linear power supply design help plz

[ant9985]

hi, i got an assignment asking me to design a linear power supply to meet the following specification:-
selectable input voltage 110/230v ac
Adjustable output voltage 24-26V
Maximum output current 1.5A
load regulation +/- 1.0%
line regulation +/- 0.5%
ripple and noise <50mV pk-pk

iv worked out the transformer is going to be rated at 39VA
and the ripple suppresion capacitor is 0.3F.
But now i need to work out what zener to use and the resistor's resistance that goes in series with it and i dont know how to work out the bias current it needs. please can anyone help. also if anyone could tell me anything else i need for this, like what load and line regulation are or how i make my power supply to fit them. My lecturer just gave me this assignment without any teaching for the subject, and i do not remember doing this kind of stuff on my previous course. thnx :shock:

 

[JimB]

You seem to be designing the most horrendous inefficient power supply imaginable.
A 300,000uF capacitor in a 1.5 amp PSU is just plain silly.
Your specification imposes limits on ripple, line and load regulation and you are trying to achieve this with an enormous capacitor and a zener diode.
You also have the output voltage as being adjustable from 24 to 26 volts, how are you going to do that with a zener?
I suggest that you think again.
In your other post on the subject of this power supply, you say that you cant use ICs. Try using a few discrete components to make the error amplifier and a pass transistor.

JimB

 

[ant9985]

iv got no idea, iv never designed a power supply before, iv been following the text book they recomended and thats what it came to, i thought it was an overly large capacitor, and i asked my teacher about it, and he said yes thats what it will be. but thnx for telling me where i went wrong. ill just tell him i cant do it.

 

[danielsmusic]

you can do this for about £40 buy a transformer that outputs 30v there sould be 3 wires on one sid and 2 on the other. this means you have 1 coil on the mains side and 2 coils on the output side if you put the ouput coils in parralell you put 240vac in the mains. if in series you put 110vac i would say use 5 caps 22000uf, 100uf, 1uf, 1pf and 1nf put them all in parralel from the output of the transformer. use a tip32c transister to vary the voltage. here is a curcuits diagram i through together.

just being nosy, how old are you.

 

[Nigel Goodwin]

you can do this for about £40 buy a transformer that outputs 30v there sould be 3 wires on one sid and 2 on the other. this means you have 1 coil on the mains side and 2 coils on the output side if you put the ouput coils in parralell you put 240vac in the mains. if in series you put 110vac i would say use 5 caps 22000uf, 100uf, 1uf, 1pf and 1nf put them all in parralel from the output of the transformer. use a tip32c transister to vary the voltage. here is a curcuits diagram i through together.

That's pretty horrible as well :lol:

And why all the capacitors in parallel?, all you need is one reasonably sized one.

The idea of using a transistor to stabilise and vary the output voltage is the correct one, but NOT just fed from a potentiometer!. It needs to be part of a proper stabilised PSU, of which there are a great many varients.

 

[ant9985]

btw if i sounded a bit rude in my first response i didnt mean to, i am greatful for all the help i can get with this.

thnx daniel.. but how do u did u come to those capacitor values and do i choose a transistor rated at 39VA just with 3 input wires and 2 output? i cant just put in a circuit without them knowing how i came to it. but thnx for the help.... why do you want to know my age? im 20

 

[ant9985]

i used the equation p = vrms x irms to get 39VA
and I x t = C x V to get 0.3F capacitor
did i use the wrong equations?

 

[Nigel Goodwin]

i used the equation p = vrms x irms to get 39VA
and I x t = C x V to get 0.3F capacitor
did i use the wrong equations?

Your whole premise is wrong!.

The regulator circuit takes a higher DC voltage, that may have a reasonable amount of ripple on it (so you don't require a massive capacitor), the regulator then outputs a lower stabilised voltage with a far lower degree of ripple.

For a simple example:

Imagine the peak output voltage on the capacitor is 40V, and under full load this drops to 30V in the troughs - so that's 10V ripple on the supply. As output you're only wanting 26V maximum, this gives 4V margin (30V-26V) from the output voltage to the lowest part of the ripple. A discrete voltage regulator circuit would be quite happy with that amount of leeway, and you would get a smooth, ripple free, output (presuming the regulator is designed OK).

In practice it wouldn't be a good idea to have 10V ripple on the incoming supply to the regulator, I used that value as an exaggerated example.

 

[danielsmusic]

smaller caps will reject smaller ripples

 

[danielsmusic]

thnx daniel.. but how do u did u come to those capacitor values and do i choose a transistor rated at 39VA just with 3 input wires and 2 output? i cant just put in a circuit without them knowing how i came to it. but thnx for the help.... why do you want to know my age? im 20

the caps just get smaller, as long as they get smaller it should work. not transistor, transformer. i have built this circuit and it works ok for me. it does vary from 30v to 0 and has no short circuit protection. the bigger the transistor and transformer the more wattage you will get. i used a power transistor but a medium transistor will do for what you want, i wanted 20amps at 30v. hope that helps.

why do you want to know my age? im 20

just curious. i was thinking gcse level.

 

[Roff]

smaller caps will reject smaller ripples

Nope.
Sometimes it is necessary to do a similar thing on a PC board with high speed logic or RF, because different-valued caps become inductive at different frequencies. This isn't necessary on a power supply preregulator filter.

 

[Roff]

thnx daniel.. but how do u did u come to those capacitor values and do i choose a transistor rated at 39VA just with 3 input wires and 2 output? i cant just put in a circuit without them knowing how i came to it. but thnx for the help.... why do you want to know my age? im 20

the caps just get smaller, as long as they get smaller it should work. not transistor, transformer. i have built this circuit and it works ok for me. it does vary from 30v to 0 and has no short circuit protection. the bigger the transistor and transformer the more wattage you will get. i used a power transistor but a medium transistor will do for what you want, i wanted 20amps at 30v. hope that helps.

why do you want to know my age? im 20

just curious. i was thinking gcse level.

You have no regulation with this circuit. All it provides is a way to vary the output voltage.

 

[audioguru]

The pot feeding a power emitter follower isn't a variable voltage regulator. It is a current limiter with a set max voltage without a load. As the load current increases, the transistor's base current also increases which causes the pot's slider voltage to drop and the output voltage too.

 

[danielsmusic]

yes you are right but this was the quikest thing i could think of at that time.

 

[JimB]

Ant9985

I am sending a PM to you with some pages scanned from a book:

"Solid State Design for the Radio Amateur"

The pages take you through a design process for simple power supplies, this will get you a lot nearer to your goal than your current path.

If your "teacher" thinks that your current approach is fine for a PSU with the specification you have shown here, he is VERY MISTAKEN.

JimB

Bummer!!
There is no obvious way to add an attachment!
Can you send me a PM with your e-mail address?

 

[ant9985]

ok thnx jimb and thnx all who replyed