In fact, the project is not mine (https://elinux.org/RPi_5V_PSU_construction#cite_note-6). I wanted to make a diy power supply for my raspberry. I am not an electronics expert. I realized it’s not a good project. I will desist. 
Thank you all for the answers.
Some of the Raspberry models are power hungry and will draw more than the 1A capacity of the 7805. You’d be better off getting a power supply that’s specified as suitable for your model.
I didn’t know that. Thanks.
Life is choices. That said, it is hard to beat some of the low end ready to buy choices now. I like to do as much as I can but the cost of getting started in doing my own circuit boards vs the cheap alternatives? I’d get an inferior product at a higher price.
Still, why not at least breadboard the project and see how it goes. That is the fun part. Then if you like it go the next step.
From that sentence one can conclude that Raspberry model needing ‘only’ 1A can be easily powered by this supply.
7805 without radiator with 1W dissipated in it is very hot - you won’t keep it in your fingers for long.
And here we have (12-5)*1=7W.
The good radiator is certainly needed.
Nowadays the cheapest source of 5V supply with 2A or more are cell phone chargers.
It is a good project. Forum members have just suggested improvements. There is an awful lot of experience and knowledge in this forum and most 
comments are very useful.
The 7805 will supply at least 1A @ 5V. R1 & R2 will make that 7805 adjustable, but only upwards, so you will be able to get at the output a minimum of 5V and a maximum of probably about 6V.
If you only want 5V, remove R1 & R2 and tie pin 2 of the regulator to Gnd.
C6, C5 & C3 are unnecessary… remove them.
The manufacturer suggests C1 should be 220nF, so change that.
The 7805 has thermal and output short-circuit protection at 1.5A, so you don’t need the fuse there.
Maybe place the fuse between C4 & C1 to protect your transformer and rectifier diodes. Don’t put the fuse before C4 as C4 will give it a hard time when power is applied to the circuit.
D4 is not needed, that can go also.
If the LED is not bright enough, decrease the value of R3, but don’t go below 470Ω.
D2 is debatable. I’ll probably be shot down by someone 
, but I have never seen a regulator chip go high (Vout =Vin) when failed, so you don’t really need D2.
D1 is a must. This protects the 7805 if the 12V Input is shorted, but it needs to be more robust… see Davidsrsb’s comment.
Finally, many cheap Chinese 5V phone chargers do not meet their published specifications.
The worst I’ve found could only supply 300mA before the output voltage started to drop, so if you decide to go this way, TEST your wall-wart/phone charger before connecting it to your Raspberry.
Cheers & good luck with your project!
EDIT: Use a TO220 package 7805 with a decent heatsink attached. See comment by @Piotr .
1 Like
By all means proceed with your project if you want a learning experience. However several years ago I noticed that 1. I was accumulating a lot of 12V power supplies for external disk drives, modems, and that sort of thing, and 2. Buck converters were only a buck or two. I used one here:
These provide up to 2A. No heat dissipation issue. Larger ones go up to 3A.
However if your circuit is part of a larger project involving say a RPi Pico, then it’ll be fine. No need to “boost” the voltage, those MCUs have tolerance in both directions.
Yes it can be problematic. Sometimes the ground pin current can cause the ground pin to rise above the output, and that defeats the junction isolation on the chip so it provides no output. I agree with the posting that if you want to tweak the voltage like that, use an LM317. That is a really nice chip.
Personally, to power a product that required a set voltage input, I’d question the use of using an easily adjustable (via a pot.) power supply of any type.
@retiredfeline
I must also confess to hoarding redundant mains power supplies (probably enough for three lifetimes), but I generally connect them to @BobZ 's recommended 317s. It is hard to go to the trouble of buying or building small buck regs. with another three lifetimes worth of 317s still out in the shed. 
1 Like
Hah, I have 9 lifetimes worth of junk^Wspares. 
But do you have over 250Kgs of Mains transformers?.. I’m thinking of advertising them as boat anchors!
Maybe we could trade: Rubbish of yours for rubbish of mine. 
Usually I ideally like to have a voltage rail that is 0% to 3% high, such as 5.12 V to power (obsolete) TTL logic for example. If you use a 7805 regulator, it might give you 4.82 volts so it leaves you without a lot of low end margin. Usually I will set up an LM317 with fixed resistors and it comes out pretty close to what I want. But if you use a potentiometer to adjust an LM317, it is probably better to set it up as a rheostat with one end connected to the wiper. And use it in conjunction with fixed resistors so that it can only give you a +/- 10% adjustment range (for example). This assuming that the LM317 is in a fixed application and you are not making an adjustable supply for use on your lab bench.
2 Likes
The Raspberry Pi will never work reliably from a 7805 regulator as it quite simply cannot supply the required current required.
Taken from the Raspberry pi website
A Raspberry Pi must be powered with a compatible power supply. All models run on 5.1V power, but Raspberry Pi 4 and 400 require 3A while other models can work fine on 2.5A. It’s not recommended to use a phone charger or USB cable to power a Raspberry Pi.
Although you can get away with slightly less than this it is not advised.
1 Like
The family also includes the Zero and the Pico though.
‘Pretty close’ is what prompted me to write…
LM317 is one of very good, never aging ICs (like 555 and 7805). But I had some problems with it.
Our first product (in 1988) was Eprom Programmer “Piccolo”. Since 2004 we don’t have it in our offer.
I found its picture here:
http://cyfronika.com.pl/micromade2/piccolo.html
Programmer was powered from PC keyboard socket (5V). Then I build step-up converter based on NE555 (those time in Poland you had limited access to modern ICs and even its datasheets) to get about 30V and used LM317 to have source switched between 5V, 12.5V, 21V and 25V.
When we started to produce a bit more, we quickly found that only one way to have those voltages enough precise was to measure each LM317 output voltage (nominally 1.25) and find the right resistor for it (we had THT resistor tapes with each resistor value written on tape).
In datasheet 1.25 is specified as 1.20 to 1.30 so ±4%. It happened to find not one LM317 with that voltage even out of specified range.
So in next revision I replaced LM317 with 7805. I didn’t used resistors to set the voltage but rather Zener diodes. There were one 7805 type we had to exclude from using in our programmer. When you wanted to switch VPP from 25V or 21V down to 5V (by shorting the reference pin to GND with npn transistor) they didn’t get their output down, but rather up to possible maximum. As input was 30V it was about 28V. That state was stable - IC didn’t noticed that has too high voltage at output. But it was only one manufacturer 7805 (don’t remember which) that had such a property.
OMG! I was wondering what to do with my 20kg of mains transformers so that they do not go to waste (it would be a pity).
When I was young you couldn’t buy what you want but you have to buy what was available. Once a week, I visited a special electronics store with liquidated redundant elements from various factories and bought what I thought would be useful one day. I did not know then that future power supplies would be complete without a power transformer.
Those time in all designs I used only elements I already have to avoid waiting years till I will get them.
You win. I’m not into heavy metal. 
What about rasperry cake with vanilla ice cream? 