MOSFET (measure the voltage between source and drain pins).
It's supposed to have an on resistance RDS(ON) of <85mΩ at VGS =-2.5V, giving you a VSD or <25 mV at 300 mA, and a dissipation of <8 mW. That won't cause any noticeable heating.
If you switch it properly you have a VGS of -4.2V, and the on resistance is even lower.
i have measured Vds and it hardly dropped at all. should i now use the voltage divider equations to get the voltage as close to 0 as possible? are there any instruments for measuring the heat of components? thanks again
gonadgranny:
are there any instruments for measuring the heat of components? thanks again
Yes - most people have 10 of them. They're called fingers.
I thought you mentioned that you actually do get heating of your part - putting a finger on a part is the easiest way to check whether it heats up (but be careful: it can be very hot).
gonadgranny:
i have measured Vds and it hardly dropped at all. should i now use the voltage divider equations to get the voltage as close to 0 as possible? are there any instruments for measuring the heat of components? thanks again
So the MOSFET is just barely switched on and Rds is high?
Yes. A thermometer can show the temperature of your components. I use a small infrared thermometer when I need to check this kind of thing without burning my fingers.
MorganS:
So the MOSFET is just barely switched on and Rds is high?
Yes. A thermometer can show the temperature of your components. I use a small infrared thermometer when I need to check this kind of thing without burning my fingers.
no when the mosfet is switched on it read 0.1V between d and s....would a smaller resistor help improve this perhaps?
yes i have one of those also, its just a bit difficult to know whether its actually trained on the part. i was envisioning some kind of probe with a soft end which you can push onto it. if they dont exist then maybe thats an invention there!
