Building and understanding a linear power supply

[Atom]

Hello to everyone, so here i am , and i need a bench power supply , a lot of people would say buy a dcdc converter and you are done, yes that's true but they have always some kind of noise, i even built myself the pcbs for some Simple Switcher ...that's how ti calls them...like LM2575, LM2596 LM2677 LM2679 the chinese copy XL4005 (it works really well).
But enough of those, now i need a linear bench power supply, so designing it from scratch ins't an option, also because right now i am studing the opamps at school and i don't fully understand them... so i started by an already made desing, that's is fairly simple.. but i still don't understand some things. you can find the schematics under there, so this is the guy that made it



here it is another guy that remade the schematics into an eagle file

https://github.com/SausageCode/Linear-power-supply-with-CV-CC

the circuit is using the transistor as a ''variable resistor'' by feeding a certain amount of current in it's base (the costant current source, snd the op amps).

now here are my questions....

1 i undestood that the guy is using the v out as the ground for the 7805 and the 7905 but why is he doing it?

2 the shunt resistor after the transistor are used to sense current .... they generate a 'drop' of voltage across them that is proportional to the current flowing thus we can know how much current it's flowing... shouldn't we measure the voltage across the power resistors?

3 i was hoping to understand the current limiting feedback so i could understand where those 4*1kohm resistor are attached because right now i have no clue, and the little arrow under them?

4 should i add a switching preregulator (i was thinking about lm2677) to diminuite the power transistor heat disspipation, what do you think is the preregulator going inject high noise into my output voltage?
(the preregulator is going to have a voltage always 3 volt higher than vout)or should i go full linear?

5 could i use an smps , right now i have a meanwell sp320-11 (paid 10 euros   ) the questionis the same as before.

Fell free to correct me because i am here to learn , Thanks in advice

 

[Atom]

Sorry this is the right video:

[David Hess]

1 i undestood that the guy is using the v out as the ground for the 7805 and the 7905 but why is he doing it?

Since the supply voltages for the control circuits follow the output voltage, the control circuits never see any high voltages and no level shifting or high voltage gain is required.  This is how a "floating" voltage regulator is made.

The complication when doing this is that a floating auxiliary supply is required which in this case means another transformer winding.

2 the shunt resistor after the transistor are used to sense current .... they generate a 'drop' of voltage across them that is proportional to the current flowing thus we can know how much current it's flowing... shouldn't we measure the voltage across the power resistors?

The voltage across the power resistors *is* being measured.  Specifically, the non-inverting input of the current regulation operational amplifier is connected to Vout plus a variable offset derived from the +5 volt supply.  That offset becomes the voltage across the current shunts in constant current mode.

3 i was hoping to understand the current limiting feedback so i could understand where those 4*1kohm resistor are attached because right now i have no clue, and the little arrow under them?

Each 1k resistor goes to one of the emitters so the average voltage across all of the current shunt resistors is being measured.

4 should i add a switching preregulator (i was thinking about lm2677) to diminish the power transistor heat dissipation, what do you think is the preregulator going inject high noise into my output voltage?
(the preregulator is going to have a voltage always 3 volt higher than vout)or should i go full linear?

That is up to you but noise is a definite concern.  The design as shown uses 4 transistors in parallel to handle the high power dissipation.

An alternative which would cut the maximum power dissipation in half is to use dual high power secondaries and switch them for a low voltage output range and a high voltage output range.  This could also allow doubling the output current on the low voltage range.

5 could i use an smps , right now i have a meanwell sp320-11 (paid 10 euros   ) the questionis the same as before.

This would be difficult to impossible.  Fixed output switching power supplies do not generally work over a wide output voltage range.  It could replace the big power transformer, rectifier, and input filter capacitor though saving weight.

[Atom]

for the question number 5 i wanted to replace the main transformer + diode bridge + capacitors with the smps, but then i should use another transformer for generating the +-5V required for the op amps and the voltage references.

ok then last question would you (or everyone else ) reccomend to continue to make the pcb for this psu or should i go with something else?

in a recent thread :https://www.eevblog.com/forum/beginners/0-30v-2ma-3a-dc-ps-kit-messed-up!/25/

at page 2 was mentioned the  Velleman K7200, is it a reliable psu? should i go with this one and make my own custom pcb?

do you reccomend some other power supply desings?

[David Hess]

for the question number 5 i wanted to replace the main transformer + diode bridge + capacitors with the smps, but then i should use another transformer for generating the +-5V required for the op amps and the voltage references.

Exactly.

ok then last question would you (or everyone else ) recommend to continue to make the pcb for this psu or should i go with something else?

This design is better than most of the designs I have seen on the forums here.

It is not really complicated enough to require a printed circuit board unless you are planning on selling the boards or going into production.

in a recent thread :https://www.eevblog.com/forum/beginners/0-30v-2ma-3a-dc-ps-kit-messed-up!/25/

at page 2 was mentioned the  Velleman K7200, is it a reliable psu? should i go with this one and make my own custom pcb?

The Velleman design looks nice also.  It uses a voltage gain stage after the error amplifiers so the control circuits can operate at a low voltage referenced to ground.

do you recommend some other power supply designs?

I would want a slightly different feature set but offhand I do have any better examples to point out unless you want to consider ancient datasheet or application note designs.

[Atom]

i don't know how to quote an answer, i'm still new to the forum  but i really appreciated your help David and for the pcb...i mean i enjoy doing them .. probably when i will finish it i will post it on here so evereyone could look through it and maybe give me some feedback..and btw it's all practice, i will never lern how to desing a good pcb if i don't even start

Thanks again!