Ir2117 mosfet driver

[mah]

i used ir2117 to drive mosfet for dc dc buck converter the input for the driver is pwm from microcontroller i attached the circuit but there is a problem the output from this driver is 9v pure dc not pwm like the input

 

[ronsimpson]

This type of "hi side" driver can not be on 100"% of the time. Start off at 50% to see if it is working.

I think 90 to 95% of the time is maximum.

 

[mah]

This type of "hi side" driver can not be on 100"% of the time. Start off at 50% to see if it is working.

I think 90 to 95% of the time is maximum.

so what to do now ?it should be working at any pulse duration.

 

[misterT]

so what to do now ?it should be working at any pulse duration.

You need to replace the bootstrap circuit with an active voltage multiplier. A Villard Cascade etc.
https://en.wikipedia.org/wiki/Voltage_multiplier
**broken link removed**

 

[ronsimpson]

so what to do now ?it should be working at any pulse duration.

That driver way never built for 100%.
Most PWM applications don't need 100%.
Are you sure you need 100%?







%

 

[mah]

100% of what ? what do you mean?

 

[misterT]

100% of what ? what do you mean?

Duty cycle.

 

[ronsimpson]

What you have will do: all the time off through 90% on. It will not reach all the time on.
What frequency will you be running? See MrT's pictures.
What voltage? What current? Is the load inductive or resistive?
It makes a difference.

 

[mah]

but i didn't mention that i need 100% of duty cycle . i need portion of duty , the problem is that the output was 100% , there were not pulses

 

[mah]

What you have will do: all the time off through 90% on. It will not reach all the time on.
What frequency will you be running? See MrT's pictures.
What voltage? What current? Is the load inductive or resistive?
It makes a difference.

pwm will be changed from 0-90% with constant frequency 100Khz
the load for now is resistive i will upgrade to full buck converter
input voltage is ranging from 14-17 v

 

[misterT]

but i didn't mention that i need 100% of duty cycle .

You said "it should be working at any pulse duration."

So the problem is that the circuit is not working at all? How did you measure it?

 

[ronsimpson]

but i didn't mention that i need 100% of duty cycle . i need portion of duty , the problem is that the output was 100% , there were not pulses

In MrT's last picture (100%) "there is no pulses" and in his first picture (0%) there is no pulses.
The driver you chose can handle 0%. It can not handle 100%!

100% (and 0%) is "no pulses".

You must not allow the input to be high all the time!

 

[ronsimpson]

pwm will be changed from 0-90% with constant frequency 100Khz
the load for now is resistive i will upgrade to full buck converter
input voltage is ranging from 14-17 v

It should work fine from 0-90%.
100khz is good.
14-17V is good.
BUCK will need a diode added! Don't allow the MOSFET to go negative.
What is the problem?

 

[mah]

its output is dc(100%) not pulses like the input

 

[ronsimpson]

I see the problem now.
The input threshold for the IR2117 is very high.
You probably are sending a signal 0/3.3V or 0/5V.
The IR2117 thinks any signal less that 40% of VCC is low (15Vx40%)
Any signal above 50% of VCC is high. 50% of 15V is 7.5 volts.
So you need to send a signal into the IR2117 that is 0/10 volts. (0/vcc) where vcc is in the range of 10 to 20V
The driver was never built to interface to a micro.

 

[mah]

oh my GOD!! will i send it to garbage now? or there is modification?

 

[ronsimpson]

I have used a very small MOASFET like a 2N7000A. Any small "logic level" mosfet. The 2n700a had a gate turn on voltage of 2.5V max.

S=gnd, G to micro. D to IR2117 in put and a pull up resistor to 15V. Burn 1/4 watt in the resistor.
Problem: this will invert the signal! Can you fix that in software?

If you can't, I can think of another way.
---edit----
Is your micro at 3.3V or 5.0V?
---edit again---
With out inversion:
Gate to micro's supply (3.3 ro 5.0V)
Drain to resistor and resistor to 15V (10 to 20V)
Source to micro.
When the PWM pin of micro is high there will be 0 volts (G-D) on the little mosfet and thus the transistor will be off. The (S) will be at 15V.
When the PWM pin is at 0V there will be more than 2.5V (G-D) and the little mosfet will be on and pull the (S) down. Warning the micro will have to handle all the current that is going through the resistor. Maybe 10mA.

 

[mah]

this is a great idea my micro is working on 5 v , it is atmega 8

 

[mah]

could i use BJT instead of 2n7000a

 

[mah]

this is the complete circuit with BC549 instead of 2n7000 , it is working but collector resistor is 200 ohm to give this response while increasing this resistor the output on the mosfet gate is restorted . i need to know the safe limits for this resistor .is there any way to get the same signal without restorion