0
\$\begingroup\$

My question pertains specifically to devices like an EcoSPARK®2 400V, N-Channel Ignition IGBT. I'm investigating the requirements for building an electronic ignition system for a go cart. This would energise the coil and thus generate a spark. You'd be firing at a maximum of 175Hz.

The datasheet uses the following test setup for inductive load testing:-

test set up

The datasheet also says that the IGBT has a logic level gate drive. And from the performance curves and switching characteristics table, it appears the RG = 1KOhm. I'm not sure how to assess it's 21 nC gate capacitance wrt other similar devices.

However this seems to contradict most of the answers on the IGBT tag. Lots of gate drive is recommended there for quick up and down switching. Some even suggest going below 0V to guarantee turn off. A 1KOhm resistor would cap the drive to an absolute limit of 5mA (5V logic) which doesn't seem a lot.

In all honesty, I had made preparations to use a TC4422 drive chip but maybe I don't need to. Is this just a newer form of IGBT and I can directly drive it from a micro controller output pin? Too good to be true?

\$\endgroup\$
  • \$\begingroup\$ According to the spec sheet, yes you can assuming the micro has compatible logic levels. At 175Hz the gate capacitance should not be a great issue assuming you do not use a huge gate resistor. \$\endgroup\$ – Trevor_G Dec 5 '17 at 1:25
  • \$\begingroup\$ RG will limit how fast the IGBT turns off, until RG gets fairly small, and then the sinking currentn capability of the micro pin limits it. You only have to turn the IGBT off fast enough for the ignition system to work. So, how fast should that be? 21nC gate charge (not gate capacitance) will come out mostly at the nonimally 2.8v Vge plateau voltage of 2.8v, at 2.8mA for 1k rg, which takes 7.5uS. They've tested it with 1k and a 1mH load, is that 1mH right for an iggy coil? Is 7.5uS fast enough? You could always test it! \$\endgroup\$ – Neil_UK Dec 5 '17 at 4:02
  • \$\begingroup\$ That is a test circuit not an application circuit. The high RG allows detailed examination of gate drive voltage and current, not optimal function. Use the supplier's app notes, white papers, or evaluation board schematics to learn how to drive it. \$\endgroup\$ – Brian Drummond Dec 5 '17 at 12:07
  • \$\begingroup\$ @BrianDrummond See, I don't understand that at all. With my marketing hat on, I would have thought that the test circuits develop absolutely the best performance that the DUT can generate. Otherwise you'd have no data at all as to how well the device behaves. It would be like having a car brochure say 50 mpg (so you can study the engine) but it actually does 75mpg that would directly encourage people to buy it. Why keep performance secret? Not at all what the sales department would want. \$\endgroup\$ – Paul Uszak Dec 5 '17 at 13:35
  • \$\begingroup\$ Testing is about establishing what can be supported or guaranteed. Don't you remember the days when car mfgrs advertised implausibly high mpg (and if you magnified the small print, it said "at a constant 56mph")? Now they advertise much lower mileage and the small print says "simulated urban cycle". Want to guess how happy the marketing depts were when they were forced to do that? \$\endgroup\$ – Brian Drummond Dec 5 '17 at 14:18

Your Answer

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

Browse other questions tagged or ask your own question.