I am working on making my own "mod" for an e-cigaret (A mod is the device that controls when and how much the coil is heated up to)

I have read a lot about it during the last couple of days, so i have a few different ways to drive it, each with their own ups and downs. Here is what i have learned so far.

First is the mechanical mod, basically a switch to allow and deny current to flow through the coil: Mechanical mod Ups:

  • Simple to build
  • No power wasted


  • No safety features (No current limit, other than the coil)
  • No power regulation

Next is the PWM approach: PWM driven mod Ups:

  • No wasted power
  • Able to regulate current, but not fully


  • Coil will get full current, no matter what the mod is set to, just not all the time while firing

Last is the DAC approach: Analog driven mod Ups:

  • Able to fully regulate current


  • Potentially lots of wasted power
  • Lots of heat generated in the wrong place (MOSFET)

I want to note the schematics are very simplified, of course additional components are needed in order to have a functional design.

I am most inclined to use the PWM apprach, but the fact that i cant measure current directly with a shunt puts me a little off.

Anyhow, here is my questions:

  • Is there a way to smooth the current running through the MOSFET so i can measure current with a shunt?
  • I have two 18650 batteries in series in order to achieve high enough voltage to drive my coil at the wanted power level (110W in this case), but one cell is limited at 25A in its datasheet, that means i can pull 25A at 7,4V nominal voltage, right?
  • Is there another way to drive the coil that i dont know about, where there is smooth current to the coil and minimal power loss in the switch component?
  • \$\begingroup\$ DAC approach will not work as shown. MOSFET cannot be reliably set to a known resistance or current open loop. You could monitor the coil-drain voltagfe or the voltage across the coil;d or measure current in load+ FET OR create a constant current driver (or even a constant voltage driver). BUT, PWM is fine. \$\endgroup\$ – Russell McMahon Aug 13 '19 at 10:38
  • \$\begingroup\$ @Russel McMahon Yes, the DAC circuit needs a feedback of some sort, thats for sure, else it wouldnt work at all. But i am planning on using PWM after the feedback i got from the people in here, so that is great :) \$\endgroup\$ – Fiskelord Aug 13 '19 at 10:53
  • \$\begingroup\$ @Fiskelord -- According to my calculations and simulations, you only need 4.2 volts (and 26 Amps) into 160 mOhm to achieve 110 Watts. With your stated 7.4 volts into 160 mOhm, you end up with about 340 Watts (and 46 amps, all parasitics ignored). Are you open to coils with different resistances? \$\endgroup\$ – MicroservicesOnDDD Aug 18 '19 at 2:13
  • \$\begingroup\$ @MicroservicesOnDDD that's true, but im not trying to build a mechanical mod, I'm trying to build a regulated one. If I only went with 1S, I would only be able to run the coil at 110W for a few puffs. So instead, I want to make a regulated mod, 2S, so I can run with pretty much whatever resistance and power limit I want (within reason, of course!) \$\endgroup\$ – Fiskelord Aug 18 '19 at 9:46

Heating coils integrate the energy dumped into them to make an overall average dissipation. The PWM happens fast enough that the variation of thermal output is insignificant.

The power dumped into the coil is proportional to the square of the PWM duty cycle. So at 50% PWM, you get about 25% of the max power. No need really to sense this directly, you can compute it.

That said, the coil will vary its resistance depending on temperature (it will increase some). You could characterize this by measuring the average battery current over the PWM range, and map this curve into your PWM setting.

Here’s some info about driving a nichrome wire heater with PWM. http://www.brysonics.com/heating-a-nichrome-wire-with-math/

If you want to measure current, it’s possible to do so by measuring the IR drop across the FET when the switch is on. As long as you know the Rds(on) of the FET you can get a good reading. Use a low-side diff sense amp circuit to boost the voltage to a useable value.

| improve this answer | |
  • \$\begingroup\$ Funny thing about kanthal wire is the temperature coefficient is small enough that it doesn't matter in this context, so that's great, just a shame it makes it unsuitable for temperature control. But yeah, it just dawned upon me what you are saying, makes good sense! \$\endgroup\$ – Fiskelord Aug 13 '19 at 7:42

Simple as that. You do lowpass filtering with a RC filter and then buffer. Only the PWM is the right approach if you want to control the current with minimum loses. Still, regulating current won't give regulted temperature as you may need.


simulate this circuit – Schematic created using CircuitLab

| improve this answer | |
  • \$\begingroup\$ How do i calculate rc filter? If I dimension it such the pwm frequency will pass, will the pwm signal be smoothed out? \$\endgroup\$ – Fiskelord Aug 13 '19 at 7:39
  • \$\begingroup\$ For this to be useful, the sense amp needs to have gain so that Rsh can be a reasonable low value. But, see my amended answer above. \$\endgroup\$ – hacktastical Aug 13 '19 at 7:57

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.