That or use a p-channel instead. It will be inverted, but then the gate voltage has to be 10V lower than Vcc (the plus power supply side) for it to work.

If I get a chance tonight I'll draw a hypothetical circuit to show how this could work.

 

This is a crude circuit I came up with that can drive the NMOS:


And a similarly crude PMOS circuit:


Neither gate drive is very good but they work to some degree (at least in simulation ). The gate-source voltage, and hence the on-resistance, will vary with input voltage.
I picked whatever random suitable parts LTSpice had and resistor values are just picked for the correct ratio.
One problem is the PMOS driver is asymmetric from storage time.

With the PMOS note the orientation of the symbol. A common mistake is to put it down backwards and then it obviously wouldn't work.

 

I modified the original circuit so that the load is at the drain of the original N-channel Mosfet. Its gate needs 10V pulses that can be from the 555 that has two diodes so that its PWM is from almost 0% to almost 100%.
For 50kHz its filter capacitor value is much too high.

 

to bill marsden;

could give me the conditons for how the switch will be on, ie like previous post the nmos Vd = 30V, Vg needs to be 10V so i need 45 on Vg or something? I mean could explain the conditions for PMOS

Ghar:

I will simualte n see what happens thanks!

audioguru:

50Khz is too high meaning????... but will that circuit work?

 

to audioguru... is that still a buck converter?

 

sorry guys the problem i get from using PMOs on the buck converter is that... for 18VDC ill get a boost instead?!?!? when i change the switch to NMOS i get 6V... this is without the timer.. im only ttesting the buck converter... and my critieria for this experiment is to step it down to 6V.. when i change the switch to pmos... it doesnt work... why is EEE so hrad

 

Hi guys, (attached) buck w/o 555timer... however could you explain to me more about PMOS like how you guys did with NMOS? regarding the parameters on the datasheet like to see which is important and also why for PMOS when i increase the duty cycle of the pulse... it actually decrease voltage instead of increasing it.... im going to implement it on the 555timer.. reason y i need 555timer is because i actually hafta build it later on.....

hope to hear from you helpful people agian soon. really much appretciated

 

to audioguru... is that still a buck converter?

No.
It is a boost converter like you said in post #59. It can also be a buck converter if the PWM is narrow pulses.
You should have said you want a buck converter in your first post.

When the Mosfet is turned on most of the time then it is a boost converter.
When the Mosfet is turned off most of the time it is a buck converter.

 

OK here is how it would work for a positive power supply...

For a negative power supply the MOSFET polarity would need to be reversed.

.

 

could give me the conditons for how the switch will be on, ie like previous post the nmos Vd = 30V, Vg needs to be 10V so i need 45 on Vg or something? I mean could explain the conditions for PMOS

Vs = 20V, it's the input voltage.

You want Vgs = -10, or Vsg = 10

This means you need the gate 10V below the source.

The schematic you posted in post #68 is wrong. You have the voltage source across gate-drain, not gate-source. Again, note the symbol. In a PMOS the source is connected to the higher voltage.

The reason the voltage vs duty cycle behaviour is inverted is because the PMOS turns on when the gate voltage is low.

Also, I'd recommend you switch to using LTSpice. It's free without any limits first of all, and second it works infinitely better than OrCAD for switchers. OrCAD takes forever and often doesn't converge. LTSpice is great for this, it was actually designed for switchers.

No.
It is a boost converter like you said in post #59. It can also be a buck converter if the PWM is narrow pulses.
You should have said you want a buck converter in your first post.

When the Mosfet is turned on most of the time then it is a boost converter.
When the Mosfet is turned off most of the time it is a buck converter.

That is definitely a boost converter. With the FET off forever the output voltage is equal to the input voltage, minus some losses (like a diode drop).
It's not ever a buck. And he did say buck in the first post...

You're thinking of a buck-boost converter, which still has the high-side switch and the inductor goes where you have the FET, or you need more switches if you want it non-inverting.

 

sorry... i didnt make it clear.. its a buck converter.. ... 19V -> 6V....

 

hey ghar, thanks. i will download LTspice and try... ORCAD seriously is making me wait a long time just to simulate this...

 

hi guys, so ive tried it using pmos.. however, now i realized to make the buck work... i need to supply a -5V to the gate... anything less than that wouldnt trigger it..... do you have any idea how i could invert the pulse coming out from the 555timer output to be negative? will nand gate work? but adc??!!? :/

 

im gonna try it using op-amp as an inverter... wish me luck

 

What?

You need about 5V gate below source minimum... not -5V.
Take a look at the circuit I posted with the PMOS.

 

yah i havent tried it on 555timer... but if i put 5V for the circuit as shown on post #73... it wont work... dont know why it needs to be -5 tho !!!!

 

ive tried >5V ... but after only 15V i get the an output of 16V whcih is not a correct value.. sigh.

 

okok ive changed the VPulse polarity to the gate.. to -ve... gonna see what happens if i simulate it with 555timer..

 

The PMOS needs gate below source... your voltage source in post 73 has the plus on the gate, which means yes, you need to have negative voltage in that situation.

 

from datasheet: On state drain source IRF1931 = -12V
gate threshold voltage = -2V
therefore -12V+-2V = Vgs?

Let me clear this up... My Vs = 19V, Vg = ?, Vgs =-14V is that right?
Thus be Vg has to be more than 5V? Vgs > Vg - Vs? Yes?

I have changed the initial without 555timer Vpulse to generate the same pulse as the 555timer +10V to 0V (picture 22), therefore now my Vg is 10V, however, output still no change as shown (picture 23).

When i change it to +15V, my results settles down to 7~8V (as shown picture 24)

Tested it with the 555timer (picture 21)... i need 15V, for the output...however my concern now is that...im using ts555cn... its a cmos timer... can it produce that much voltage at the output or is cmos fixed at 5V output? The datasheet says that Vcc ranges from 2V to 16V but thats the input of the 555timer...

Another thing is that i have to invert the dut cycle of the pulse.. meaning for buck D = Vout/Vin now its inverse... isit because of the PMOS switch im using? hmm? meaning the larger the larger the pulse width the lesser the output..

ive been workinng on this could you help me to analyse? Thanks alot for keeping up with me guys, hear from you all soon...