mosfet voltage drop

Hi

from a long time I am using that schematic : http://adam-meyer.com/arduino/N-Channel_MOSFET
using a RFP50N06 instead of the RFP30N06 (which is mostly obsolete).

I made several set-up with Teensy3.2 or LC with that circuit and work as it should for driving DC motors or others DC stuff.

Lately I want to control (12v) solenoids using a Teensy4 and I get a big voltage drop: voltage at the solenoid wire where +/- 3v. If I'm not attaching the solenoid the voltage rise to the 12v.
if I mount a DC motor instead, the motor turns correctly and I measure well +/- 12v.
What made the use of the solenoid wrong!

What made me really confused is when I plug the solenoid on an old set-up running from several years with that same mosfet circuit and a LC, it work ! the solenoid is 'moving'...

I see that LC and T.4 have different PWM frequencies. Could it be related to my issue ?

As the mosfet assembly was just soldering, I suspected my circuit and plug an Arduino micro. it worked straight away.

datasheet says for 50N06 : Gate to Source Threshold Voltage :: Min : 2v / Max 4v
Is the teensy4 3.3v too 'low' for that Mosfet ?
If yes ! do you have any N-channel replacement suggestion ! (thanks...)
I would like to keep the T.4 as I need usbMidi and almost 12PWM... seems easier to change the mosfet !

During my tests the teensy was powered from the Usb cable plug on my computer... All the set-up I have made before where external powered, I mean the teensy is not powered from the Usb plug but from an external 5v. Does that could change something ?


thanks for your feedback !
best
0-0
 
Last edited:
The basic issue as you guessed is that the 3.3V drive is not high enough to fully turn on that MOSFET while the 5V was. If the MOSFET only partially turns on, it starts dropping a lot of voltage when you drive something that requires a lot of current.

The easiest path would be to use something like a transistor or other logic level shifter like a 74HCT125 to translate the drive signal from 3.3V up to 5V. Finding high power MOSFETS that are happy with only a 3V drive is difficult.
 
thanks for your reply.

74HCT125 seems to be out of product but texas Instrument have replace it with CD74HCT125.

if I well understand:
- OE pins / Gnd have to be connected together and join to GND.
- VCC is +5v supply
- A pins are 3.3v logic input (PWM from teensy)
- Y pins are 5v out (to mosfet)

is that rigth ?
thanks
0-0
 
The CD74HCT125 will work fine and your basic pinout looks correct.

Since the device has 4 separate channels, if you have any unused A inputs, they should also be tied to GND or VCC and the outputs left unconnected. Basically just don't leave inputs floating per the datasheet.
 
As others have said, you could use a logic chip to change the voltage.
Or you could use two transistors in a row. You said that the current part is turning on partly, enough for low load signals. If you used one transistor and a pull up resistor to 5V you could create a signal that toggles between 0 and 5V. You then use that to drive the gate of the second transistor. This isn't a very nice solution, isn't very power efficient and may have issues at high switching speeds but has the big advantage of only requiring parts you have already plus a resistor.

The cleanest option would be the to change the FET to one that works with your voltage levels. Most component suppliers will have a parametric search that you can use to find a suitable part.
e.g. https://uk.farnell.com/w/c/semiconductors-discretes/fets/single-mosfets?channel-type=n-channel&range=inc-in-stock
You can then filter by threshold voltage, package, max current etc... to find parts that meet your requirement. Just about any supplier worth using will have a search option like that, who your options are will depend on where in the world you are.

With a 3.3V output you really want the listed threshold voltage to be 2.5V or less, you don't want to be anywhere close to the threshold if you're trying to use the part as a switch for anything high current. If you want to be 100% certain the distributor sites will also have links to the datasheets which will include a chart showing the on resistance (RDS on) for different gate voltages.
 
Hi AndyA
thanks for your reply.

The only one Mosfet I could find yet is that one: https://www.mouser.fr/datasheet/2/196/Infineon_IRLB8743_DataSheet_v01_01_EN-1228483.pdf

data sheet :
VGS : gate threshold voltage 1.35 (min) / 2.35V
RDS(on): 3.2v
Where I'm lost is on suppliers list most of the time RDS is express with resistance (ohm) and I haven't the knowledge for.

After reading lots of other thread, it seems a recurrent 3.3v logic issue...
I think I will go 1st with logic level shifter ! probably with the 74HCT125 which seems the more efficient for my purpose and the easier to implement...

Best
0-0
 
You should have lots of options beyond the part you found. After applying some rough filters mouser is showing me over 900 possible options. Not all of them will be suitable but I'm sure plenty would be.

The first item in the list is: https://www.mouser.co.uk/datasheet/2/408/TK4R3E06PL_datasheet_en_20210120-1075493.pdf
Looking at the chart on the middle left of page 5 of the data sheet at 25C with a gate-source voltage of 3V it can pass around 10A with (and this is extrapolating a bit from other charts) an RDS(on) in the region of 10mOhms.
Without knowing the full specs of the parts you're using it's hard to be sure but that looks like it should work for you.

Or looking a little lower https://www.mouser.co.uk/ProductDetail/onsemi-Fairchild/NDP6060L?qs=7reVrgTiN3Za28IzvP1Tjw== looks to be able to cope with 20A with a 3V gate voltage abut it's hard to see exactly what the typical RDS(on) would be in that situation.


RDS(on) is always measured in ohms, it is short for Resistance - Drain to Source when on. Resistance is measured in ohms.
If RDS(on) was 100 mOhms (0.1 Ohms) and you were trying to drive a device that normally took 12V / 1A then you would lose 0.1V in the transistor (Voltage = current * resistance) leaving 11.9V for the device. OK, the devices current draw may now be a bit lower now because the voltage to it has dropped to 11.9V but if the drop due to the transistor is enough to changes the results meaningfully then you're using the wrong transistor.
RDS also matters for power reasons. Using the above example with 1 A current and 0.1 ohms RDS the power lost into the transistor will be 0.1 Watts (Power = current squared * resistance). This gives you an idea of both how much power is getting wasted and also how hot the transistor is going to get.


Yes, this can be an issue when using 3.3V parts but only because the people supplying the drivers are using the cheapest, easiest to source parts they can find.
 
Back
Top