1
\$\begingroup\$

I'm designing a heater controller that has a primary input voltage between 12 and 24V. Everything except the heater runs on 3V3, including microcontroller, LED's, potentiometers and the MOSFET.

The problem is the MOSFET wont be operating at it's full potential, http://www.onsemi.com/pub_link/Collateral/NTD5406N-D.PDF Figure 2 shows that 3V3 limits the current to about 5-6 Amps. I need to switch up to 8 Amps.

As a work around I've considered the following:

  • Using a 5V Regulator with a Shutdown Pin (LP2951) not ideal for PWM.
  • Using a 5V Regulator a level shifter.

I'm planning to manufacture 1,000's so my goal is low parts count and reliable, cost is always a consideration but at around $0.50 (either of the above solutions) isn't too expensive.

Since I don't need much current (<1mA) and I only need a little more voltage (<4V total) on the MOSFET could I safely use a Zener diode or a voltage reference? What would that circuit look like? The couple of attempts I made at simulating these types of circuits didn't work.

Edit:

I created a simulation of what I think is Alex's answer could somebody verify my understanding? Simulation of Alex's Circuit If this would work in the "real world" I like it. Simple and doesn't require another Voltage Regulator. FYI: The Zener pictured is 5V1 and the Input ("L") is 3V3.

Still of the circuit in simulation

\$\endgroup\$
  • \$\begingroup\$ Do you need high-frequency PWM for a heater? I think that rather low frequency will suffice, as the heat will take quite some time to spread evenly. \$\endgroup\$ – DThought Apr 20 '15 at 9:45
  • \$\begingroup\$ My plan is 500-1kHz PWM and use a set Duty Cycle % run the heater. It's a small element so a stiff breeze can make it too cold to use. So a 1Hz PWM might work, but might be a bit slow. \$\endgroup\$ – James Apr 20 '15 at 16:48
  • 1
    \$\begingroup\$ You might want to consider an alternate MOSFET. The NTMFS4897 has similar specs to the one you selected, but can handle 40A Id at 3v Vgs (Figure 2 in the datasheet). \$\endgroup\$ – tcrosley Apr 20 '15 at 18:14
  • \$\begingroup\$ @tcrosley Of course simpler is better, if you would like to add your comment as an answer I will mark it. Thanks! \$\endgroup\$ – James Apr 23 '15 at 23:08
  • \$\begingroup\$ Done. I also added a link to the Digi-Key page where you can buy it. \$\endgroup\$ – tcrosley Apr 23 '15 at 23:30
3
\$\begingroup\$

You might want to consider an alternate MOSFET. The NTMFS4897 has similar specs to the one you selected, but can handle 40A Id at 3v Vgs (Figure 2 in the datasheet).

You can get it from Digi_Key for $1.75.

\$\endgroup\$
3
\$\begingroup\$

What you probably want to do is add another small transistor (NPN or NMOS) in an open-drain configuration as a buffer. So what you would do is add a pull up resistor to the gate on your current transistor and then use the new transistor to pull the gate to ground to turn it off. You will also want to add a pull up to the new transistor as well so that it turns off the power transistor by default. You can add a Zener diode to limit the gate voltage. The control voltage will end up being inverted, but you should be able to turn the transistor all the way on.

\$\endgroup\$
  • \$\begingroup\$ Exactly. VGS needs a minimum of 5V for ops solution, and can take up to 20 V, so as long as Vin isn't past that, it will work. \$\endgroup\$ – Passerby Apr 20 '15 at 16:54
  • 1
    \$\begingroup\$ See question Edit for follow-up. \$\endgroup\$ – James Apr 20 '15 at 18:39
  • 1
    \$\begingroup\$ Yep, that's exactly what I was getting at. \$\endgroup\$ – alex.forencich Apr 20 '15 at 18:44
  • \$\begingroup\$ The only issue I can think of with that circuit might be power supply sequencing. If the 24 volt rail comes up before the microcontroller rail, then the base of that transistor will be held low through the ESD protection diodes in your microcontroller. It might be a good idea to add one more NPN with a pull down on the base in front of what you have there to invert it again and prevent this issue. \$\endgroup\$ – alex.forencich Apr 20 '15 at 18:50
  • \$\begingroup\$ @alex.forencich Good point, perhaps an open collector output pin on the microcontroller could also be used (some PICs have them). The base voltage is never going to get very high so the pin is safe. \$\endgroup\$ – KalleMP Feb 3 '17 at 6:09
2
\$\begingroup\$

If you consider < $0.50 cost acceptable, an elegant solution is to use a MOSFET gate driver such as the Microchip MCP1401/2.

It provides level shifting as well as a buffered output that can drive about half an ampere into the MOSFET gate in order to get it to switch quickly.

enter image description here

Of course you could just build this with discrete transistors- 3 BJTs and a couple resistors, which would have a BOM of a few cents.

\$\endgroup\$
  • \$\begingroup\$ Agree, I have used these chips on multiple heater circuits, they always work great. Nice having the integrated Schmidt trigger as well. \$\endgroup\$ – justing Apr 23 '15 at 23:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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