Control 10 amps in solenoid with Arduino and 2n3055

Thread Starter

ealbers

Joined May 7, 2017
3
Hello,
My son is creating a science fair project which needs to smoothly vary current through a solenoid coil from 0 to 10 amps, controlled by a Arduino.
Note we cant use pwm, as this would cause the coil field to pulse, so it needs to be able to smoothly vary the current.

How would one drive a 2n3055 from a digital to analog converter? Is there a better way? Is it possible to drive a MOSFET in its linear region to smoothly vary 10 amps?
Thanks!
 

crutschow

Joined Mar 14, 2008
34,281
You can use PMW as that's an easy and efficient way to control the current.
You just need the PWM frequency to be high enough so that the coil inductance smooths the current (typically a few kHz is sufficient).

If you vary the current linearly, the transistor will dissipate a lot of heat.
What is the maximum coil voltage to give 10 amps?
 

Thread Starter

ealbers

Joined May 7, 2017
3
You can use PMW as that's an easy and efficient way to control the current.
You just need the PWM frequency to be high enough so that the coil inductance smooths the current (typically a few kHz is sufficient).

If you vary the current linearly, the transistor will dissipate a lot of heat.
What is the maximum coil voltage to give 10 amps?
We really want to vary the current without pwm, the hall sensors he's using to map the magnetic field would 'see' the pulses even with the filtering the iron core would help with....

We are looking at a 12volt system (old PC power supply), running through 2 2ohm resistors in parallel plus the winding resistance of about 0.2ohms should give 12/1.2ohms or 10 amps through the system....want to smoothly vary the power so my son can map the magnetic field at various power levels for his science fair project....
I know the 2n3055 would get hot, would it work? Could I drive it with a 2n2222 from a d-a converters 0-5v output?
Don't want to fry anything :), its a science fair project, if it dissapates a bit of heat I'm not too worried, I'll heat sink the thing.

Can a MOSFET be used in its linear area??? how about a normal FET?
 

Alec_t

Joined Sep 17, 2013
14,280
How about this?
CurrentController.PNG
Note that the resistor and MOSFET would be dissipating up to 100W and 30W, respectively, as heat, so would need to be mounted on very large fan-cooled heatsinks. Since this is for a science fair you would also need to screen the heatsinks to prevent anyone coming in contact with them and getting burnt.
Now you see why PWM control was suggested. Any measurement signals could be filtered to smooth out ripple arising from the pulsing.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,281
The transistor will be dissipating a peak power of over 30W and the resistors will be dissipating a maximum of 100W. :eek:
The cost of those power resistors alone will be considerable.

I think you need to reconsider PWM.
The frequency can be high enough that the current ripple is small and the frequency is likely well above what the hall sensors can see.

What is the expected inductance of the inductor?
What is the hall sensor frequency response?

If you still want an analog approach below is a circuit that should work.
It uses a power N-MOSFET to control the current with two BJTs to cancel the input offset voltage.
The circuit acts as a low offset, source follower.
It gives 0-10A through the inductor for a 0-5V input (for a 0.5Ω source resistor).
Note the two high powered resistors needed.
The N-MOSFET can be any type of sufficient voltage and current, in a TO-220 type case.

upload_2017-5-7_12-54-38.png
 

Thread Starter

ealbers

Joined May 7, 2017
3
How about this?
View attachment 126214
Note that the resistor and MOSFET would be dissipating up to 100W and 30W, respectively, as heat, so would need to be mounted on very large fan-cooled heatsinks. Since this is for a science fair you would also need to screen the heatsinks to prevent anyone coming in contact with them and getting burnt.
Now you see why PWM control was suggested. Any measurement signals could be filtered to smooth out ripple arising from the pulsing.
Very cool, the resistors I ordered are 100 watt 2ohm's, so paralleled its 200watts at 1ohm, so I think that will do well, wow 30watts out of a little mosfet, that will be hot!

Would a 2n3055 be a better choice??
 
Top