Hello All,

I'm trying to convert direct dc to pulsed dc. Here goes the method I used so far for conversion.
First I used 555 timer in Astable mode to generate PWM pulses.(0-100% duty cycle).
These pulses are used to drive the MOSFET's gate. (SiHP6N40D), the output of the mosfet is used to drive electromagnet coil(load).

I got few problems with above circuit:
1) the mosfet output is inverted as compared to 555 output.
2) if I supply 10V to mosfet drain the max vtg read is 12V (CRO measurement). this happens for every increment of voltage eg 11V means 13V it reads)
3) after connecting load I see high voltage spikes at the rising edge of mosfet output, this also causes spikes in 555 timer output.
4) to solve problem(3) I placed a electrolytic capacitor across drain and source, the spikes are removed but there is change in the mosfet output waveform i.e the low state seems to be discharging of capacitor.
5) the mosfet is rated for 6A current handling capacity. the mosfet is too much heated with supply of 3A of current even though heat sink is provided.

Pls help me out . (sorry for my english)
Attaching the circuit for reference.


Regards,
Priya

 

Also I found that only at low duty cycles its getting hot, while at higher duty cycles MOSFET is not heated a bit. What might be the possible reasons?

 

A p-channel mosfet driven in conjunction with a NPN transistor might be a better solution. That way your load will always be grounded and the mosfet will control power to it. This will also solve your inverted output problem. Possibly something like this. This was for a low voltage setup and you'll need to choose a higher Vbe transistor and a higher current sustaining mosfet.

 

4) to solve problem(3) I placed a electrolytic capacitor across drain and source, the spikes are removed but there is change in the mosfet output waveform i.e the low state seems to be discharging of capacitor.

This is not the way to do it and may contribute to the MOSFET overheating. Instead, connect a diode across the electromagnet to catch the back-emf pulse.

 

I connected a diode MUR860G across electromagnet to catch the back emf pulse. is it something to do with gate source capacitance that leads to heating of the MOSFET

 

What you could try is to add a complementary driver stage between the 555 and the mosfet. http://www.next.gr/uploads/64/135-9914_434816708.png

 

is it something to do with gate source capacitance that leads to heating of the MOSFET

No. The mosfet should be either fully on or fully off and the transitions between the two should be as fast as possible. When the mosfet is off there is no current and no pwer is dissipated in the mosfet. When the mosfet is fully on the drain source voltage will be very low and so the power will be very low. While the mosfet is only partly switched on there will be both voltge and current and so more power to heat the mosfet.

Even if the fully on or fully off conditions are met there is still some power in the on condition (the voltage will not be zero) and how hot the mosfet gets depends on the RDSon (the on resistance) of the mosfet and heatsinking of the mosfet.

What current does the electromagnet use, which mosfet are you using, and is it attached to a heatsink?

P.S. At what frequency are you switching?

 

The mosfet used is SiHP6N40D vishay's make. The switching frequency is 1 Hz. Heat sink is attached to mosfet.

 

From the datasheet for that MOSFET, RDSon is 1Ω at Vgs=10V and Id=3A which is close to your usage I think. That is a dissipation of 9W in the MOSFET so it will need significant heatsinking. You might do well to search for a lower on resistance MOSFET to get the heat down.

 

Why are you using a 450V mosfet to switch 120V? Use a lower rated part to get better Rdson, like for example ipb107n20n3g

 

A p-channel mosfet driven in conjunction with a NPN transistor might be a better solution. That way your load will always be grounded and the mosfet will control power to it. This will also solve your inverted output problem. Possibly something like this. This was for a low voltage setup and you'll need to choose a higher Vbe transistor and a higher current sustaining mosfet.

No current limits resistor to protect the Vb of Q1?

 

will replacement of 1k to 100 ohm and 10k to 1k (as per circuit diagram attached) serve for a lower on resistance MOSFET to get the heat down?

 

does another MOSFET in parallel is required to get the heat down?

 

does another MOSFET in parallel is required to get the heat down?

Like I said before, use a different mosfet. The one I posted in #10 will easily take 6A without heatsinking.

 

For a high current coil load, the Vds better choose as 120V*3 = 360V, Vds=≥360V, Ids=3A*5 =15A, Ids=≥15V, Rds≤20mΩ.

 

I tried npn transistor(2N2222A) in conjuction with N channel MOSFET to solve inverted problem. but it didnt worked. Pls suggest how can I resolve.

 

There is some problem in my 555 timer circuit.
If I place 10uF capacitor between 2 and 6, waveform seems to be distorted. but if 100nF is placed it seems to be exact square wave. what could be the reason?

Attaching the screenshot of waveforms. circuit simulated in proteus 8 demo s/w.

 

Why are you using a 450V mosfet to switch 120V? Use a lower rated part to get better Rdson, like for example ipb107n20n3g

If I use IPP110N20N3 G (T0-220 package) what should be the gate resistance value? is it 2.4 ohm OK?

 

If I place 10uF capacitor between 2 and 6, waveform seems to be distorted.

The waveform isn't distorted but your 'oscilloscope' is distorting it. Set the coupling for channel A to DC (it is set to AC in your picture).

 

Im magnetizing em coil with pulsating supply. I use an AC supply to demagnetize the same coil. How can I integrate both the supplies such that there should be switching between them which should not affect the pulsating supply?