1
\$\begingroup\$

I'm trying to put together a boost converter to take a 11-14.4v input and boost it to ~17v with a max current of 2A.

Most of my designs (made either by calculations done by me or a calculator tool from the web, all of which have actually given me different values) work fine at low loads (~100-150 mA), but the voltage drops off quite quickly as I increase the load and I just get a small ripple followed by a flat voltage somewhere below 12v after about 300mA upwards.

I've made the circuit physically once, but now I am using LTspice to simulate it.

I'll be using a 100KHz switching frequency (LM2585T-ADJ) and the online calculators I've used have said to use: Min 64uH inductor with 38% duty cycle (http://www.coilcraft.com/apps/selector/selector_2.cfm) Min 6.28uH inductor with 15-35% duty cycle(https://learn.adafruit.com/diy-boost-calc/the-calculator)

EDIT: Good point about the schematics, don't know why I didn't include them straight away. LTspice schematic Eagle schematic

\$\endgroup\$
  • \$\begingroup\$ schematic & LTspice-schematic would be helpful. does it fail under heavier load both in sim & on the bench? or only one - if so, which one? \$\endgroup\$ – Techydude Jun 25 '15 at 3:07
  • \$\begingroup\$ There's no magic to a 2A boost converter, so if yours isn't working there's something wrong with your design or implementation. Remember your switch current limit has to be higher than the 2A since the switch is switching the input current (which is around Vout/(Vin*efficiency). As @Techydude said a schematic would be very helpful. \$\endgroup\$ – John D Jun 26 '15 at 3:02
  • \$\begingroup\$ Yeah, fails under heavier load on both. Real circuit worked exactly as expected under very light load, the sim doesn't level out at the desire voltage, but I assume that's because there is no feedback circuitry to control the output. Good point about the switch current limit, but in my simulation there shouldn't be any current limit on the switch, should there? I am pretty new to LTspice though so I may be wrong. I assume I have misunderstood something about boost converters because they do seem fairly simple in operation. \$\endgroup\$ – Eddy Jun 26 '15 at 4:53
0
\$\begingroup\$

If the Spice model of the IRFP4668 is at all accurate, it will not be turned on well with only 5V gate drive. You should increase the gate drive to 10V to support higher current. Expect the power modulator to have a resonant frequency of ~ 1200 Hz, with a Q of ~ 4. So, it will take a lot of simulated time to settle, probably at least a second and the output will be ringy.

Peak switch current with a 34W output and 11V input and 64uH inductor will be ~3.4A. The LM2585 switch can limit as low as 3A, so don't expect to get 2A output. It might be able to put out 1.75A with a 3A switch limit. The inductor will need at least a 5A rating to avoid saturation and thermal problems.

Also, the 12V input will need to be able to supply the 3+ amps.

\$\endgroup\$
  • \$\begingroup\$ Ah thank you so much! Just needed to up the gate voltage to 10v and the simulation works exactly as expected, although doesn't take very long to settle like you predicted. Yeah, I've since realised I will need to either lower my current requirements or choose a new regulator. When looking at inductor current ratings, I am looking for the saturation current to be greater than 5A to avoid saturation occurring, correct? \$\endgroup\$ – Eddy Jul 2 '15 at 0:36
  • \$\begingroup\$ I'm glad it worked out @Eddy Yes, the inductor needs to be rated >5A for saturation, and also to keep power dissipation down. \$\endgroup\$ – gsills Jul 3 '15 at 2:15

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.