Needing help with a PS design.

[chipperi]

At work I spend my days building and fixing test equipment and analyzing, under supervision, failures that come back from the production line. (we manufacture battery chargers and DC to DC converters.) I am a pure hobbyist when it comes to electronics, I am a general class amateur radio operator, so I have some basic electronic experience.

I was telling the boss that I wanted to build a bench top power supply from scratch just to better my understanding of how some of what we do works....later that day he gave me a 120/55ct transformer and said if I use the whole secondary I can pull 6A, if I use the center tap I could pull 12A. He also gave me a gbj2506 bridge rectifier to get me started.

SO... I am thinking variable 0-30Vdc 10A controlled by a atmega328p-u because I already have one (not Arduino) I want to be able to set the CC and Voltage much like the one I use at work. Dave did a similar build in video 222 a while back and liked it. I have seen so many schematics and posts but most are 5A and under and many use old obsolete parts.

Needless to say I will be picking your brains fairly frequently, but for now I would like some ideas and pointers to get me going.

So Specs:
Variable 0-30Vdc 10A
Transformer 120Vac Pri  55Vac Sec center tapped CTvolts = 27.5Vrms*1.414= 38.885Vdc
Rectifier        GBJ 2506   600V Vfwd 1.05v per element  25A
Adjustable Voltage and Constant Current
As precise as affordable
Atmega328p-u control
Good readable LCD
Relatively current parts and techniques.

Hope to hear from you guys.

Chip

[Landrew2390]

The architecture that Dave uses in the video is one of the simpler, fully adjustable ones I've seen.  The bigger problem the amperage you're asking the power supply to handle.  In a switched mode PSU, it's going to be less of an issue, but linear designs are going to run into thermal problems quickly.  Some back of the envelope numbers are coming out at 375 watts of power dissipation at maximum amperage and minimum voltage.

The way most commercial units get around this problem is having multiple taps on the transformer.  Each tap is switched in or out depending on the output required.  In this manner, the voltage being fed into the regulator stage remains closer to the output voltage.  If a proper multiple tap transformer is used, you could potentially get your maximum power dissipation down to ~125 Watts.  That's still a lot of power for a regulator and heatsink to handle, but it's far more reasonable.

My advice is reexamine your specifications.  If you need a full 10 Amp output, then you need to design for it.  I sat down and looked at my projects and determined I only needed a 1 Amp supply for my designs.  You'll still be able to use the parts you've been given on a smaller supply, they'll just be oversized.

[dom0]

Variable 0-30Vdc 10A

Almost no one actually needs this.

You probably want/need some fixed range at 10 A, maybe something around 24 V? But you probably never need .5 V and 10 A. Or 2 V and 10 A. Even 5 V and 10 A would probably be very, very rare today.

If all of this is true, then you actually want a supply with said 2x V and 10 A, with foldback current limiting. This avoids blowing stuff up by limiting current the more the lower the output voltage is. E.g. at 24 V you can pull 10 A, but at 1 V you only get .1 A.

[Kleinstein]

For so much power you should have a transformer with more taps. Already having a second useful tap could reduce the heat production to about half.

The the more complicated alternative would be having a step down stage to reduce the voltage if needed. However this would produce more noise, so a lower quality output. It can be still possibly to have a clean output at lower current limit (e.g. up to 2 A).

[AlxDroidDev]

What kind of stuff do you need 10 amps for, if you don't mind me asking? That really is quite a lot, and such a PSU needs an excellent (read "complex") project not to mess up your day.

[chipperi]

What kind of stuff do you need 10 amps for, if you don't mind me asking? That really is quite a lot, and such a PSU needs an excellent (read "complex") project not to mess up your day.

I based the specs on the bench top supplies we have at work... they are (I think) 0-50Vdc 50A I doubt I would ever use 10A at home...at work I regularly go over 10A.  The specs are just what I would like to do but I am always game for change.

This is the one I use frequently at work...http://www.volteq.com/volteq-power-supply-hy5050ex-50v-50a-over-voltage-over-current-protection-110v-input.html

[tron9000]

What kind of stuff do you need 10 amps for, if you don't mind me asking? That really is quite a lot, and such a PSU needs an excellent (read "complex") project not to mess up your day.

I based the specs on the bench top supplies we have at work... they are (I think) 0-50Vdc 50A I doubt I would ever use 10A at home...at work I regularly go over 10A.  The specs are just what I would like to do but I am always game for change.

This is the one I use frequently at work...http://www.volteq.com/volteq-power-supply-hy5050ex-50v-50a-over-voltage-over-current-protection-110v-input.html

JESUS! I really hope the 50A comes out the back terminal cos I know for a fact the 4mm banana plug jacks are only specified for 10A maximum! that's no PSU; that's a DC arc welder!

[Landrew2390]

My advice is building a smaller PSU first to make sure you understand how they work and to get your control theory down.  A 30 Watt PSU belching smoke and flames is a lot easier to deal with that a 300 Watt unit.

And as a disclaimer, I did have that happen to me.  I'm really strong at hardware, but am struggling to progress beyond "Arduino Noobie" when it comes to software.  As a result, I messed up the programming on the controller and the transistor I was using to regulate the output went past it's design specs by a full order of magnitude.  Lots of smoke and flaming bits of plastic.

I don't know where your abilities lie in hardware or software, but I recommend you start small unless you're comfortable with all the concepts involved.

[dom0]

What kind of stuff do you need 10 amps for, if you don't mind me asking? That really is quite a lot, and such a PSU needs an excellent (read "complex") project not to mess up your day.

There really isn't much of a difference between a 30 Watt and a 300 Watt lab PSU. The  30 Watt unit will need a solid thermal design, as will the 300 Watt unit. Tap switching may be implemented in both. Instantaneous SOAR protection is always a must, and so is thermal protection. Numbers change, parts change, but the design, the topology, is the same.

What IS different is that you will need multiple pass transistors in the 300 Watt unit, and you will likely need a beefier driver, or multiple drivers, especially with MOSFET pass elements. And you might want a soft-start circuit, not only due to the large transformer (especially if it's a toroidal), but also because of the large filter caps, to avoid blowing fuses (or be able to use a useful fuse rating, not 15 AT like in some Chinese 300 W supplies).

Oh, and failures during build / evaluation, if any, tend to be more violent in nature, as expected.