The source of command is a 74HC541 and before that is a microcontroller.That is quite a few parts for an apparently simple application. What is the source of the "command" signal that goes to pin 3 of the Ir2104?
What kind of diode is D01?
Why does the "C" underneath D01 have two values?
You say there is not fast switching, but there should be for a device that does not want to spend an excessive amount of time in the linear region.
How do you plan to implement Res1 so that it can dissipate 550 watts?
Did you engineer this design or are you cutting and pasting?
Thank you, If we want to switch 4 MOSFETs by a driver and the distance on the PCB is approx 10 cm, in your opinion I must change the circuitI assume R? is the load.
Place the IC near the MOSFET and remove most of the Gate parts. Reduce R2 but probably not important.
If you have a resistive load you do not need the RCD snubber. Maybe just the diode across the load as shown if you think the load will have some inductance.
View attachment 283870
What spike are you expecting? Resistive loads don’t cause spikes as far as I know.I tried to reduce spike voltage with an increase in the turning on time.
The load might has a little inductance and may this L and parasitic capacitor cause spike voltage in turning onWhat spike are you expecting? Resistive loads don’t cause spikes as far as I know.
100V 200mA diode seems small. I would have used a 200V 1A fast diode.D01 is 1n4148
In fact, maybe we drive 4 MOSFETs by a driver, and because the driver is compact and we chose it.I find it odd to use a gate driver, whose purpose is to supply lots of current for fast switching, then limit the current out to 25mA to slow down switching.
We didn't check it, but we encountered a problem where the MOSFET burned out in the past.That raises the question of whether the spike was observed or hypothetical
The connection is a little long and the load is a heater and the entire path of the load circuit is approx 1 meter.If the connection to your load is very long or if your load is inductive the spike you mentioned might be capacitive coupling of the gate signal into the drain. An image of the spike compared in time with the gate drive signal would tell you a lot.
Because the current of gate is small we choose 1n4148 and it's enough fast, in your opinion why is necessary 1A diode?100V 200mA diode seems small. I would have used a 200V 1A fast diode.
If you are switching with 470 ohms of gate resistance, the turn off speed is slow and there probably won't be any spike.
Ron said:Mousivandhossein said:
D01 is 1n4148