0
\$\begingroup\$

I need to make a voltage controlled current source for nichrome wire as part of an experiment. It seems like I need a circuit which can translate a microcontroller's 0...5V range voltage to a 0...8A current. I mean a voltage controlled current source. I can use a DC power supply which can supply this current.

Is there a topology or any suggestion circuit for example composed of an opamp and a power transistor? Or any specific IC would work as well.

I would be glad if one can share a circuit diagram for this purpose. (An opmap followed by a power transistor was in my mind but Im not expert in the topic)

MAJOR EDIT:

I think PWM control is better for power consumption. So the uC's PWM pin is the input signal. Here is what I have in mind so far. But never tried such high current circuit before:

enter image description here

And here is simulation results for current, instantaneous power and average power: enter image description here

Some more info:

PWM will be 490Hz frequency.

I'm planning to use this power MOSFET.

Flyback diode is a 1N4007.

I would be glad to hear about any suggestions such as the MOSFET ratings, heatsink required or not or any circuitwise mistakes here.

EDIT 2

Here is the new circuit with a MOSFET with a lower Rds(ON): enter image description here

I had to modify Rg because it wasnt saturating the new MOSFET sharp enough. I reduced it from 10k to 220 Ohm. And R_col was too low for a quarter watt resistor. I can only make comments on it by looking at the simulation below:

enter image description here

\$\endgroup\$

closed as off-topic by Bimpelrekkie, Trevor_G, Dmitry Grigoryev, Autistic, uint128_t Apr 23 '17 at 22:19

  • This question does not appear to be about electronics design within the scope defined in the help center.
If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
    \$\begingroup\$ 8 A at what voltage? \$\endgroup\$ – Olin Lathrop Apr 20 '17 at 19:40
  • \$\begingroup\$ There are lots of ways to generate a voltage controlled current source. Google is your friend. Try "voltage controlled current source" and click images.. then take your pick. \$\endgroup\$ – Trevor_G Apr 20 '17 at 19:41
  • 3
    \$\begingroup\$ I'm voting to close this question as off-topic because this site is not a design service. You are welcome to propose a design here and then we can discuss that design. \$\endgroup\$ – Bimpelrekkie Apr 20 '17 at 19:41
  • \$\begingroup\$ @FakeMoustache I need a simple solution for experimental use not a professional design but you are right. Please give me sometime. I will edit this question with a circuit I try to simulate just trying to find proper components. I will add the diagram maybe you might have suggestions on it. \$\endgroup\$ – HelpMee Apr 20 '17 at 20:52
  • 1
    \$\begingroup\$ I just made an edit with a circuit I have in my mind. \$\endgroup\$ – HelpMee Apr 20 '17 at 21:42
1
\$\begingroup\$

You basically have two or three choices for this.

You need to use pulse width modulation, PWM, controller of some sort since the currents are too large for an efficient linear solution that does not need a heat-sink the size of a shoe-box.

The issue with PWM is you are driving a purely resistive load. That means you are switching large amps at large voltages with sharp edges. As we all know that translates into a rather effective EMI transmitter.

As such you have two options.

1. High Frequency PWM with an added inductor.

schematic

simulate this circuit – Schematic created using CircuitLab

Using the above technique you can maintain the current, all be it with a ripple, in the coil and wire at a fairly steady rate and the PWM action will be switching voltage and a sustaining current level.

However, since the load wire resistance is very low, the coil needs to be large so it's DC resistance is much less than the wire so you do not lose power by heating the inductor.

2. Low Frequency PWM

The alternative is to use low frequency PWM.

schematic

simulate this circuit

Low frequency means drive the wire like it would be driven from the AC line, i.e 100, or 200Hz. In this scenario you are swill switching the full current but now you are doing it at a frequency that is much less problematic from an EMI point of view.

However, it may still be prudent to add a small inductor coil to reduce the current rise time a just a little.

3. A fully integrated current driver.

You could implement a fully integrated current control system with feedback using an LED driver chip like the LT3086 from Linear Technology. That would be a high frequency PWM system but again you are going to need a large inductor, magnetically and physically.

This would be the ultimate design for precise control of current, but may well be overkill for your application.

Implementation

In all cases it is very important to design your circuit so the MOSFET turns on and off as quickly as it can, especially with a high frequency PWM system. During transition the resistance of the MOSFET changes rapidly, however there is a considerable amount of power lost as heat during that transition. As such you can find the MOSFETS will overheat despite the temperature you think they should be running at. That may mean adding a push-pull pre-driver before the MOSFET to deliver sufficient current to fill and dump the gate capacitances quickly.

Obviously your power supply needs to be capable to supply the full current on your wire, but you also need a significant charge reservoir capacitor close to the driver circuit. Although capable of providing the max current, the power supply itself may not be capable of adapting to such high transitions in current immediately.

Recommendation

Of the three approaches I mention, I personally would opt for the low frequency PWM method with fast MOSFETS. If need be, I would also add some small inductance into the line to limit the rise time and noise spikes a little.

ADDED THOUGHT.

If you are driving this thing from an AC transformer you may want to consider a normal "dimmer" type triac circuit.

\$\endgroup\$
  • \$\begingroup\$ Hi thank you I opened a new question about this, I would be glad to have your suggestions. I will also use low freq. 489Hz. Should the 220u parallel cap be close to the nichrome wire? I dont need inductor right? Please see the modified circuit here: electronics.stackexchange.com/questions/301144/… \$\endgroup\$ – HelpMee Apr 22 '17 at 19:51
  • \$\begingroup\$ Hi again, I want to use driver as you suggested for the MOSFET at 500Hz freq. I want to use a TC427. You mentioned to add a reservoir capacitor close to the driver circuit. You mean adding it to the power rail of the gate driver. And would 470u be enough? And do I really need an inductor for 500Hz in series with the wire and how big should it be. I want to try your second commendation for the low frequencies. \$\endgroup\$ – HelpMee Jun 12 '17 at 10:05
2
\$\begingroup\$

Such high currents are best handled efficiently. Getting rid of the heat from inefficiency will take more trouble, expense, and space than designing a better circuit in the first place.

My first reaction is to have the microcontroller control a low side switch with PWM. The thermal time constant of even a thin wire will be long compared to what modest PWM hardware in a micro can do.

The nichrome wire won't get hurt by pulses, but you can put a inductor in series with it if you really want to filter the current for some reason. Since the wire looks like a resistor electrically, all you need is a power supply, the wire, and a switch in series.

Added in response to schematic

I wouldn't try to invent your own FET gate driver. You want the edges to be nice and crisp. 8 A is a bit too much to reasonably find a FET that can be driven directly from a digital signal. With 8 A thru 1.5 Ω, you obviously have a 12 V power supply available. That would be ideal for powering a FET driver. The switching edges should be nice and sharp, and the result quite efficient. No heat sink should be needed.

For example, let's say you get a FET with maximum Rdson of 15 mΩ. That would only dissipate just about 1 W. Something like a DPAK with the right copper pads can dissipate that directly. Since nearly 100 W would be going into the wire, the overall setup is quite efficient. And, at these low voltages, 15 mΩ is no stretch at all.

Note that in this case, the PWM duty cycle controls the average power, not the average voltage, as it would with a inductor and diode in there. However, that doesn't make it harder to control. If this is part of a larger control loop, linearly controlling the power is probably better than controlling the square of the power, which is what voltage control would be.

Supply voltage

I just noticed due to a comment from Trevor that you plan to use a 24 V supply. Don't do that. Use a supply that is the voltage the wire needs at full on. A 24 V supply will cause twice the dissipation in the same FET as a 12 V supply when using the purely resistance-switching scheme I'm talking about. Fortunately, 12 V is common DC power supply value. Then you can also run the FET driver directly from the same supply. 24 V would be too high for that.

\$\endgroup\$
  • \$\begingroup\$ Agreed. Issue with ni-chrome drivers though is the load resistance is so low... that pretty much makes the add of an inductor almost a must... Playing around with it, the inductor gets large pretty quickly and I am having a new respect for the depth of the issue. \$\endgroup\$ – Trevor_G Apr 20 '17 at 21:31
  • \$\begingroup\$ Please see my Major Edit. I used a MOSFET and tried to make a circuit and gave more details. I would be very glad to have your suggestions. \$\endgroup\$ – HelpMee Apr 20 '17 at 21:42
  • \$\begingroup\$ @Trevor: I just noticed that the OP said in a comment that the wire resistance is 1.5 Ohms, so 12 V at 8 A. A 20 mOhm FET, for example, as low side switch would work fine. There is no need to smooth the current unless you're worried about capacitively coupled noise maybe. Just turning the FET on and off with PWM would give good control over the wire heating power. \$\endgroup\$ – Olin Lathrop Apr 20 '17 at 21:44
  • \$\begingroup\$ @OlinLathrop true though his 24V PSU needs to be able to deliver those 14.5A current pulses... And yes, that's some noisy signal. \$\endgroup\$ – Trevor_G Apr 20 '17 at 21:50
  • \$\begingroup\$ @Trevor: He should use a 12 V supply. For the noise, you can use a small inductor, diode, and capacitor across the wire. These aren't big enough to sustain the current for long during the off period, but will reduce radiated noise significantly. \$\endgroup\$ – Olin Lathrop Apr 20 '17 at 21:59

Not the answer you're looking for? Browse other questions tagged or ask your own question.