hello , i made a program on PIC to generate a pwm signal to control the MOSFETs on a H bridge to control a DC motor , that was the entire point behind the H bridge but our teacher told me that i must have a dead time or there will be a short circuit , which means that when i control the mosfets 1 & 3 i should turn them off then start mosfets 2 & 4 , but with my PWMs this is done simultaneously which will cause a short circuit according to my teacher , any idea how to make this dead time in PIC ???

 

The PICs with a 'Motor PWM' module can generally do the dead time automatically, you just set a register.

What PIC are you using and what pins are connected to the H bridge stages?

 

the pic i am using is 16f877A and the pin connected to the H bridge is pin 17 (which is CCP1 pin) , i only have one pin connected to the H bridge because i will use an inverter wired to the same pin to feed the other 2 transistors of the bridge

 

OK, swap whatever inverter you are using for a 40106 and add two more sections between that and the bridge. The 40106 is a hex schmtt trigger inverter.

Use a series resistor with a shunt diode, plus a cap to ground, on the inputs of the extra inverters.

Something like 10K + 100pF should give around a microsecond or so delay, with the diode across the resistor so it still turns off instantly.

That should give a delay at each changeover so both halves of the bridge are off, allowing things to settle. You can adjust the dead time by changing the capacitor values.

40106 info:
http://ics.nxp.com/products/hef/datasheet/hef40106b.pdf

 

u seem like an expert in these stuff , however english is not my first language and i didn't really understood what you meant
first things first , i am not using any special functions in the schmitt trigger inverter ???
and how can i calculate the delay exactly ??
and what about the PICS which can calculate dead time ?? can u recommend one ??? isn't it easier to use one of those than to make a complex circuit to get a dead time ??
i know these are lots of questions but i would appreciate any help , and one more question please , programming a 18f pic is the same as programming a 16f pic ?

 

If you are OK using a different PIC, look on the microchip web site -

This is a table of parts that have a Motor control PWM, have a look through to see which have any other features you need:

http://www.microchip.com/ParamChartSearch/chart.aspx?branchID=55&mid=10&lang=en&pageId=74

The PIC16 and PIC18 are similar language-wise and you would have no problems using MPLab, but if you program in C you need to check if your compiler supports PIC18. Some do both, some have different versions or options.

 

do u know any free compilers that work for pic 16F and pic 18F ???
i found a project on a DC motor control , but it's on a microcontroller called ATMEGA16 , i don't know if i can find this in my country but i would like to know if these microcontrollers are better than standard PICs ..... here's the link to the project

 

Sorry, I use the CCS PICC compiler, which is a commercial one.
It may have a free demo? - www.ccsinfo.com

Hopefully other users on the board will be able to provide information on others.

 

I've just seen a free compiler mentioned in another thread - The HiTech one here:
http://www.htsoft.com/downloads/demos.php

 

OK, swap whatever inverter you are using for a 40106 and add two more sections between that and the bridge. The 40106 is a hex schmtt trigger inverter.

Use a series resistor with a shunt diode, plus a cap to ground, on the inputs of the extra inverters.

Something like 10K + 100pF should give around a microsecond or so delay, with the diode across the resistor so it still turns off instantly.

That should give a delay at each changeover so both halves of the bridge are off, allowing things to settle. You can adjust the dead time by changing the capacitor values.

40106 info:
http://ics.nxp.com/products/hef/datasheet/hef40106b.pdf

okay now before everything i will try this method , if it works then i will not need to swap PICs , which is something i am really trying to do
now if i want to apply this on proteus
what did you mean by " add two more sections between that and the bridge"
on the output of pin 17 i will have 2 wires that will feed the bridge , one of them will be connected to the inverter , now i will use a 40106 inverter in place of the normal inverter , then what ???
and do u know how can i calculate the delay time ?? i am operating on a frequency of 10Khz , which means a period of 100 micro seconds , so i am going to need almost 2 micro seconds as a delay (on each up and down) , do u think it's enough ?? and what about the value of the capacitor
if u cant help in calculating the delay time just , please, explain to me how to wire the circuit you described above
now we're talking electronics

 

Rather than feed the exiisting two signal direct to the H bridge, connect each to a 10K resistor and the the other end of the resistor to the input of a new inverter section. Add a diode across the resistor and a 220pF capacitor from the invertor input to 0V.

The output of that new invertor then goes to the original feed point on the bridge.

That's what I meant by 'between the invertor and bridge'.

(So two new R-C + inverter sections, one in each half of the bridge feed).

The 10K plus 220pF should hopefully give about a 2uS delay in each transition.
By fitting the diodes the correct way round (depending on the bridge setup), you can bypass the resistors for the turn-off but still delay the turn-on, so you get a dead time.

 

thx for the help but i must ask , can i find a driver that does all this without building this circuit , i mean that i will give the driver (IC) the pwm input and i got an output of 2 PWMs that are inverted with the dead time i need?
i need a driver that does this and that i can find on proteus in order to make a simulation