Continue to Site

Welcome to

Welcome to our site! is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

IR2184S SD and IN pins swapped?

Not open for further replies.


Apr 29, 2022
Reaction score
Trophy points
Activity points
Hi everyone,

I'm new to this forum and from the Philippines. I have this customized requirement for a 3-phase BLDC drive. I can't use any of the commercially available electronic speed controllers available on the market. They are only capable up to +60Vdc as motor supply voltage and I never stumbled for a much higher supply voltage. With that, I designed and created my own BLDC driver, using at first L6384 mosfet/igbt gate drivers. The circuitry ran well but I have been using +24Vdc as initial motor supply voltage. This is to ensure that any idiocy in my part will not fry the devices. I made it to a point that the whole application is modularized for easy troubleshooting and/or repair when the time come. By the way, there was no datasheet for the 3-phase BLDC motor whatsoever (3-wire). My measurements revealed that the stator winding was organized in a Wye arrangement with no neutral point. At any given time, any two winding has a DC resistance of 9 Ohms. And that, its nameplate has 270 Vdc, 400 Hz terminal voltage. I have provided a PWM block with adjustable frequency from 60 Hz to 24 kHz, and an adjustable duty cycle of 5%~95%. I made sure with that using an oscilloscope from a local colleague. I have also provided a chopping control using LM393 comparator and a 74HC08 AND gate for gating the shutdown pin of the L6384. All worked out perfectly well, but I cannot spin the motor way beyond more than 3 rotations per second, sweeping the frequency from the lowest setting. At higher frequencies, of phase input signals, the motor growls and stalls. The idea is to make a sensorless back-EMF, but I haven't implemented that one yet. The goal was to first run the motor in an open loop trying to characterize the motor then. When I increased the motor supply voltage to +60Vdc, all was well but there was an audible rattling noise on my board as well (not on the motor itself). That has caused and raised a red flag to me. But when I adjusted the duty cycle to say 25%, that noise diminishes. Please take note, that at any time, not of the 6 mosfets was heating up. As I raised the motor supply voltage to +150Vdc, and duty cycle of pwm signal to about 20%, no evident misbehavior on the circuit board, except that the mor was stalled (maybe because of too low applied power)? As I slowly ramped up the duty cycle, there goes BOOM!!!! All 6 mosfets were doomed.. All gate to source junctions and gate to drain junctions broke down, they shorted internally. But the worst was that, my mosfet gate drivers suffered the catastrophe.


Unfortunately, when I try to recover the board, our local supplier has advised that they ran out of L6384 ICs. The closest alternative is IR2184S. As I compared their datasheets, both input signals are the same, but with a much higher output current for driving mosfet gates. So, I temporarily created a workaround by stuffing the components on a perf-board just to revive the ongoing implementation. Here is my schematic diagram (I intentionally removed the bypass capacitors.):


Right now, I am using +24Vdc for the motor supply voltage for concept proofing. I put on pull-down resistors for ensuring that when any input pin was left floating, there should be low input on IN pin and that shutdown pin SD is pulled low. This will ensure that the mosfet pair will float, and will not cause any trouble. With the current bootstrap values, I am able to turn on the mosfet without any PWM, for as long as 30 seconds. The weird tings are:

1.) For any given IR2184S, the supposedly HIGH input on IN pin, and LOW input on SD pin will float the mosfet pair. It only turned off, and not floating. I made some dummy load, a 24V incandescent bulb. One of the leg was tied to +24Vd, and the other end to the drain and source pins of mosfet pair. The bulb lighted up, because if the junction was floating, the bulb should not light at all.
2.) A HIGH input on IN, and HIGH input on SD turns on this mosfet gate drive. with high side mosfet ON, and low side mosfet off.
3.) A LOW input on IN does not affect the operation at all.
4.) The SD now controls the turning on/off of high side and low side mosfets.

I apologize for a long description of the problem and its background. I just want to find out if someone out there has encountered the same problem that I am encountering now? It seems that IN and SD pins were interchanged, especially their functionality. I am very sure and very confident that I wired the IR2184S correctly. My colleagues had verified the schematic and the actual circuit board to be consistent, and that connections to IN and SD pins were correct, as per IR2184S datasheet. I tried to search for any erratum regarding this device but found none.
Any positive insights and advises will be greatly appreciated. Thanking in advance..



Running a BLDC motor without feedback is quite difficult. You need to know all the motor parameters and the input voltage.
PWM duty cycle, as well as 3 phase rotation frequency needs to fit motor RPM as well as input voltage and in best case even torque.

You can't just lower the bus voltage and it still will work. You get a lot of motor current or low torque.
All parameters need to fit. You can't change one parameter while keeping the others.

Feedbacked operation is much easier. Just ramp up the duty cycle and the RPM will adjust itself.

There are dedicated BLDC driver ICs. I'd use them.


Not open for further replies.

Part and Inventory Search

Welcome to