Mosfet Driver using 555timer ( Buck-Boost converter)

[Nilesh]

Hello all!
I was working on to design a simple buck-boost converter to drive a constant load with a maximum power output of 50W.
I simulated my design in LtSpice initially driving the mosfet using pulse function of LtSpice(as shown in attachment). Everything worked fine.
But as soon as I used the 555timer in it's astable mode to switch the mosfet at 50kHz, something strange was observed. LtSpicce shows power dissipation of >200W across both the mosfet and the input source to the converter circuit!!!.


Can somebody help me with the issue???

I'm attaching both the schematic with and without the 555timer driver . 
I'm sorry for the poor design of the circuit. I'm a beginer......

[Zero999]

It's because, with the 555 timer, the P-channel MOSFET never turns off. The MOSFET's gate needs to be taken to near 24V to turn off, which the 555 can't do, because it's only powered by 6V.

Connect the 555 timer's supply to +24V and its negative rail to +12V and you'll find it will work.

The circuit also lacks feedback, so the output voltage is dependant on the load resistance and supply voltage.

Anyway use the 555 timer? It's one of the worst choices. 

[Nilesh]

Thanks for your reply...

I want to mention that I have used an N-channel MOSFET.
And, I'm going to use a constant load and constant input voltage so I did not provide any feedback

I'm attaching few more waveforms for a more detailed view of the circuit. Please have a look!
The waveforms include...
(i) PWM signal from the 555 timer IC.
(ii) The voltage across the N-MOSFET.
(iii) The output voltage across the load in BUCK mode.
(iv) The current through the 24V source.

(i,ii,iii are steady state waveforms)

Thanks.

[Zero999]

Thanks for your reply...

I want to mention that I have used an N-channel MOSFET.
And, I'm going to use a constant load and constant input voltage so I did not provide any feedback

I'm attaching few more waveforms for a more detailed view of the circuit. Please have a look!
The waveforms include...
(i) PWM signal from the 555 timer IC.
(ii) The voltage across the N-MOSFET.
(iii) The output voltage across the load in BUCK mode.
(iv) The current through the 24V source.

(i,ii,iii are steady state waveforms)

Thanks.

Oh, I misread the schematic. 

Then the problem is the reverse: the N-channel MOSFET is acting as a voltage follower and isn't turning on enough. You need to use a P-channel MOSFET, with the 555 timer's positive rail connected to +24V and its negative rail to a +12V supply.

Refer to the link below: notice how the buck-boost converter uses a P-channel MOSFET. If an N-channel device is used, then the gate voltage will need to be driven above the supply voltage, in order for it to turn fully on.
https://www.allaboutcircuits.com/technical-articles/analysis-of-four-dc-dc-converters-in-equilibrium/

Please post the .asc file an use labels on the nodes, so it's obvious which plot is which.

[T3sl4co1l]

You misread again   The driver is referenced to the transistor source with a floating supply; it should be fine.

It's not clear if the 200W dissipation is due to including startup transients in the calculation.  10ms is a very short duration to average over, in realistic terms!  You simply need to select the steady part of the waveform and do the power calculation out there.  (You didn't show the power calculation so I have no idea if you made this mistake, or did all this already and still got the large result.  Just guessing here.)

I have a different complaint, however:

This is a naive circuit -- while PWM is part of the operation of a switching converter, it's not what we're doing; it's secondary to the intent.  What are we actually doing?  Switching inductor current.  Why?  Because inductor current dictates transistor and diode dissipation, and if that current is allowed to hit crazy levels (like during the startup transient pictured!), transistors and diodes will explode!

So we prefer circuits which control PWM, to set inductor current, and then we can just about guarantee our transistors/diodes will survive even when the end user accidentally shorts the output.  We don't care what the PWM% actually is, as long as it gets the inductor current where we want it.

This is called current mode control.  (The fixed-PWM example is called "open loop", while a controlled-PWM circuit is called "voltage mode".  Both are inferior, or require protection circuitry hacked on, just to emulate what a current-mode controller does from the start. )

This is kind of hard to do with a 555, but it can be done.  Here's an example in the boost topology:
https://www.seventransistorlabs.com/Images/555%20Boost.pdf
Note that R13 is a shunt resistor that senses Q1's drain-source current, and Q3 (relative to the threshold set by R4+R8) measures that current, and shuts off IC1 when that current goes above the setpoint.  The setpoint is in turn determined by IC2A, the error amplifier which ultimately regulates output voltage.  This is all a bit convoluted I'm afraid, hence Q2 and support components -- it's a forced fit to make the 555 operate this way, and there are much better ICs out there to do this same job, like UC3842.

For your application (inverting buck-boost), most controllers will need a high side gate driver and supply.  This isn't hard to solve, but does increase the component count.  Alternately, a more common buck controller can be used, in "bootstrap" connection, to make the negative supply.  Or a Ćuk converter (which uses a double inductor cleverly) can use the same boost circuit, while creating a negative output.

Tim

[Nilesh]

Thank for your reply!
It took me much time to understand your view,sorry,I'm still a beginner and I believe you guessed it looking my design

 (i) To clear any confusion about the power dissipation(as shown in ltspice), I'm attaching the .asc file. There are two of them. One with the 555 driver and one with the pulse signal used. Please have a look.

(ii) I learned that the PWM is not there just to switch the MOSFET, but to switch the inductor current at safe values. Thanks.
But I got a doubt. The circuit here only has a duty ratio between 0.3 and 0.7(since I'm not going to allow the value of R3+R5 to go beyond 100 ) . And the converter circuit without the 555 driver circuit can be taken to much higher or lower duty ratios. But still, the one without the 555 driver shows the power dissipation in nW range!!! . And as soon as I add the driver in the same circuit in place of the pulse signal feature of ltspice, I get that Watts of energy dissipation.
The point is, if controlling the PWM, to get safe ranges of the inductor current, is the problem here than the results must be the same in both the cases if I'm not wrong!.

Only one strange thing I noticed is that in the 555 driver circuit the inductor current is starting from 11A and finally settling to lower values but in case of pulse driver circuit(without 555 driver), the inductor current starts from zero, undergoes an undamped oscillations with 11A of peak current and settles at lower values
The point is, if controlling the PWM, to get safe ranges of inductor current, is the problem here than the results must be same in both the cases if I'm not wrong!.

Please have a look at both the circuits 

Thank You!

[Nilesh]

I'm attaching the .asc file.
Please have a look!

[Zero999]

Yes I misread it again. In my defence, it's poorly drawn, which makes it difficult to read and easier to misread, although I should have waited until I had more time to look at the schematic before replying.

Here's the last schematic neatened up a bit to make it easier to follow.

The MOSFET only dissipates an average of 3W. Large spikes in power dissipation are to be expected, especially in start-up.

[Nilesh]

Thank you very much Zero999 for your time

[Nilesh]

I got all my solutions 
Thanks everyone

[Siwastaja]

Large spikes in power dissipation are to be expected, especially in start-up.

Are to be expected, but may possibly blow up the MOSFET as well, since they are real, not simulation artifacts. Hence the earlier suggestion of current mode control (highly recommended), or minimally, a corner-case hack (like softstart PWM rampup) that works 99% of the time and only blows up 1% of the time.