Hi,

My requirement is that I have an input of 10VC to 32VDC and I have to design a power supply that can convert the said input to 12VDC with a capability of sustaining load of 310mA minimum.
As i am looking for a cost effective solution it will require me to design a discreet buck boost regulator with no ICs used. The switching should done by some kind of feed back circuit.
Can anyone please suggest a design or a similar design that i can use here.

 

Hi,

My requirement is that I have an input of 10VC to 32VDC and I have to design a power supply that can convert the said input to 12VDC with a capability of sustaining load of 310mA minimum.
As i am looking for a cost effective solution it will require me to design a discreet buck boost regulator with no ICs used. The switching should done by some kind of feed back circuit.
Can anyone please suggest a design or a similar design that i can use here.

You can't do that with a buck regulator - buck regulators can only reduce the voltage, it can't increase it.
A discrete circuit that can step both up and down, that's cheaper than doing it with an IC? Now that's quite a challenge (and quite a pointless challenge)

 

Can we use a combination LDO(Discreet) and Buck regulator(Discreet) and switch between those once the voltage is less than 18V and More than 18V respectively

 

If efficiency is not important, the simplest solution would be a joule-thief type boost converter combined with a shunt regulator.
Your overall efficiency will be terrible, (especially with 32 V input) but it would have a very low parts count.

 

Hi,

My requirement is that I have an input of 10VC to 32VDC and I have to design a power supply that can convert the said input to 12VDC with a capability of sustaining load of 310mA minimum.
As i am looking for a cost effective solution it will require me to design a discreet buck boost regulator with no ICs used. The switching should done by some kind of feed back circuit.
Can anyone please suggest a design or a similar design that i can use here.

Hello,

Here is a schematic of the basic buck-boost circuit. The switch would be where the transistor goes.
Turning the transistor on and off is PWM modulation and that varies the output voltage.
To get this PWM, you would use a triangle generator along with a simple transistor made comparator to compare the triangle to the output voltage and generate the pulse width like that.
For the triangle generator you can get away with just charging and discharging a capacitor with a resistor and another transistor, or use a constant current transistor circuit to charge the capacitor and another transistor to discharge.
The comparator might be as simple as a zener diode and another transistor.

The accuracy is not going to be anything like using an IC for this or even a few separate IC chips of the more basic type like comparators and op amps.

 

i am looking for a cost effective solution it will require me to design a discreet buck boost regulator with no ICs used.

Sorry but those two requirements are contradictory.
The cheapest way to build a buck-boost regulator is using an IC designed for that purpose.
An all discrete design will likely never be cheaper.

 

I this is probably a school project. We have had a number of people asking about " My requirement is .........with no ICs ".
45+ years ago we made things with no ICs and it was hard.

 

Hi @ronsimpson, unfortunately this not a school project . It is required for a real-world product. As there are cost constraints that are to be adhered to I am more inclined on a discreet solution rather than choosing an IC.

 

The problem with controlling buck-boost is the Vout goes negative in respect to power supply Gnd. Of course it can be solved but why bother if an flyback topology can avoid this problem.
In flyback you can connect secondary Gnd to power supply Gnd.
Of course when galvanic isolation is no needed.

 

Hi @ronsimpson, unfortunately this not a school project . It is required for a real-world product. As there are cost constraints that are to be adhered to I am more inclined on a discreet solution rather than choosing an IC.

This is a patently false economy. A dedicated IC solution will ultimately be cheaper than ANY possible discreet design. If you don't believe me cost it out for yourself. Don't forget to factor in the cost of extra PCB area and the cost of hand assembling the thru hole components.

 

flyback topology can avoid this problem.

So a discrete flyback converter could work, but it still may not be cheaper than a buck-boost with an IC controller.
Below is an example circuit:
The opto coupler in the feedback is not needed if you don't need isolation between the input and output.

 

So you think you can build a discrete circuit for less than this?

buck boost converter

 

It cost me more to have a resistor put on a board and soldered down then the resistor costs. I try hard to not use many small parts.

 

Our budding crop of wannabe electronics moguls might want to dedicate a modicum of study to the "economies of scale". It explains why small fry cannot hope to compete with the "deep pockets".

 

Using a cheap IC can make it very simple. This is a buck-boost converter.
Details are in the attached application note.


Modern Chinese versions of the MC34063 can cost as little as U.S. 5¢ each.

 

Using a cheap IC can make it very simple. This is a buck-boost converter.
Details are in the attached application note.

View attachment 321914
Modern Chinese versions of the MC34063 can cost as little as U.S. 5¢ each.

Hi,

I hate those chips (ha ha) but yeah cheap and easy to use.