Hi,
Can anyone here think of a good logic level P channel MOSFET that can deal with about 40v, 1A?
Have been searching the internet for hours and thought to try here..
Thanks!
diode:
Can anyone here think of a good logic level P channel MOSFET that can deal with about 40v, 1A?
I think most of them can handle that much power. Probably quite a few BJTs, too.
You can do parametric searches at both DigiKey.com and Mopuser.com.
At DigiKey I put in the search term 'logic level mosfet'.
Then clicked on "FETs - Single".
Then under FET Type selected "MOSFET P-Channel"
Under Vdss I selected for 40V,45V,or 50V
You can figure out the rest.
I'd say nearly all of them. It'd be more trouble to find one than CAN'T handle that than one that can.
1A at 40V is nothing for a MOSFET, provided they have a cooling rib.
Here's a brief list from Vishay: http://www.vishay.com/mosfets/p-channel/v-ds-gteq-neg31-v-lteq-neg80-v/
Thanks all! Will look at those resources - there are so many.
Was just wondering if there is one or some that are particularly good / easy to interface / power efficient etc.
(like for example the IRF540 N Channel seems to come recommended as an N Channel)
Reading into it also seems they have become much better over time.
ps. yes the IRF520 is N.
You are using this as a high side switch device?
To drive from an Arduino pin, you will need NPN or logic level N-channel to pull the P-channel's gate low to turn it on, with pullup resistor to 40V to turn it off.
I posted a circuit yesterdy to do that, for driving a halogen lamp.
CrossRoads - I want to connect them to a TLC5940. Am presuming a P channel MOSFET can connect straight to a TLC.
TLC5940 datasheet says it needs some minimal load to switch properly.
Use a low enough value resistor as the Pullup resistor to 40V to meet that.
The TLC5940 will act as the NPN transistor here.
Have not tried this myself, wire up just 1 MOSFET to start and see how it works.
Thanks for your thoughts!
Not sure what you mean with the minimal load?
Haven't got a P channel MOSFET at hand so have to get one and try tomorrow.
What exactly are you trying to do with the mosfet ? if your switching loads then low side switching is best if possible because N channel devices are cheaper and have better characteristics.
sparkylabs:
What exactly are you trying to do with the mosfet ? if your switching loads then low side switching is best if possible because N channel devices are cheaper and have better characteristics.
But a TLC5940 is already a low-side PWM controller that sinks current. So for a matrix, as an example, if the columns are cathodes you use the TLC9540 to control the columns. This leaves you to control the rows (anodes), on the high side. How do you use N-channel for that? The general way to solve that from what I have seen is P-channel MOSFETs or maybe darlingtons on the high side to source current.
yes in that case it has to be high side. unless the LED's are powered from 5v he needs level shifting too as suggested above
Minimal Load:
See Page 14 of data sheet:
The maximum output current per channel is programmed by a single resistor, R(IREF), which is placed between
IREF pin and GND pin. The voltage on IREF is set by an internal band gap V(IREF) with a typical value of
1.24V. The maximum channel current is equivalent to the current flowing through R(IREF) multiplied by a factor of
31.5. The maximum output current per channel can be calculated by Equation 6:
Imax = V(iref)/R(iref) x 3.15
where:
V(IREF) = 1.24 V
R(IREF) = User-selected external resistor.
Imax must be set between 5 mA and 120 mA. The output current may be unstable if Imax is set lower than 5 mA.
Output currents lower than 5 mA can be achieved by setting Imax to 5 mA or higher and then using dot
correction.
If you are choosing a MOSFET by drain-current parameter you are probably going about things
the wrong way.
No-one would normally take a MOSFET near its max rated current value - it will need LARGE heatsink
and will dump tonnes of power.
Use the Rds(on) figure to decide if a device is suitable - you can calculate the power dissipation and
voltage drop directly from that - make sure both are sensible. Also remember power goes as the
square of current so that 10A is four times worse than 5A for the same device - this really can bite
the unwary. Do the maths to avoid surprises.