Another ATX to bench power supply question?

Hi everyone, I'm very inexperienced with electronics (I'm currently in high school physics and we haven't gotten up to the electricity unit yet, so I have no formal education) and decided to build a bench power supply out of a computer ATX supply because it should be a good learning experience and would provide me with a cheap, high-current power supply. I understand the concept of everything and it all seems straightforward; bundle same-colored wires into a common output, solder power and ground to a switch, solder sense to +3.3V, place a 10 ohm 10W resistor between a +5V and ground. There's just one problem I'm running into and it involves the outputs:

How should I go about making outputs for the positive voltages and ground? I've considered banana plugs and binding posts, but these tend to only be rated around 5/10A and my +5V rail (which is the output I expect to get the most use out of) outputs 25A. I'd like to be able to perform experiments such as electrolytic cells, so I'll probably need to (or rather it would be much more convenient if I could) pull almost all of that current.Is there any sort of convenient, high-current-capable connector (I was considering maybe alligator clips?) that I could buy either from Radioshack or Home Depot? Sadly, those are the only stores of this type in my area and ordering online is too much of a hassle for me.

Thanks in advance for any help anyone can provide!

 

Yes, I know that switched mode power supplies such as an ATX PSU needs a load to keep running (and I think some of them will keep running but don't output the correct voltage when there's no load present on the +5V rail or something like that). That's why I'm going to use the 10 ohm 10W resistors, which most guides I've read have suggested. I think it's just because someone decided to use them and found that they worked and everyone else just copied that one guide and followed suit, but if it works, I don't feel like messing around with other types of resistors. Plus, those resistors ARE pretty cheap.

Also, I tried looking for that type of connector by searching "wheelchair" on Home Depot and Radioshack's websites and got nothing. I do see it from other random sites and yeah, they're rated around 50-350A depending on the model you get, but I really would like to stay local with this purchase, especially because it's not an essential part. If I can't find anything anything I guess I could just solder all of the leads for each voltage to one thick gauge wire (which I know Home Depot does sell) and solder that to whatever I want to power, which is cumbersome but it does work. I'd just like to know if there's any way to do this less painfully.

EDIT: Oh, wait, I didn't see that you specified I should add a load to ALL of my outputs. Would it work if I just bought two resistors and hooked one up to, say, 0V and +5V, and the other to +3.3V and +12V? I don't plan on using -5V or -12V because their current rating is so low (0.3A and 0.8A respectively).

 

For the connectors, I believe Bill was referring to something like P/N 7043K21 at McMaster-Carr (http://www.mcmaster.com/#battery-connectors/=3uz5vc). That's a 50A battery connector. There is no male or female housing. The way they are keyed keeps them from being connected with reversed polarity.

Unless you are doing some awfully high powered work, I think you can get by with simple, high-quality banana jacks.

The load you have on the 5V supply may be enough. My PSU, which is made from a 430W Thermaltake supply, works just fine with only the 5V load.

John

 

Hmm, thanks for the link. But I do think I'll end up utilizing most of that 25A capability, especially since my primary goal (at the moment) is to run an electrolytic cell off it, and the speed of the reactions in the cell correlate directly to amperage. Though I can't really use connectors in an electrolytic cell either way because electrodes can't exactly accept connectors. So I'm at a loss of what to do. Yes, I could always just solder wires to an electrode, but what if I want to do something else with the supply later, like powering an actual device? Also, soldering is a pretty permanent solution, and an inconvenient one at that if I decide that I need to move the cell because I'll have to move the PSU with it.

 

Go to your local hardware store and buy a 30 amp rated outlet and plug for a dryer. You won't use all the pins, but the connectors are pretty rugged.

 

Well, I just got back from Radioshack because I always pass it on my way home and I saw connectors in their Power Connectors bin called "2-Position Interlocking Connectors": http://www.radioshack.com/product/index.jsp?productId=2103448

I bought those and a female version of those too. I have no idea how they work in the slightest, but they are rated for 20A and I figure that's the best I'm gonna get at that price, so I'm going to try to figure out how to use these things. I haven't even opened the box yet. Should I or are they useless? Also, if they're useful, will I need to ensure that the 5V rail never puts out more than 20A somehow (the others are all below 20A) so that the connector doesn't get damaged?

 

If you need a 25 amp connection and use the Rat Shack battery connector, it will melt at that level of current. You can always try two in parallel, or else get a connector rated for the current.

 

There are also what are called "Deans" connectors at your local hobby shop. They are low-resistance connectors designed for modern high-power DC electric motors. They don't cost much more than the cheap ones you got, and they are soldered, not crimped.

John

 

Thanks, I looked them up and now I'm deciding between these two:

http://www.toysonics.com/10-pairs-2-pin-golden-plated-deans-connector-t-plug.html
http://www.toysonics.com/10-pairs-3.5mm-gold-plated-bullet-connector.html

Which one is preferred, if either? The Deans connector looks safer because of the plastic, but the bullet connector is cheaper and withstands more current (not that I'm exceeding 25A anyway).

 

If you are working on HHO cells, use stainless steel machine screws to connect your leads to the cell, and make sure your connections are well made in the container, any spark can explode your cell (HHO burns over 4000 times faster than gasoline vapors.) See my HHO videos on youtube >>>>

http://www.youtube.com/watch?v=5yNl6CeZeSc
http://www.youtube.com/watch?v=7fcC_xZv-dw

and another thing too, have you actually tried to power your cell from the PC PSU?? It will probably trip every time you connect your cell to it (The cell will create a short on the PSU's power rail, I have tried the same thing). You would be better off connecting a car battery to it, and just keep a charger on hand to charge the battery when needed.

My .02

 

Well I'm not able to go on Youtube right now, but I'll check those videos when I get home.

But yes, one cell I was planning on running is an HHO cell. I was also possibly planning on running a chlorate cell. Yes, I'm aware that stuff is rather dangerous, but I'm not exactly making fireworks or anything like that. The site that most other sites I've read refer to for chlorate (http://www.geocities.com/CapeCanaveral/Campus/5361/chlorate/chlorate.html) does suggest an ATX PSU as a possible power supply, so as long as the cell has enough resistance to draw less than 25A I would think it should work. I can't run that cell above around 15A anyway because the only anode I have access to is graphite.

 

@shadowdude77,

The first connector is a Deans copy; the second is a "bullet" connector. Either will work. On a 40A+ electric model airplane, I use bullet connectors to the motor, because there are 3 wires. For the battery connector, I use a Deans. The Deans is just one connector, and the geometry prevents connecting anything with reversed polarity.

John

 

Hmm, okay, thanks. I think I will actually go with Deans connectors because they are only about 50 cents more expensive for the 10-pack and they're not only better looking but polarized.