# Design concepts for a high performance current source.

#### Just Another Sparky

Joined Dec 8, 2019
234
I'm working on a project which will require the construction of a current source capable of producing rapid current changes in a DC inductor. I'm working with a proportional valve having a 13 ohm, 0-1000mA DC coil. The valve will respond only as quickly as I can drive the current up or down. E.G. applying 13 volts across it results in a quarter-second delay while the armature magnetizes, whereas breaking the circuit drives the valve shut hard with no hesitation.

Any pointers as to what sort of design concepts might work best for this? Perhaps some sort of freewheeling half bridge having a bus voltage substantially higher than 13 volts? Would a full H-bridge work any better than a half bridge for driving the current down rapidly? Or would I be better off with some manner of circuit to regulate the forward voltage drop across a suitably large series transistor to act as a variable 'brake' resistor?

Joined Feb 20, 2016
4,177
You may try a switched constant current source but the supply voltage will need to be greater than the 13V you have.
Also, try turning it on with 24V then dropping down to the 13V for holding. That will turn it on faster.
But anyway, a controlled current is the way to go. Just have a limit to the max value.

#### Sensacell

Joined Jun 19, 2012
3,034
A linear current source is simple- but will be a big power waster requiring a large heatsink.
Use the highest source voltage you can, this is what determines the rate of current rise - heat and power problems must be considered.

#### DickCappels

Joined Aug 21, 2008
9,086
That brings me to ask - how long and how often is the current on?

Do you need more than on/off and can you rely on "flyback" to discharge the inductor or do you need to actively drive it until the current is effectively off?

#### Just Another Sparky

Joined Dec 8, 2019
234
For all practical intents and purposes the coil will be a continuous load. It is rated for 13 watts. I'd like to be able to drive the current both up and down since the application is a three-port proportional valve - meaning it has a 'fill' range (e.g. 100-400mA) and a 'discharge' range (e.g. 600-900mA) with a center 'lapped' position in which all ports are blocked. Being able to traverse the lapped position from 'fill' to 'discharge' or vise-versa in a mighty hurry will be very valuable for maintaining stable and accurate PID control.

I did a little more thinking. What are your thoughts on something like this? I'm going to have a microcontroller at my disposal so generating PWM and designing a proportional forward-VD regulating circuit to drive the analog "brake" transistor in relation to a second PWM signal shouldn't be the end of the world.

It's been ages since I've dealt with transistors - not since tech school as far as I can remember. Would basic BJTs be good choices for this application or could I do better with a different technology like a FET without confusing the issue too much?

I'll determine the inductance of the coil and calculate my DC bus voltage to achieve a maximum 0-100% response time of roughly 30ms and do the same for the power rating & forward voltage rating of my "brake" transistor.

A variac connected to the coil and adjusted to pass a current of 1000mA should be accurate enough for the purposes of my calculations, no?

Last edited:

#### MrAl

Joined Jun 17, 2014
9,344
I'm working on a project which will require the construction of a current source capable of producing rapid current changes in a DC inductor. I'm working with a proportional valve having a 13 ohm, 0-1000mA DC coil. The valve will respond only as quickly as I can drive the current up or down. E.G. applying 13 volts across it results in a quarter-second delay while the armature magnetizes, whereas breaking the circuit drives the valve shut hard with no hesitation.

Any pointers as to what sort of design concepts might work best for this? Perhaps some sort of freewheeling half bridge having a bus voltage substantially higher than 13 volts? Would a full H-bridge work any better than a half bridge for driving the current down rapidly? Or would I be better off with some manner of circuit to regulate the forward voltage drop across a suitably large series transistor to act as a variable 'brake' resistor?
Hello,

One of the definitions used for an inductor is very very common:
v=L*di/dt
where
v is the voltage (sometimes assumed to be constant for simplicity),
L is the inductance,
di is the change of current,
dt is the time interval over which the voltage is applied.

We can solve for the change in current di very simply:
di=v*dt/L
and so the change in current is brought about by the voltage, the time that voltage is applied, and the inductance.
Now since we cant change the value of L in most cases that is a constant, and dt is often predefined by the application too, so all we can do is change the voltage. If we want a larger increase (or decrease) in the change of current 'di' we need to increase (or decrease) the voltage 'v'.
Now since 'v' enters into the right side as a linear factor, if we double 'v' we double 'di'. Very simple.
So if we had di=2 amps when the voltage was 5 volts and we need di=4 amps instead, then we need to increase the voltage to 10 volts (double the current required double the voltage).

There is a catch though because there is a limit to how high we can go with the voltage. Often the construction of the inductor L has some limit on the voltage it can take before the wire insulation starts to break down and that ends up either shorting out or partially shorting out some of the windings. So you have to increase the voltage but be aware that there is a limit to how high you can go. If you find that you have to go higher than you need a different inductance which would mean a different motor.

To find out more about the circuitry used to do this, look into the circuits used to drive stepper motors. The control chips used for that have internal measurement circuitry that measures the current and provides feedback that allows a higher drive voltage if the current does not reach the required level.
That's the basic idea really. Measure the current and if it is not high enough increase the voltage and the input DC voltage to that circuit would be limited to the max for the construction of the motor or other inductance.
The force of a stepper motor is related to the current, so if the current is not high enough we dont get enough force, and if we dont get enough force the stepper motor would either not turn or take too long to move, so a control circuit is used to get it moving faster and more accurately.

#### LowQCab

Joined Nov 6, 2012
2,337
• Just Another Sparky and dendad

#### Orson_Cart

Joined Jan 1, 2020
85
The best way would be to have an adjustable switch mode supply, e.g. a Sepic, which you can quite quickly set to 13V, 15V, 18V 20V, 30V etc to force a fast ramp of current into the solendoid - and a current limit for each step - - to reduce the current fast you need to apply or allow volts the other way round - you could have a series of fets with different zeners in the drain ckt across the solenoid ( possibly via bridge rectifier to sort out the polarity issue ) and energising the correct fet gives the ramp down of current associated with that zener, an over arching zener of 39V would be the default fastest ramp down ...

#### kaindub

Joined Oct 28, 2019
100
In a relay, the coil inductance is a variable ie its not constant. Think of the relay coil as being a coil of wire that has a magnetic circuit through it (like a transformer) When the relay is not energised, the inductance is small as the airgap is large - the armature has not closed. When fully energised, the airgap is small or non existent, therefore the inductance rises. In steady state operation, the inductor plays no part in determining the coil current .
The relay current upon energisation will be determined by the coil inductance and the coil resistance. The closing force and hence closing time are determined by the coil current(remember the relay is just a simple magnetic circuit)
Therefore to have the relay close twice as fast required twice the voltage to be applied. To apply an infinite amount of current would require an infinite amount of voltage.
In my opinion, the simplest solution is to have a power supply several times the size of the coil voltage capable of supplying the coil current. Insert a dropping resistor between the power supply and the coil and bridge it with a normally closed contact of the relay. When the relay opens the resistor is in circuit and controls the col current to the rated level.

#### LowQCab

Joined Nov 6, 2012
2,337
In a relay, the coil inductance is a variable ie its not constant. Think of the relay coil as being a coil of wire that has a magnetic circuit through it (like a transformer) When the relay is not energised, the inductance is small as the airgap is large - the armature has not closed. When fully energised, the airgap is small or non existent, therefore the inductance rises. In steady state operation, the inductor plays no part in determining the coil current .
The relay current upon energisation will be determined by the coil inductance and the coil resistance. The closing force and hence closing time are determined by the coil current(remember the relay is just a simple magnetic circuit)
Therefore to have the relay close twice as fast required twice the voltage to be applied. To apply an infinite amount of current would require an infinite amount of voltage.
In my opinion, the simplest solution is to have a power supply several times the size of the coil voltage capable of supplying the coil current. Insert a dropping resistor between the power supply and the coil and bridge it with a normally closed contact of the relay. When the relay opens the resistor is in circuit and controls the col current to the rated level.
Did you bother to read any of the other posts in this Thread ?
Did you bother to analyse the suggested Circuit ?

The Thread Starter has a Linear-Actuating-Proportional-Valve operated by a Solenoid.
Having a Relay involved in the control of this Solenoid would result in
completely unpredictable behavior of the Valve in question.

If you would like to ask questions regarding the Linear operation of the above Circuit,