Do my logic level mosfets still require a gate driver?
Capt. Quirk
Posts: 872
I bought 30 FDS6670a N-channel logic mosfets, and I am not sure how well they will work with the propeller? (1) Mainly, will they require a gate driver despite being a logic level mosfet?
I am trying to make a 3 phase h-bridge, with small homemade motors using 3.3v power supply. My future mosfets will need perform with whatever wire size & pwm frequency necessary, to make sufficient torque for a micro sumo bot. But for now, I am experimenting with my FDS6670A mosfets.
(2) In a post from 2009, Mike Green said:
“The Propeller provides about 3V for a HIGH output voltage and this isn't enough for many MOSFETs, even those that are designed for logic level switching. You need a gate to source voltage that's well below 3V. Typically the datasheet for the MOSFET will have a graph for the amount of current switched vs. the gate voltage. The gate voltage for the amount of current you want to switch has to be well below 3V”
My impression from reading app-notes and articles on the subject of choosing a mosfet; I thought the Vgs(th) was the most important factor. That it represented the logic voltage required from the Propeller to open the gate. But from what Mike said, and what the name implies; it is the difference between the voltage supplied to the gate, and the voltage that is going through the mosfet’s source pin, to ground. Or it is the residual voltage at the mosfet’s source pin, before the gate opens?
Mike suggested the FDFS6N754 as an example. Compared to my mosfet the Vgs(th) is similar, but Vgs is much different.
(3) In another post from 2010, Phil Pilgrim stated:
“For PWM-ing, the MOSFET's gate capacitance is also a spec that needs consideration. If the gate capacitance and PWM frequency are high, a MOSFET driver will be required for efficient high-speed switching”.
I assume he is referring to “Qg” and detailed in fig. 7 and possibly fig. 8 of my data sheet. In Mike G’s FDFS6n754 data sheet, the Qg values are 1/10 my mosfet. Do these lower values represent a normal range for my application?
(4) Lastly, while I am experimenting with mosfets and h bridges, should I use a use an opto isolator or a series resistor to protect the Propeller? Especially if I start with 7 to 12 volts, and work down to 3.3volts.
Thanks
Bill M.
I am trying to make a 3 phase h-bridge, with small homemade motors using 3.3v power supply. My future mosfets will need perform with whatever wire size & pwm frequency necessary, to make sufficient torque for a micro sumo bot. But for now, I am experimenting with my FDS6670A mosfets.
(2) In a post from 2009, Mike Green said:
“The Propeller provides about 3V for a HIGH output voltage and this isn't enough for many MOSFETs, even those that are designed for logic level switching. You need a gate to source voltage that's well below 3V. Typically the datasheet for the MOSFET will have a graph for the amount of current switched vs. the gate voltage. The gate voltage for the amount of current you want to switch has to be well below 3V”
My impression from reading app-notes and articles on the subject of choosing a mosfet; I thought the Vgs(th) was the most important factor. That it represented the logic voltage required from the Propeller to open the gate. But from what Mike said, and what the name implies; it is the difference between the voltage supplied to the gate, and the voltage that is going through the mosfet’s source pin, to ground. Or it is the residual voltage at the mosfet’s source pin, before the gate opens?
Mike suggested the FDFS6N754 as an example. Compared to my mosfet the Vgs(th) is similar, but Vgs is much different.
(3) In another post from 2010, Phil Pilgrim stated:
“For PWM-ing, the MOSFET's gate capacitance is also a spec that needs consideration. If the gate capacitance and PWM frequency are high, a MOSFET driver will be required for efficient high-speed switching”.
I assume he is referring to “Qg” and detailed in fig. 7 and possibly fig. 8 of my data sheet. In Mike G’s FDFS6n754 data sheet, the Qg values are 1/10 my mosfet. Do these lower values represent a normal range for my application?
(4) Lastly, while I am experimenting with mosfets and h bridges, should I use a use an opto isolator or a series resistor to protect the Propeller? Especially if I start with 7 to 12 volts, and work down to 3.3volts.
Thanks
Bill M.
Comments
min 1
typical 1.8
max 3 V
You are set to go, as the typical 1.8v should be the norm
This comes up all the time. Protection from what?
We have people concerned with "protecting their pins" on the one hand and these bizarre testimonies as to the Propeller's "robustness" and being "rock hard" when it comes to dead shorts and absolute stupidity on the other hand.
It's amazing.
What is the fear? Not knowing what you're doing? I'm sorry, but opto-isolation won't help.
If a guy can do something dopey on the output side of a opto-isolator what keeps him from just as easily doing a dopey thing on the input side of an opto-isolator ("protection" foiled)?
-Phil
You are talking about why people use a N-channel mosfet in place of a P-channel mosfet. That the Rds value of a N channel is lower. But in this case, the high side mosfet driving the load would require a driver to increase the gate voltage in order to obtain a lower Rds value of my mosfet?
And the low side mosfets would work fine, as they are?
When choosing the high side mosfet; the "drain-source on-resistance vs drain-current," graph is the most important value. Is there a "ballpark" acceptable Rds value for a high side mosfet connected directly to the Propeller? Since many companies tend to boldly display this value on their data sheets.
And does the forum editing tools create those nice subscripts, or do you import them from word?
Thanks
Bill M.
No. You have to take the RDS value from the graph, multiply it by the square of your maximum current, and make sure the result does not exceed the MOSFET's power dissipation rating. It's better if it's no more than half that -- especially if you're PWMing the MOSFET.
I've found that interfacing a micro to a load via a MOSFET requires some design work. Due to the tradeoffs, there are no "cookbook" recipes. Although it would be easy to way overspecify the MOSFET, you then run into the problem of high gate capacitance and how to drive it.
-Phil
Sincerely
Bill M.