4

I've designed the boost converter below:

enter image description here

Aside from the irrelevant voltmeter parts which are not connected, I'm asking if the way I'm driving the MOSFET is correct.

Working is simple, when TL494 let's current from its transistors, voltage will occur on the TC4420. My concern is that I've put a 20k large resistor which would limit the base current of NPN transistors of TL494 if they even exist. I don't know how exactly TL494 works but in the datasheet it mentions about NPN transistor, collectors and emitters so I assumed it has ordinary NPN transistors inside.

To remedy my concerns, I can move the TC4420 input to collectors and add a 20k resistance between collector and VCC. Since the output is inverted in that case I would replace TC4420 with TC4429 which is inverting version. Two inversions cancel out. I can connect emitter directly to ground and thus let TC494 decide the base current of its transistors.

On the other side, I am open to any suggestions about the converter in general.

Thanks.

CaveScientist
  • 375
  • 1
  • 10
  • 1
    Something like a uc3842 would be much easier methinks. Controller and mosfet driver in one chip and dirt cheap. What does the drive waveform look like to the tc4420? You can go lower on the value as the tl494 can switch a 100mA or so - its been some time since i even looked at that part. – Kartman Nov 14 '22 at 02:35
  • What's the Vcc voltage in this schematic? – Nick Alexeev Nov 14 '22 at 03:43
  • VCC = 7V-18V I haven't tested yet so I dont know the waveform of TC4420. True uc3842 would be more modern but I still want to know about the npn transistors in these ICs – CaveScientist Nov 14 '22 at 12:02
  • @CaveScientist Not necessarily more modern, the UC3842 has been around for almost 30 years. Nothing wrong with using it, (don't forget to add external slope compensation if needed) but there are much newer controllers that might be easier to use, harder to get and more expensive :) – John D Nov 14 '22 at 19:23
  • @JohnD I recall using a 3842 in a project about 38 years ago :-) . – Russell McMahon Nov 14 '22 at 22:50
  • @RussellMcMahon You're right, now that I think about it, I remember back about then the arguments about the "new current mode" topology and the 3842 being the main current mode controller. – John D Nov 14 '22 at 22:56
  • 1
    @JohnD I appear to have overstated that slightly :-) - in [this fascinating post](https://www.righto.com/2021/08/) Ken Schirriff says it was introduced in 1984 (say 38 years ago) and it would have not been new when I used it. Ken reverse engineers the chip by removing the package and inspecting the die. Then discusses the design indetail. Very worthwhile. – Russell McMahon Nov 15 '22 at 06:23

0 Answers0