NOT gate (transistor logic), high frequency operation

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
Hi,
Iam generating pwm pulses by comparing sine and triangular signals.
I need two pwm signals which are complementary to eachother. I'll use these two to drive a H bridge MOSFET inverter.
thankyou
 

SgtWookie

Joined Jul 17, 2007
22,230
It will invert the input signal, but you will need some provision to ensure that both the high and low side MOSFETS cannot be conducting at the same time. Otherwise, you will wind up with the dreaded "shoot-through" condition, where the MOSFETs both being turned ON cause a direct short across the voltage supply.

The 470 Ohm resistor limits the maximum gate charge current to 19mA. As the gate charges, the current will decrease. This will result in very slow turn-on times, and as a result, a lot of power being dissipated across the MOSFET as heat.

It would be much easier to use a MOSFET H-bridge driver IC than try to design one yourself.
 

mik3

Joined Feb 4, 2008
4,843
what is the maximum frequency i can work with this?
It depends on the transistor used and other capacitance in the circuit (mainly the input capacitance of the H bridge driver). However, PWM frequencies are usually below 500 kHz thus it will be ok.
 

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
the MOSFETs both being turned ON cause a direct short across the voltage supply.

It would be much easier to use a MOSFET H-bridge driver IC than try to design one yourself.
ok. u mean overlap of conduction periods! i want build a low cost circuit.
IS there any alternative for driver IC?

Any suggestions on this link.
http://www.bobblick.com/techref/projects/hbridge/hbridge.html
no driver IC used here.
 

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
i have seen some datasheets of driver ICs reg.
But im unable to slect the cheap n best one.
can u suggest me a good one here.
 

SgtWookie

Joined Jul 17, 2007
22,230
MOSFET driver IC's are actually pretty inexpensive, and they can source/sink a good deal of current all by themselves; some can go several Amperes.

You need drivers that can source/sink plenty of current in order to turn your power MOSFETs on or off in the shortest amount of time possible. It would take you quite a bit of time to attempt to even approach the performance of a MOSFET driver IC, and your time must be worth something.
 

SgtWookie

Joined Jul 17, 2007
22,230
We have absolutely no clue what your requirements are.

We don't know what voltage and current levels you are planning on switching.

We don't know what MOSFETs you are considering.

How do you expect us to make a recommendation based on nothing?
 

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
Ok. here are my details.
mosfets: z44
inverter output frequency : 50Hz
triangular signal for PWM : 1KHz
Dc input to the inverter: 12V
 

beenthere

Joined Apr 20, 2004
15,819
Driving a FET with a triangle wave will lead to huge losses in the device due to ohmic heating. They are efficient only if turned on and off as quickly as possible. That makes a square wave drive very important.

You can use that triangle waveform applied to a voltage comparator to vary the PWM duty cycle.

What current will you be driving?
 

SgtWookie

Joined Jul 17, 2007
22,230
Do you mean IRFZ44? Vdss=55, Rds(on)=17.5mOhms, Id=49A?
Total gate charge is 63nC. That's going to take some current.

Why don't you post your schematic thus far?
 

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
Hi,
actually i dont have any particular schematic.
im implementing my idea here. i already posted a thread about inverter design.

just check this.
http://forum.allaboutcircuits.com/showthread.php?t=21571

i have designed the pwm circuit.
two XR2206 (sine & triangular), a comparator (741 with +/- 12v).
Iam ready with four Z44 mosfets assembled on a single heat sink and are electrically isolated with mica strips.

What i need is a not gate. (it is confirmed with the transistor logic)
and now a new problem with overlaping of conduction periods of mosfets.

finally,
im not going to connect any load here. my intention is to show the output in oscilloscope.


now can u suggest me a simple n cheap mosfet driver?
 

Attachments

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
hello.
check the above post. can any body suggest me a simple low cost mosfet driver. my circuit is ready with pwm pulses.
tahnkyou.
 

SgtWookie

Joined Jul 17, 2007
22,230
Why don't you take a look at ST Microelectronics VNH2SP30-E IC.

It's a fully integrated automotive H-bridge, capable of sinking/sourcing up to 30A @ 41v.
The inputs are at logic level, so it would be simple to drive. You would not have to worry about fooling around with discrete power MOSFETs, cross-conduction (shoot-through), or many of the other things you may experience with your current plan.

About the only concern will be adequate heat sinking for the IC itself. There are suggested layouts and thermal calculations beginning around page 19 in the datasheet.

Digikey stocks this part for under $12 US.
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=497-4500-5-ND
I have no idea what the availability would be in India.
 

Thread Starter

onlyvinod56

Joined Oct 14, 2008
369
hi
Why don't you take a look at ST Microelectronics VNH2SP30-E IC.
i tried for that. But it takes 4 ~ 5 days for shipping. Anyways i dont have that much time. I have only two days left. I have an idea.

Shal i replace the H-bridge with a push pull arrangement (a centre taped transformer with two mosfets)? There will be no chance for 'sohot-through'. If both the switches are conducting, it forms a parallel path but not short circuit.

see the attachment. I have simulated the circuit. im attaching the results here. to get the an approximate sine wave, i need to use a capacitor.

here are my available hardware details:

centre taped transformer: 230 / 12-0-12 (actually it is stepdwn T/F. iam going to use it as step up, as i dont need ato drive any loads here)
R(load) : 1K ohm (used in simulation)
Capacitor : 1000uf ( -do- )

i dont know how much stress the transformer primary (12-0-12) can withstand. So i'll use a 7805 or 7809 to feed the transformer.
as the turns ratio is 230/12 = 19.1666,
if i feed 5v, the secondary will be 5*19.1667 = 95.835 (100v approx)
9v, the secondary will be 9*19.1667 = 172.5003

If my calculations are correct, i'll be happy. But here is a problem.
I need to use a capacitor to get an approx sine. the voltage rating of the cap will be approx 200v if the input is 5v DC.

i dontknow whether i'll get a 1000uf/200v within these two days.

please help me. what i need is, an approx sine in oscilloscope. leave the harmonic content/THD
 

Attachments

Top