I'm building a portable charger for my phone using a large 2.7v supercapacitor, which uses a very simple boost converter to create 5v. I am not concerned about input current, however, when a load (my phone) is connected, the voltage falls to 4.53v, regardless of any changes I make to the duty cycle, frequency, or inductance. My best guess is that my phone only begins charging when the voltage increases over this point, but this causes the voltage to fall again and oscillate. My question is, how can I increase the amount of current that the boost converter can source, so that the output voltage will be nearer to 5v?

The boost converter circuit disagram

The actual boost converter

  • 2
    \$\begingroup\$ 1) You can FORGET to get this working properly on a solderless breadboard because the contact resistance will be too high for what you want. On a breadboard limit yourself to currents of less than 100 mA to avoid issues. A phone charging at 100 mA will take very long to charge. Normally phones at 5 to 10 times 100 mA ( 500 mA to 1 A some even more). 2) That inductor is completely unsuitable for the currents needed to charge a phone. That inductor is only for very small currents like 1 mA. It saturates at high currents making it unusable. \$\endgroup\$ Commented Oct 17, 2018 at 14:18
  • \$\begingroup\$ Much easier and much more workable solution is to buy a ready made module like: ebay.com/itm/… \$\endgroup\$ Commented Oct 17, 2018 at 14:24
  • \$\begingroup\$ Could you post the part numbers of the parts used? \$\endgroup\$ Commented Oct 17, 2018 at 14:30
  • \$\begingroup\$ Also a BS170 isn't going to work for this at all. A BS170 has a maximum drain current of 0.5 A. Your phone charges at 5 V and will probably need at least 500 mA, if it gets less current it might not charge at all. 5 V, 500 mA is 2.5 W. At an input voltage of 2.7 V that means more than 900 mA has to flow through the BS170, that will break it! At lower input voltages this gets even worse. \$\endgroup\$ Commented Oct 17, 2018 at 14:31

3 Answers 3


I find this question close to me, as I some years ago I was doing the same kind of mistakes on a very similar project (basically getting a boost convertor on a breadboard, without the knowledge about the MANY pitfalls, as others have already mentioned). I asked in a forum and I got a lot of surprised looks from the elders, just like you :)

There are many things in this project that can be fixed with more reading and dedication, however I feel that a major hurdle that may stop you is the need to transfer it to a real PCB (best case) or at least very tightly packed prototype board (with a lots of caveats there). If you want to charge a phone, that is.

If my gut feeling is right, and you don't want to leave the convenience of the breadboard, may I suggest doing something slightly different, that will get you the most of the intuition about boost converters¹, but still be feasible on a breadboard, and with the added benefit of avoiding the risk of killing a potentially expensive equipment (the smartphone)?

How about boosting the supercap to power a bright white LED (the ones that require 3+ volts)?

You still need the boost converter (and you can demonstrate that the supercap alone doesn't light the LED even the tiniest bit), but the currents required will be within the capability of jellybeans like the BS170, and if your switching frequency is low enough², within the capabilities of breadboard.

¹ Sadly, to really get the hang of how boost converters work, it would be best if you can poke around with an oscilloscope. If you can get your hands on one, at least for a while - go for it!

² In general, stay below 100 kHz; less is not an issue, you just need larger caps and inductors. More, and you get into the area that makes breadboards inappropriate, and the MCU wouldn't cope with driving the MOSFET efficiently too.


None of the components you have selected are appropriate for that circuit.

Loosely speaking you will want all components to be rated for 2-3 A.

Your switching NMOS has an \$R_{DSON}\$ of 5 Ohms with a 10 Vgs drive signal.

Before you attempt to connect your phone to your new circuit, I recommend testing with a 10 Ohm resistor as a 500 mA load. As an unregulated boost converter can generate high voltages on the output.

A 5.6V zener on the output for over-voltage protection would also be wise.


The inductor can only store a certain amount of energy in the field established by the current when the FET is on, which is then discharged to the output cap and load during the off cycle, so increasing the switching frequency will increase the power transferred. The size of the core of the inductor determines the amount of energy storeable in each cycle, a small core will saturate at a very low current. There's an Isat rating available for most inductors.

With USB charging you also face an issue where the load is reacting to the supply voltage - in the absence of any enumeration (identification of the connected device over the data lines), the charging circuit resorts to stepping the current up until the input voltage falls to indicate the limit of the charger's output.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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