Hi, I am building a SMPS using IRF840 MOSFETS. Using PWM from a PIC at 5V, this drives a MOSFET driver TC4420 whic is at 5V as well but can run at 12V if necessary. The TC4420 then drives a high side driver for my MOSFETS shown in the picture below. The problem is the output from this high side driver is giving the wave form I have drawn along side. The MOSFETS just blow when I turn the voltage up and I think it may have something to do with the fact that the high side driver has the wave form in the picture.

If any one has any ideas, please help. I'm using a two switch forward topology to make the SMPS.

 

Gee, you're using a logic-level low-side driver for standard-level MOSFETs. I wonder whatever could be the problem?

 

Could you explain a bit please? This setup worked fine for someone a couple of years ago and i've just modified it to suit my specs.

 

As far as I can see, the IRF840 has a 500V drain source break down voltage and has a gate threshold voltage of 2-4V so it should work fine with the logic levels??

 

Here are the actual readings off the scope. The outputs here are not connected to the MOSFET however, they are just floating so i'm not sure if that has anything to with the fact that the outputs get so distorted under higher pwm values.

 

I'll try modify to this and see if that helps.

 

I used N=(V*t)/(B*A) which is number of turns for the transformer is equal to the voltage applied times how long its applied for (maximum) divided by the product of the area of the core and the maximum flux desity which I set to lower than that of the core for safety reasons.

If i add more turns, but the voltage ,time and area stay the same, then the B value which i set to be lower thatn the max allowable value, must get lower. So would that help in reducing the amount of built up energy in the core? Allowing the core to reset?

 

This is when the frequency is set to 50khz instead of 5khz. It seems better as there is no build of charg befor the nex cycle of pulses. Can any one suggest why it works better like this?

 

If you're using 5v for the TC4420's Vdd, you will only get the IRF840 into the linear region. Download International Rectifier's datasheet, and look at the plot for Figure 1.

You need to give it 10v+ to get the IRF840 fully turned on.

Meanwhile, do you really need that high of a Vdss? If not, you should use a MOSFET that has a Vdss rating perhaps 25% higher than you really need. Otherwise, you will have an unnecessarily high gate charge for your Rds(on); as Vdss, Rds(on) and gate charge are all related.

 

Well, its connected straight to the 220V mains, so i figure with DC rectification before and a saftey factor 500V is perfect?

 

Hello,

We do not support circuits connected directly to the mains.
We do not want ANYBODY to get hurt.

Bertus