Variable Power Supply

[Mati256]

Hello. I'm looking to build a variable power supply to replace the "wall wart" I'm currently using. I have build one before, but my notes are lacking some specifications for some reason.

I have googled around and I can't find a good schematic to build one, I don't know what I'm doing wrong.
Finding components where I live is no easy task, they usually stock some common components, nothing too fancy.
For example, I wanted to build this Supply from Make Magazine: http://makezine.com/projects/0-24-Volt-2-Amp-Bench-Top-Power-Supply/
But I can't find a BR805D Bridge Rectifier in the electronics store. 
I want to know with what individual diodes I can replace this bridge?
What about that mini volt meter? I have never used one, would any voltmeter be useful in this situation?

Also, if you happen to know a better, simpler VPS, please point me to the schematics.
What I'm actually looking for is a 12V or 20V 1A VPS with a voltmeter display.

Thanks

[liquibyte]

BR805D = 2A 50V bridge rectifier.  Over thinking leads to madness.  1N4001 - 1N4007 gets you there.  Mini Voltmeter on Ebay.

Sometimes you just have to experiment with component values based on what you have on hand or is readily available.  If you toggle your country flag, you'll get much more useful information from the various folks that are more local to you.

[Mati256]

Thanks liquibyte I was guessing I could use a 1n4007, but looks like the BR805D Max. Average Rectified Current is 2A and the 1n400X is 1A? Is that ok?

[denelec]

I built a similar power supply more than 25 years ago.  It still works.
You can use any 2A, 50V or more rated diodes or bridge.
Use a big heatsink on the LM317. And make sure your case has ventilation holes.
And add a fuse for safety.

I think any voltmeter should do.  I made mine with a recycled needle VU meter.  Same for the ammeter.

[liquibyte]

Thanks liquibyte I was guessing I could use a 1n4007, but looks like the BR805D Max. Average Rectified Current is 2A and the 1n400X is 1A? Is that ok?

Since the max of the 317 is 1.5A I'd say on average you'd probably never want to draw more than 1A continuous as a small supply anyway.  Others with more experience than I will probably give you much more detailed info but I've used the 1N400x diodes many times in cases like this without issue.  If you really need a 2A diode, try using the parametric search on any of the electronics parts supply houses like Digikey or Mouser.  There are almost too many choices with discrete components and the best way to learn is to learn to read the datasheets.  It's daunting for a newbie but eventually you'll get to the point where you'll understand things enough to make informed choices.  Of course, you could always use a 1N540x series rated at 3A.

[homebrew]

http://makezine.com/projects/0-24-Volt-2-Amp-Bench-Top-Power-Supply/

That sir, is a VERY CRAPPY build instruction!
It is dangerous for various reasons:

(1) No fusing whatsoever, neither on the primary side, nor on the secondary side. I have my doubts that the transformer shown even has a temperature fuse in it nor that it would be inherently short circuit proof.

(2) Ridiculous case! Wood is definitively not recommended to build enclosures that house components operating on the mains...

(3) Standard LM317 are only rated for 1.5A not 2.0 (http://www.ti.com/lit/ds/symlink/lm117.pdf)

It has severe functional limitations:

(4) Minimum voltage would be 1.25V

(5) The pisspoor heatsink shown would run red hot on some 100mA! (And set the wood on fire  )

There are however tons of good DIY projects out there.
For example: http://hpm-elektronik.de/nt30-4-netzteil.htm

Or just use the google image search to find tons of good schematics. (google for variable voltage current power supply schematic)

Play safe!

[macboy]

100% agree with homebrew. That  design, and especially the construction used by the author, are complete garbage. The heatsink is a bad joke, the lack of fusing is dangerous, and the over-rating of the LM317 shows that the author knows less than he purports to know.

[liquibyte]

Yeah, sorry I didn't even look at the article but did a bit of searching based on OP's first post.  There are good and bad 317 circuits out there and this is one of the poorly thought out ones.  To be honest, there's way too many 317 versions out there because they're so easy to make and very few of them center around safety concerns or even good explanations as to why what is being done is why the way it is.  Not only is fusing important but earthing the chassis and transformer the right way are as well.  The ubiquitous LM317 is a great chip in my opinion though.  I keep a stock of them on hand for breadboarding stuff up but to be honest, they don't make great variable supplies for the most part but that doesn't mean that it's not possible.  No one here has ever really put forth a great hobby design to be honest even though bad ones flow through all the time.  Perhaps we need to put together a low part count, solid performing first build for the new folks here and point them to that build when these things get proposed.  I've yet to see anything like that offered even though it gets mentioned rather often.

As far as this build goes, isn't that fuse for the variable side a little underrated at 1A for a 5A transformer?

[homebrew]

As far as this build goes, isn't that fuse for the variable side a little underrated at 1A for a 5A transformer?

Why should that be the case? 1A at 230V gives approx. 230W. That should be plenty of power for the main transformer's output of 28V @ 5A ...

[homebrew]

No one here has ever really put forth a great hobby design to be honest even though bad ones flow through all the time.  Perhaps we need to put together a low part count, solid performing first build for the new folks here and point them to that build when these things get proposed. 

Great idea! I would immediately be willing to contribute. But what would be the specs?
Something like 0-15V @ 1.5A linear regulated. Maybe some additional fixed voltages, too like +/- 15V and 5V for analog stuff and logic ...

[liquibyte]

No one here has ever really put forth a great hobby design to be honest even though bad ones flow through all the time.  Perhaps we need to put together a low part count, solid performing first build for the new folks here and point them to that build when these things get proposed. 

Great idea! I would immediately be willing to contribute. But what would be the specs?
Something like 0-15V @ 1.5A linear regulated. Maybe some additional fixed voltages, too like +/- 15V and 5V for analog stuff and logic ...

I'd say that a first build is usually a single variable supply, I know mine were.  Then I started doing things like 317/337 stuff and then graduated into learning ways to increase the amps.  One of the issues that's always run into with the 3 terminal regulators though is heat, one of the often underestimated side effects of doing these designs.  If something that has multiple outputs is proposed, I'm sure it would just be passed over for something simpler for a first build.

I'd like to see a solid 0-20V @ 0-1 or 2A variable voltage and current design with short circuit protection that doesn't suffer from the issues I've been running into with the design I've been trying to fix.  What I often see is that people start out looking at 12,15, or 18 volt at 1 amp and then as research progresses they inevitably graduate to higher voltages anywhere from 20 to 30 or even 50 volts at 2, 3, or 5 amps or more.  More often than not, this is in the belief that bigger is always better and that room to spare is a good thing.

Eventually you end up knowing what you're after and start lowering those expectations back down into something reasonably realistic for working with, say, op amps and the like.  20V @ 2A is usually plenty of room for anyone just starting out but a design that did 2X this with a tracking mode plus a 5V rail might be a good thing if it was reasonably simple to build and cheap enough to put together.

[Vito_R]

You might want to consider going with a LM350/K which can do up to 3 amps.  I built one of those over 25 years ago too still using it today.  You might need a bigger transformer and higher amperage bridge rectifier of course but for the small difference in cost you benefit from the flexibility of having 3 times more current.

[Vito_R]

hmmm im building an upgrade : 0-24v, 0-5A variable current limit ... the PCB is on the way from SEEEDstudio
the 5A is lm317 piggy backed with power PNP. one of those generic 5A types
current limiting is using PMOS, but ... im not sure if the PMOS will work, its a conversion from a simulation
this is the pic of the heatsink to be tested on (10cmx15cmx3cm) ... ribbed fins. this is actually the most important component i think 
the supply comes from a generic 24V 200w PWM power blocks adjusted upwards to about 28v
the worse part of any power supply is sinking heat ... its like vomit ! arrgh

Nice design, are you adding a fan or just sticking with convection cooling?

[liquibyte]

oh def a fan ... @ full draw = more than 125w of very angry electrons
and it will be inside a plastic "bin" of sorts lol
(upgrade from this "bin" of volts not bad for 1.5years of use , lm317 7805 n a 7812 http://3roomlab.blogspot.sg/2013/04/036-most-unlikely-psu-mod-lol.html)

Man, that's some ghetto right there.  Hey it works though right?

[Mati256]

Thanks for all the answers, that schematics proposed is way more complicated than what I'm looking for.
It's great to see my simple question sparked this design idea, I might wait and see what comes out of it and build that.
But, if you happen to come across some neat design, please let me know.

[Mati256]

No, I'm not giving up. But I'm still looking for an approachable Power Supply schematic, the one posted by homebrew is too overwhelming for me. I'm looking for something like the MAKE Powe Supply but safer. 

[liquibyte]

Thanks for all the answers, that schematics proposed is way more complicated than what I'm looking for.
It's great to see my simple question sparked this design idea, I might wait and see what comes out of it and build that.
But, if you happen to come across some neat design, please let me know.

Here, try this one.  It's got a few issues that we're still working on but it works rather well for the most part and should be a good first power supply.  If you really don't want to take the time to have the boards made or don't quite have the confidence to try this out on a perfboard or something similar, I can send you a board I had made.  I keep trying to get someone with the right test equipment to build and test this and help us figure out where the failure on the transient is coming from but I've had no takers so far so I'm willing to give the last two boards I have left away to anyone that wants them.

[Vito_R]

An LM317 or LM350/K design would be a lot simpler then that for him. fuse, transfo, bridge rectifier, a few capacitors, LM 350 or K on a large heatsink, resistor, trim pot to control the voltage and you got yourself a power supply.  I built one when I was a kid and even etched my own very simple pc board at the time.  Was my first electronics project and it's still one of my regularly used power supplies today. The LM350K has built in short circuit protection/current limiting and reliable as hell, never failed on me.

[liquibyte]

what kind of failure on transient? can you illustrate via a ... ermmm graph?

There's a transient that occurs on power on.  Here's a screenshot from the scope of a guy that built one.

An LM317 or LM350/K design would be a lot simpler then that for him. fuse, transfo, bridge rectifier, a few capacitors, LM 350 or K on a large heatsink, resistor, trim pot to control the voltage and you got yourself a power supply.  I built one when I was a kid and even etched my own very simple pcb board at the time.  Was my first electronics project and it's still one of my regularly used power supplies today. The LM350K has built in short circuit protection/current limiting and reliable as hell, never failed on me.

The actual circuit isn't all that bad to build but is certainly more complicated than your average linear voltage reg circuit.  Plus, I offered to send a board, just add components.  The most expensive part is the transformer at about $30 or so.  Next up would be the two 10-turn pots I used that were $12 each and then the three trimmers at $4.62 each and then the three TLE2141's at about $2 each.  The output cap was $2.64 but a different cap could be used if wanted.  The 2N3055's were $2.34 and the design needs two.  The rest of the stuff is just bog standard things most people have laying around.

[kxenos]

Have you tried to delay the power up of the op-amps for 40-50mSec?

[liquibyte]

Have you tried to delay the power up of the op-amps for 40-50mSec?

Yep, I even worked out this PFET circuit for it.  I haven't had the chance to try it yet though, too much going on right now.

[homebrew]

Here, try this one.  It's got a few issues that we're still working on but it works rather well for the most part and should be a good first power supply.  If you really don't want to take the time to have the boards made or don't quite have the confidence to try this out on a perfboard or something similar, I can send you a board I had made.  I keep trying to get someone with the right test equipment to build and test this and help us figure out where the failure on the transient is coming from but I've had no takers so far so I'm willing to give the last two boards I have left away to anyone that wants them.

Nice supply!

But I can't wrap my mind around the supply of the current-limiting opamp. Why is the zener in series with the opamp supply? There will be a constant 10V drop over the zener but the supply of the opamp would not be regulated at all?

But apart from that, I also believe that the inrush-current of the main filter-cap is the cause of the glitch. The cap might have a quite low ESR and therefore the input voltage will drop a lot until the cap is charged. Hence the series pass transistors are saturated more under this condition. But a few ms later the cap is fully charged and the regulation circuit is not fast enough to compensate ...

My suggeststion: Insert some delay to turn on the output. Maybe as simple as an RC timing element and one FETto clamp the set voltage to zero or something the like ...

[richard.cs]

http://makezine.com/projects/0-24-Volt-2-Amp-Bench-Top-Power-Supply/

That sir, is a VERY CRAPPY build instruction!
...
(2) Ridiculous case! Wood is definitively not recommended to build enclosures that house components operating on the mains...

While I agree that it's a pretty crappy project for most of the reasons given I would argue against your point about the wooden case. A well made wooden case is less of a fire hazard than most plastic boxes, is mechanically stronger for something like a power supply with heavy transformers and is insulating enough to give good secondary protection against electric shock. That particular case isn't great but good cases can be made from wood. The most important thing to be careful of with insulating cases is to avoid conductive screws, transformer mounting bolts, etc passing through the enclosure unless they are prevented from becoming live by some other aspect of the design (being insulated on the inside or being earthed).

There are however tons of good DIY projects out there.
For example: http://hpm-elektronik.de/nt30-4-netzteil.htm

Agreed, way better. Though maybe a little complex for a beginner.

[liquibyte]

Here, try this one.  It's got a few issues that we're still working on but it works rather well for the most part and should be a good first power supply.  If you really don't want to take the time to have the boards made or don't quite have the confidence to try this out on a perfboard or something similar, I can send you a board I had made.  I keep trying to get someone with the right test equipment to build and test this and help us figure out where the failure on the transient is coming from but I've had no takers so far so I'm willing to give the last two boards I have left away to anyone that wants them.

Nice supply!

But I can't wrap my mind around the supply of the current-limiting opamp. Why is the zener in series with the opamp supply? There will be a constant 10V drop over the zener but the supply of the opamp would not be regulated at all?

But apart from that, I also believe that the inrush-current of the main filter-cap is the cause of the glitch. The cap might have a quite low ESR and therefore the input voltage will drop a lot until the cap is charged. Hence the series pass transistors are saturated more under this condition. But a few ms later the cap is fully charged and the regulation circuit is not fast enough to compensate ...

My suggeststion: Insert some delay to turn on the output. Maybe as simple as an RC timing element and one FETto clamp the set voltage to zero or something the like ...

The zener for the current control op amp is due to the negative rail, otherwise the total voltage would exceed the rated 44V.  I didn't design this one, I'm just trying to fix some of its issues.  I've always been concerned about being so close to the max of the op amps I'm using, especially when it comes to the transient.  I get 43.7V after rectification and filtering and the max input to the TLE2141 is 44V.  Hasn't seemed to damaged anything yet during testing but I always thought that putting all three op amps behind the zener would have been a better idea given the tight output.

I agree that the inrush is caused by the filter cap.  I get the same results as far as the transient goes whether I'm scoping the output of the PS itself or the output of the filter caps.  ESR on my caps according to the datasheet is 44 max at 100Hz, not that I entirely understand that number.

The PFET circuit was my attempt at a delay and regardless if it is a good idea or not it does simulate nicely.  I had initially settled on a soft start on the primaries but given the discussions I've had, I think it would really be better to eliminate the issue than hack around it.  3roomlab has given me a couple of ideas that I want to try out and I'll post if I get good results or not.  Simulations are nice but not always related to reality so some further testing is in order.

[homebrew]

While I agree that it's a pretty crappy project for most of the reasons given I would argue against your point about the wooden case. A well made wooden case is less of a fire hazard than most plastic boxes, is mechanically stronger for something like a power supply with heavy transformers and is insulating enough to give good secondary protection against electric shock. That particular case isn't great but good cases can be made from wood. The most important thing to be careful of with insulating cases is to avoid conductive screws, transformer mounting bolts, etc passing through the enclosure unless they are prevented from becoming live by some other aspect of the design (being insulated on the inside or being earthed).

Interesting point of view ...

Well I'm pretty sure it would not be compliant to any standard anymore (correct me if I'm wrong). But also no one instead of hobbyists are doing it nowadays (except the audio-people building boxes from it ...).
And if it burns ... it burns!

For my project I strictly use Metal enclosures only when it comes to mains stuff ...