MOSFET as a rheostat is it safe?

[Chriss]

Hi!
I have a small task to finish, to make a kind of rheostat to drive a solenoid valve.
The solenoid valve is 12V, when the valve is max opened, applied 12V it takes around 2.5A.

Here is what I have to do:
The valve should be controlled around half open and fully open, what means to feed the solenoid with around 5-6V up to 12V.

I'm thinking to do it as simple as it could be, to use a potentiometer to set up the required voltage on the G of the mosfet.

Is this kind of setup safe enough?
I mean are there some hidden thinks to take in count if I use a potentiometer to increase and reduce the amount of voltage to the G of the mosfet and like this drive the solenoid?

Thanks for any advice and ideas.

My best regards.

[Circlotron]

First off, will your solenoid actually behave the way you expect as regards applied voltage vs percentage open? Solenoids generally need a bit of encouragement to get them opening, but once they get toward the full open position they really go for it because the closing up of the magnetic gap makes the attraction stronger. The force vs distance is very nonlinear.

[bob91343]

You should have a diode to protect the FET from induction due to the valve's configuration.  Put a diode of 3A rating, any voltage is fine, in parallel with the valve coil in such a direction as to not conduct when the valve is energized.

[pqass]

According to https://tameson.com/proportional-solenoid-control-valve.html PWM is the way to go.
See bottom of https://www.electronics-tutorials.ws/waveforms/555-circuits-part-1.html for typical 555 PWM circuit.  Replace the motor with the solenoid.

You may also want to try a variable current sink https://www.codrey.com/electronic-circuits/simple-electronic-dc-load/.
Add the solenoid and +12V voltage source in series with the MOSFET drain.  Add a reversed diode across the solenoid terminals (like the motor example).

Both circuits above assume DC operated solenoid valve.

[Whales]

Check "Safe Operating Area (SOA)" in your mosfet's datasheet, make sure it can handle the expected conditions.  Expect to be disappointed by many mosfets

[NiHaoMike]

Just PWM the solenoid, simpler than making a variable current sink.

[Chriss]

Thanks to everyone for the replay.
I will solve my task with pwm.

My valve is actually a proportional solenoid valve as pqass mentioned.

Thank's to all of you.
My best regards.

[Circlotron]

My valve is actually a proportional solenoid valve as pqass mentioned.

I missed that. I didn't know there was such a thing as a proportional solenoid. I wish our washing machine had them. Depending on the size of the load, it turns the water on and off as much as 15-20 times during the fill cycle, sometimes being on for as short as a fraction of a second. One of these solenoids could make for a gentle shutoff and saved me having to fit two water hammer surge arrestors.

[Siwastaja]

Yes, definitely PWM since the solenoid itself has a lot of inductance meaning a square wave control signal becomes a triangle wave current, and given high enough frequency (depending on the solenoid, even just a few kHz could be enough), this triangle current approximates a constant value with low enough ripple not to cause mechanical vibration.

[Zero999]

Definitely control the current, rather than the voltage. A sense resistor can be used to monitor the current flowing through it.

[ajb]

I'm thinking to do it as simple as it could be, to use a potentiometer to set up the required voltage on the G of the mosfet.

I most cases it's definitely better to PWM the MOSFET than to use a linear solution like this, but if for some reason you DID want a linear driver, you wouldn't do it like this.  The relationship between Vgs and Rds is dependent on operating conditions, so you can't really get a consistent mapping across temperature range or across multiple devices.  Instead you'd want an active current regulator, as shown in this link pqass shared earlier: https://www.codrey.com/electronic-circuits/simple-electronic-dc-load/  The op amp drives the gate of the pass transistor as needed to regulate the voltage across the sense resistor, which means you get a constant current through the load.  (There are still temperature and other factors that will affect stability and precision, but these are much better controlled through the op amp's feedback than they would be if you just applied a fixed voltage to the FET gate.  You would still need to review the safe operating area chart for your transistor to ensure it will be suitable for your application, many FETs are optimized for switching applications these days and not all are suitable for linear applications like this.

It's good to know that you're using a proportional solenoid valve.  Many solenoid-actuated valves can't really be used proportionally, even with an external current driver, because the solenoid does not directly actuate the main valve.  Instead it controls the flow of a small amount of the working fluid that is used to actuate the main valve spool.  This allows the solenoid to be much smaller and lower power than a direct-acting solenoid valve would require.  Most pneumatic/hydraulic control valves operate this way, as do a lot of diaphragm valves like are used for irrigation systems.

[Chriss]

My driver should mitigate an ecu output which is gonna bad but the machinery should still work until the ecu is not switched to a new one. To get the new ecu we have to wait several weeks, so we have no option, just to solve the problem in some way.

Then I come to the ide, ok, I could make a dimmer with a mosfet and pot. But after I go back to the machine and checked again the valve, I realized it is a proportional valve 12 VDC. And when I checked back here I saw also a writing about this type of valve, and links to the valve too.

I realized quickly does I have nothing to do with my dimmer stuff here.
And you suggested also to go to pwm controller.

I didn't design lot of electronics but I have some clue about and finished many project successfully, but I like to be careful, and  like to hear others opinions as well before I burn down something. 

I will solve my task with a pwm driver which I constructed many years ago, and I know it is good design cos I tested it many times in real situations, and also made measurements during the usage.

Thank you folks for such of nice infos, I really appreciate your time spent on my question.
My best regards.