I have an IR2104 MOSFET half-bridge driver connected exactly as shown in an example circuit in its datasheet: enter image description here

I'm using two IRGPS4067DPBF IGBTs and a 100nF bootstrap capacitor. The whole circuit including the high-side IGBT collector are running on 5VDC. SD and IN are pulled to Vcc. Yet for some reason both the high-side and low-side outputs of the driver are pulled to ground at all occasions (measured with an oscilloscope) and neither of them ever goes high no matter if IN is high or low.

I tried using a different driver yesterday (L6384) with exactly the same results, so I thought that maybe it's broken and bought a different driver today, but without any success as I explained.

What am I doing wrong?

  • \$\begingroup\$ Is \SD (shutdown) pin floating? Or connected to somewhere having low voltage? \$\endgroup\$ – Rohat Kılıç May 13 '17 at 11:17

IN pin should not be pulled high continuously , this driver is a bootstrap driver , which means it needs to recharge the bootstrap capacitor through the low side mosfet when the high side mosfet is off.

buttom line , IN pin should be PWM from 0 to around 95% but never to 100%.

if you want 100% , ( for some applications you do ) there is solutions like adding a charge pump ... etc

check out this app note from Texas instruments Providing Continuous Gate Drive Using a Charge


After checking the datasheet it seems you have another mistake , that is powering the IR2104 from 5v , the IR2104 accepts 3.3v and 5v and 12v logic inputs at IN and SD , but its supply (aka VCC ) should be above between 10v to 20v. the IGBT voltage can be 5v no problem .


  • \$\begingroup\$ I swapped the 5v supply for a 12v PSU and guess what, it works! Such a stupid mistake, I thought that It'd be okay with running at 5v. Thanks! \$\endgroup\$ – DELTA12 May 13 '17 at 11:49

The bootstrap circuit steals energy from the output switching waveform. If you are testing this with no reasonable toggling frequency then there isn't enough energy stolen to feed the bootstrapped driver inside the chip. That means the top IGBT will never become activated. Try inputting a square wave of greater than 10 kHz and looking at the output with an oscilloscope to verify that things are as they should be. You might find it works better if the top IGBT is fed from a somewhat higher supply than 5 volts also.

Static testing is pointless with bootstrapped drivers.

  • \$\begingroup\$ i think the top IGBT higher supply doesnot change anything, \$\endgroup\$ – ElectronS May 13 '17 at 11:34
  • \$\begingroup\$ @ElectronS if the top supply is too low then the bootstrapped supply will be too low to generate enough gate voltage to turn on the top device so yes, that supply level is important. But, as I've said, the input needs to be toggling AND the supply needs to be high enough. Did you not understand this in my answer? \$\endgroup\$ – Andy aka May 13 '17 at 11:39
  • \$\begingroup\$ the way i understand , the igbt voltage (which can be up to 600v) doesnot affect the bootstrap operation, as i have shown in my answer , it the IC supply voltage (aka Vcc) is where the capacitor charge . but who knows i havenot tried 5v mosfet supply before. \$\endgroup\$ – ElectronS May 13 '17 at 11:42
  • \$\begingroup\$ Of course it does - if the supply is too low then there won't be enough bootstrap voltage. \$\endgroup\$ – Andy aka May 13 '17 at 11:57

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.