Continue to Site

# Driving an N channel FET?

Status
Not open for further replies.

#### bigfarmerdave

##### New Member
If I want to drive a generic N channel FET, say an Infineon IPB70N10SL-16. This is a 100V, 70 amp, 16mOhm RDSonn part. If I use it as a high side switch to a 12 volt source, I know that I need my gate voltage to be higher than the 12 volt rail, but what I don't know is how to calculate/find how much current I'll need to drive the gate. Any help is appreciated.

Is speed important?

What's this for?

You could bootstrap it but the duty cycle can never be 100%.

Speed is not important.
I'll be using it to control a 12 volt solenoid.
How do you calculate what the max duty cycle would be?

Duty cycle of a solenoid?

Max duty cycle is probably going to need to be about 85%

Max duty cycle is probably going to need to be about 85%

How did you arrive at that number?

Right now it is a rough estimate. We anticipate that with an 85% duty cycle we should be able to open the solenoid up more than what we think will be needed to get the max oil flow that we'll ever need.

Solenoids are more of an all on or all off affair. Do you have a model number?

Unfortunately, I don't actually have any model number in front of me right now. In response to a solenoid being "all on or all off", I want to regulate oil flow and turning the solenoid all on or all off would be like flipping a switch, all or nothing. I want to regulate oil flow more precisely than that.

Then a solenoid isn't going to do it, PWM on a relay or solenoid is handy for reducing power required to operate it as the holding current is lower than the pull in current.

The amount of current needed is the Qg / switching time. Your mosfet has a 160nC gate charge. If you wanted a 1us turn on, then your gate current needs to be...

160nC / 1000ns = 0.16A.

Obviously, if you want it to turn on faster, you'll need to push more current.

Last edited:
Smanches, That's what I was looking for. Thanks a lot!

Unfortunately, I don't actually have any model number in front of me right now. In response to a solenoid being "all on or all off", I want to regulate oil flow and turning the solenoid all on or all off would be like flipping a switch, all or nothing. I want to regulate oil flow more precisely than that.

How about several solenoids in parallel lines to incrementally increase the flow?
ie:
1/4, 1/2, 3/4, or full, or isn't that accurate enough?

PatM

Just PWM it like I think you're trying to do. I've done this exact same thing to implement a soft turn-off for hydrolic systems so the rams don't slam when it's shut off.

You need to find the inductance of the coil though.

Last edited:
But you still won't be able to control the flow rate. It'll be either on or off.

But you still won't be able to control the flow rate. It'll be either on or off.

No, but if he turns the solenoid on for 15 seconds and then off for 45 seconds, it will deliver 25% of the flow rate compared to if were left on. As long as the cycle rate is low, this is a valid way of controlling the average flow rate.

Depends on what system he is working on. For example, if it is just going into a resevoir or something like that. Then hey no prob. Now if it is in an application where a constant stream is needed, just at a lower rate, then problems.

As a chemist, we encounter this situation frequently. Some reactions respond fine to say addition of 10 mL of a reagent every 15 minutes (for a total of 45 min). Other reactions may require the same amount of reagent over the same amount of time, but the reagent has to be added drop-wise, otherwise bad stuff can happen (no product, rapid boiling and in extreme cases explosions).

I've actually done it before (PWMed a solenoid) and it works rather well to regulate oil flow. I just don't want it BANG!!! on or BANG!!! off but would rather ramp it up and down in a nice controlled manner like Smanches described.

Status
Not open for further replies.

Replies
14
Views
1K
Replies
7
Views
1K
Replies
166
Views
20K
Replies
6
Views
1K
Replies
22
Views
1K