Mosfet driver

It sounds like you have a problem with rise and fall times of the waveform and too much time is spent transitioning between on and off states. From your description it isn't clear how the components are configured, can you post a schematic for us please.

Brian.

Hi,

a circuit problem.
Please provide
* schematic
* PCB layout
* including all part values and part names and so on

Klaus

Time to buy an oscilloscope - there are many reasons your circuit could be inadequate.

betwixt said:

It sounds like you have a problem with rise and fall times of the waveform and too much time is spent transitioning between on and off states. From your description it isn't clear how the components are configured, can you post a schematic for us please.

Brian.

Click to expand...

Hello Brain
How the pulse timing and transition time adjusted, can you tell us little about that?, as I need help understanding that sysem

All I know is that the pwm ic generating the pulses for the mosfets or transistors or IGBT's high and low, when high are on and this for a certain time as long as the pulse is sent from the ic, the low are off and vice versa,
The mosfet take time to turn on and time to turn off, etc
I'm working on Repair, I see many converters use this configuration using mosfets, 4 mosfets, 2 high and 2 low, when they burn and need replace and I don't find original mosfet number in market, I try replacement part, but I be worry about that timing issue, some do work instead of original number, such as 59N30 instead of 50N20, although they have different tr and tf,
I don't understand how they work instead, can you explain why?, and how?
Regards

If you cannot describe the design, show a schematic, as parts are not the problem.

Check if the power supply is capable of delivering enough current without excessive ripple. Consider using decoupling capacitors close to the ICs. Add heatsinks or improve airflow around the components. your PCB design should allow sufficient heat dissipation. Using thicker copper layers or adding thermal vias. Here is another detailed MOSFET driver design: https://www.pcbway.com/project/sponsor/Mosfet_Driver__Controller.html
You can study this too.

The problem from burnout is inadequate cooling when overloaded with current, or lack of OTP or OCP protection from loads with 10x large surges like tungsten lamps and motors starting on full voltage.



However, your problem is a lack of drive voltage for switching to get low Rce ( or low RdsOn in FETs) due to capacitance loading.

Stop thinking about frequency and start thinking about min
and max pulse width. Unless you're hugging 50% duty,
which transformer-type topologies will make practical but
straight buck converters see rarely.

"Found" transistor models may or may not show you saturation
recovery time or its dependence on Vce(sat) and Ic and temp.
Switching transistors might. Ones meant for small signal HF or
audio will expect you to "don't do that" and shine on any such.

So you know nothing (or, what do you?) about shoot-through
which is #1 way to burn an output stage, if you do enough and
often enough that some condition (likely Tj, followed by you
spike a junction with aluminum in that old crusty standard linear
technology) is violated.

I have seen some crappy discrete bodges shown as MOSFET
drivers recently. There's a few good reasons why MOSFET
driver ICs are so numerous.

Transistor has good properties @ 1A Vce(sat)= 0.8 at Ic/Ib = 10 50 MHz 30V



6N137 has good properties using 10 mA input, output current = 5V/330R = 15 mA or 1k to 15V then transistor can boost 150x base current until Vce<2V then reduces to 10 when saturated.

Are you using 10 mA input and 15 mA output on Opto?

I have to wonder about that "fT" of 50MHz but said to be
"typical frequency" (???). If that's the classic fT then it's
one dog-slow transistor (even when not in saturation).
It would not be surprising if you got into "valve float" at
higher frequencies, maybe failing to fully assume the
"off" and "on" gate levels and putting things linear that
should be switched.

Where's the 'scope photos in all this - 2825 output &supply,
"driver" output / FET gate, driver supply, FET drain?

Because we could make sh!t up all day and get nowhere.