# Unregulated power supply design

Discussion in 'General Electronics Chat' started by chimera, Feb 14, 2012.

1. ### chimera Thread Starter Member

Oct 21, 2010
122
2
Hello! I been meaning to design a power supply for my work bench. Right now I have been using a modified computer supply and its just too bulky..eh I wont bore any one with that...

I been doing some extensive reading, understanding how the unregulated power supply works and what are its inherent draw backs.

I have come across so many different formulas for determining the capacitor, the bleed resistor, output regulation..etc. Despite that, I have not figured which formula(s) to use.

I dont have a schematic prepared, so for now here is what I have going on paper:

Transformer(13.5 AC)-->rectify_Full Wave-->filter_cap + bleed resistor--->19.5V (no load). For the sake of argument,

Vripple is 5%, Output Current(max)=3A, RL=6.5Ω, Output Voltage=19.5V (no load)

Like I said, I have seen tonnes of reading material, but I am not completely sure of the formulas..on the account that every one uses different one..

Can someone please provide me with the basic forumlae to find the filter cap, C. This will help me do an analysis of the output voltage.

Thanks

2. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
Filter cap amount is dictated by voltage ripple: I = C dV/dt

NOTE: RMS transformer seconadary current is about 1.8X DC load current for FWB circuit. You need at least a 5A transformer for a 3A power supply design.

NOTE 2: if transformer is 13.5VAC (no load) your unregulated DC input level will probably drop to maybe 14V under full load.

3. ### chimera Thread Starter Member

Oct 21, 2010
122
2
Okay. So the max DC Current is going to be, with the transformer I have on hand:
DC max = 3/1.8 = 1.67A

1- How did you come up with the 14V output at full 1.67 A ?

2- For instance, Vripple is 3%. According to the formula, I=C dv/dt, this means that,

1.67=C ΔV/ΔT, where ΔV=(3/100)*14 and ΔT=1/(2*60)

C=13917 μF

Is the calculation correct based on the 14V output?

4. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
Estimating the effects of loading and voltage peak "flattening" that occurs due to the high peak currents drawn by a full wave bridge/capacitor. It's ballpark guess at minimum voltage, could be a little higher.

Calculation is correct.

5. ### chimera Thread Starter Member

Oct 21, 2010
122
2
Thanks for the replys. However, i cant really seem to understand the basics of building a power supply. Not because I cant find enough resources to read up on, but because every single article which I have read, has always used different equations and the explanation is not quite adequate.

Can you or someone, please point to some good links or papers even that explain the fundamentals behind a UNREGULATED POWER SUPPLY.

I want to build one up and see how the output voltage responds to varying load (let it be resistive or inductive) and then once I have figured that out, move over to regulating it (LDO regs or switching)

6. ### K7GUH Member

Jan 28, 2011
191
23
Can't tell you why no one has suggested this before. Build a supply out of what you have on hand. Use a fuse in the primary lead, and either a half- wave or full wave rectifier. Use whatever capacitor(s) you have handy. A handful of wire-wound or sand resistors would be useful to create a varying load. Use a reliable volt-ohm-milliammeter (doesn't have to be digital), and keep notes on what you find. You don't need the Encyclopedia Galactica explanation of what to do. Be careful of any voltage, high or low, and follow the one-hand rule faithfully. If you think equations will be helpful, write down any that you discover during your experiment(s). Then let us know how you did.

7. ### chimera Thread Starter Member

Oct 21, 2010
122
2
lol..or a close approxiamtion would be simulation using a software but seriously.. a few helpful resources to study on would be great!!!

8. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,300
335
chimera likes this.
9. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
Here you go. I have published it already numerous times, so I think there are no typos (?)

• ###### APPA.PDF
File size:
40.6 KB
Views:
125
Last edited: Feb 16, 2012
chimera likes this.
10. ### pilko Senior Member

Dec 8, 2008
213
20
@ bountyhunter,
Good info.
What is the appendix A part of?

Regards

pilko

chimera likes this.
11. ### MrChips Moderator

Oct 2, 2009
12,624
3,451
I was supposed to help Bill with the power supply equations but could not find the time right now. All of this is available on the net.

The ripple voltage can be approximated to

Vripple = I/(2fC)

As a quick example, if I = 1A, f = 60Hz, C = 1000uF,
Vripple = 8V

It is easy to take it from there.

http://en.wikipedia.org/wiki/Ripple_(electrical)

http://www.zen22142.zen.co.uk/Design/dcpsu.htm

The point is not to reduce the ripple voltage to a very small value. If you tried to do so, the value of C would be very large. This becomes a short-circuit to AC and the peak current through the rectifier diodes will sky-rocket. It is instructive to calculate the peak current through the diodes. This is not easy to calculate and would be made simpler if you use some approximations.

Once you reduce the ripple voltage to a reasonable level, use a voltage regulator to eliminate the ripple.

chimera likes this.
12. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
It's my own work. It's part of a larger work on switching and linear regulator design.

chimera likes this.
13. ### chimera Thread Starter Member

Oct 21, 2010
122
2
both of you are awesome!!!! sorry i wasant able to get back to this thread.. i am currently looking for employment as a design engineer and wow..let me tell one thing.. the job market is hard!!!

I looked into the pdf bounty hunter provided and its very consistent. i like that. Majority of the info I had/have gathered over time regarding power supplies is not to the point.. some of them ramble on about specific details etc..

I have ordered some parts based on the calculations I made to designing an regulated power supply. I am supposed to get them this monday. I'll run some real simulations using an 0-scope and post back with much more detail ... possibly a pdf with screen captures and schematics and all of that good stuff.

stay tuned if you would like