All about Linear Power Supplies

[queennikki1972]

I have tons of questions about linear power supplies and the people over at Stack Exchange are rude as #$%^ to newbie's.

I have built a small linear power supply with a LM7812 and LM7805 and a 12VCT Transformer with Smoothing Capacitor. Diagram was accurate and works well for small bench circuits.

I am looking at building a 12v 5a constant power supply next using the LM338T. I have a diagram I think will work, but I have what I thought was a simple question.  Is a 6 amp bridge rectifier sufficient for a 5 amp load? The rectifier in question I already have and is rated 6amp at 800volts.

The other power supply is a variable 0-30vdc using the LM317T or the LM338T. First off, which would be a better choice for 0-30vdc variable at 1 amp max?

I have circuit links to both if needed and I apologize if I have offended anyone or asked a dumb question.

Nikki

[digsys]

Welcome to forums :-) They can get narky and rude at times, but this one is pretty ok. Just ignore a**wipes
For a 5A op at 12VDC, your filter cap is 1,000uF + 1,000uF per Amp = ~ 6,000uF, so app 8-10,000uF is the closest ie 2x 4,700uF
At that your inrush / charge current is ~ 2X+ the max op current. Worst case is if you switch on at full load or short.
Bridges are cheap, so use a MIN of 10A, 20A ones are not a lot dearer, and it gives you headroom if you want to go up a bit.
For the variable 1A PS, since you are dropping up to the entire range at min OP, you would either use a BIG heatsink with cooling, or go for the LM338T.
It still needs a heatsink though. Again, it gives you headroom.
Welcome to the forum, you should be safe here :-)

[queennikki1972]

Wow! Thats all I wanted.. A simple direct answer. Thank you. And Yes, I have a HUGE vented aluminum heatsink from a old CPU, cut in half and tapped out with a M3 screw for the variable power supply.

[ArthurDent]

You might get more responses if you combine your 2 similar threads. One way to get more ideas on the LM338 would be to check the application note on that regulator at the link below. The makers have a lot of ideas on how to use their chips.

Also there have been a number of threads on building 0-30VDC supplies. Many are probably more complicated than you need, maybe not. Do a search using the search box in the upper right on the main page, you'll get a lot of references and a lot of ideas you may be able to use.

http://www.ti.com/lit/ds/snvs771c/snvs771c.pdf

[MarkF]

Welcome.
A couple circuits you might want to study.

   

   

[digsys]

Just note: IF you are not using a fan, Heatsinks work a LOT better if the fins are vertical. That is, hot air rises so natural convection is upwards !
Don't make the mistake of mounting it horizontal or even upside down etc

[queennikki1972]

I did not intentionally make two separate threads, other than the project images of the one i had built first. Im sorry

[MarkF]

At a minimum, use a heatsink similar to a RA-T2X-64E with a fan.

(Check that.  Your big aluminum one looks good but you will probably need a fan.)

   

[ArthurDent]

"I did not intentionally make two separate threads, other than the project images of the one i had built first. Im sorry"

Not a problem, I just thought you'd get more replies if they were combined. You can also check

https://www.google.com/search?q=0-30+volt+power+supply+schematic&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjWx8CB0OzhAhVIw1kKHeAbC1sQ_AUIDigB&biw=1280&bih=611

and see more schematics than you'd probable want to. There are posters on this site who can give you good ideas as you have already seen.  One note: if you don't need the supply to go to zero volts, the supply is much simpler as shown in the schematics above. If you want it to go to zero, you need to add a negative voltage and a few extra parts.

[queennikki1972]

Thank you

[queennikki1972]

At a minimum, use a heatsink similar to a RA-T2X-64E with a fan.

(Check that.  Your big aluminum one looks good but you will probably need a fan.)

   

By fan, do you mean directly on the regulator or a case fan (exhaust fan in the rear of the case)

[MarkF]

Anything to circulate air across the heatsink.
Remember, the largest power it will need to dissipate is at low voltage and high current output.  The LM317 or bypass transistor has to drop the largest voltage at that point and still provide a high current.   Power(P) = voltage(V) * current(I).

[MarkF]

Food for thought...
Here is a small PCB I did for a LM317 adjustable supply (no current limiting) based on the datasheet.
Input is either a wall wart or a transformer.

   

[digsys]

It all about thermodynamics ! The more "efficiently" you set it up, the less air you need across the heatsink.
Ideally - either mount the fan to the Top of the HS and suck air UP, out of the enclosure - Hot air rises. Or mount to the bottom and blow up, pulling in cold air
from the outside. You can just "blow into" the HS, or just suck "out of" the enclosure - it all works, but efficiency will vary.
As long as the "hot air" is expelled, pretty much anything works. At your currents / voltages, it's not a big deal.
It gets "serious" when the power levels go up dramatically -
This is an E-Load I make for discharging EV battery packs - https://pbase.com/digsys/image/158741423
It can "push" 2,000W at peak. I have high-flows pushing and pulling, with added turbulence. If we need to run for hours, I'll jam 2 fans each side !!

[queennikki1972]

Wow! both are impressive. I have not learned to draw or etch boards yet. I started with a few simple circuits like police lights and piezo alarms and then dove right into power supplies. I am still learning the math but its difficult to search for the formula if you dont know one exist yet 

I appreciate all the help.

[MarkF]

I have been using Diptrace 3.3 Freeware version for my schematic and PCB designs.  It's free and easy to learn. 
It also generates 3D images of your PCB to check over before getting it made (Previous post as examples).

Also, I have been using Elecrow to get PCBs made.  You can get 10 PCBs for $4.90 
There are a lot here that use PCBWay for about the same price.

[queennikki1972]

Any opinions on using PLA or ABS 3D printed enclosures for linear power supplies such as a lm317T and 24vac transformer? PLA melting point is about 190c. ABS is around 240c. i have used these plastics for face plates only in the past.

[MarkF]

I would think either would be fine.  The max junction temperature of the LM317 is 150^oC.  You will have bigger problems than the case melting if it gets that hot.

For your next supply project, one thing I noticed on your first power supply is that you didn't bring out earth ground.  The 3rd prong on the power cord straight to the front panel to a green banana jack.  And tied to the case if it is metal.  Your power outputs will be floating and in some cases you might want to tie the common to ground.

Here is an example (although HP used a black jack):
   

[ArthurDent]

There is another power supply design thread on this site that goes into a lot of detail on all aspects of power supply design, including 3D printing an enclosure, in about 15 pages. It would be worth at least scanning the posts to see if there is something that you might be able to use in your design.

https://www.eevblog.com/forum/projects/linear-lab-power-supply/

https://www.eevblog.com/forum/projects/linear-lab-power-supply/?action=dlattach;attach=578924;image

[mariush]

If you want to keep it simple, you should look into other linear regulators like *1084 for example, or *1085 :

See AZ1084, AP1084, LT1084, LM1084 : https://www.digikey.com/short/pzcfr5

5A maximum current output, input voltage up to 30v (but depending on who makes them, some are maximum 12v or maximum 15-18v)

The *1085 regulators are 3A parts: https://www.digikey.com/short/pzcf2d

You can use two in parallel with a small resistance to balance current... not the best way to do things but nevertheless it's possible.

See page 15 in the Linear datasheet: https://www.analog.com/media/en/technical-documentation/data-sheets/108345fh.pdf
 



^ just replace the "two feet of awg18 wire" with a couple 0.01 ohm resistors and connect the output between the two resistors.

Also keep in mind that most regulators have a maximum power dissipation limit, like let's say 35w .... so for example you can't have 12v IN from the transformer and output 3.3v at 5A ... because the amount of energy dissipated would be (12v - 3.3v) x 5A = 43.5 watts.

You could work around that by using two 3A regulators in parallel but it would make more sense to use a transformer that has two secondary windings... for example, use a transformer that has 2 x 6.3v secondary windings.... and either add a switch on the front panel of your power supply or use a mechanical relay and a voltage comparator or something to switch between 1 winding or 2 windings.
For example, if  the output voltage is set to less than 5v DC, then switch to using only one winding, so that in worst case scenario you're gonna dissipate around  (7v - 1.25v ) x 5A = ~ 28 watts when output voltage is set to your minimum of 1.25v (the reference voltage of the linear regulator)

[queennikki1972]

I grounded to the case with the green wire and a lug under the circuit board

[queennikki1972]

Welcome.
A couple circuits you might want to study.

   

   

I found the link to this circuit and reading a little. So although the LM317 is limited to 1.5 amps, this circuit allows for 5 amp constant current and variable as well? What size transformer and smoothing would be best suited to cover this also? If I had to guess on transformer I would say 12v-0-12v 5 amp ?

I have to admit although I could probably build this circuit , it will take me a minute to comprehend its entirety.

[Mechatrommer]

I found the link to this circuit and reading a little. So although the LM317 is limited to 1.5 amps, this circuit allows for 5 amp constant current and variable as well?

yes. extra current will go through TIP3055 due to 22R and 4R77 voltage divider mechanism.

What size transformer and smoothing would be best suited to cover this also? If I had to guess on transformer I would say 12v-0-12v 5 amp ?

depends on what you want to do. it could be, it could be not. if you want PSU that goes to 30V on the output, then no.

I have to admit although I could probably build this circuit , it will take me a minute to comprehend its entirety.

Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it.

another word... if you can read it in a minute, means you can build and debug it in 2 minutes... it could also means another i have several psu projects that went under the shelve due to sub par performance based on eevblog community standard, hence they have to go behind the project/task queue, get used to it! if you think eevblog (electronics) is a friendly place fine, get deeper into the rabbit hole, there are wonderful things (not intending to demotivation) but... not with sitting on a "lazy chair" cheers

[queennikki1972]

Yes I would prefer the 30v output on the circuit if I built it.. As I understand it I would not want to go higher than the max input of 40 volts for the regulator (maybe that does not apply in this circuit I don't know) and that 12-0-12 or 24 volts is to low of an input for 30v output but a 24-0-24 or 48v transformer is to high.

[MarkF]

I'm not sure which circuit you are looking at.  But...

• I don't think 30V is a big deal.  You probably won't ever need much over 24V.
• Having adjustable current limiting is a bigger deal.
• Hammond has many voltages to pick from if you need to buy a transformer (See attached).
• Looking at the 1st circuit, I'd try a LM358 opamp so you don't need a negative supply.  You would not be able to get completely to zero voltage.  Probably 1.5V as minimum.

I have not built either circuit.  I don't know how they perform.  I have 4 of the little HP supplies that I posted the picture of.