hands without shadows

I got my hands on a free computer so I took it apart and found an ATX power supply and decided to try to convert it to a lab power supply with common, 12v, 5.5v, and 3.3v posts. I put a 10 ohm 10 watt wire coil resistor between black and red (negative and +5.5v) and an on/off toggle switch between green and black (power on and negative). I bundled and soldered the leftover red (+5.5v), orange (+3.3v), black (negative) and yellow (+12v) wires to separate posts. The other wires were blue (-12v), purple (+5.5standby), grey (power ok) and white (-5.5v) which I cut short and made sure they aren't touching anything. There were no remote sensing wires. Now when I turn it on I get nothing, whats wrong? It worked before.

Andrew

Your conversion process is correct, providing that it was not a DELL computer. Occasionally during conversion, a component on the PSU circuit board may end up with a broken lead or have a solder joint part, but this is fairly unusual.

Since the PSU did not have sense wires, the next most common problem is grounding one of the binding posts on the case. This happens if the post moves when being tightened or with posts not remaining centered after being tightened. If you have a meter, check for continuity between each post and the case, including ground. Some power supplies isolate the DC ground from the case, some do not.

The symtoms are most like a missing sense wire or a grounded post causing the overload circuitry to capture.

hands without shadows

I checked and made sure there are no sense wires and that the plugs arent grounded. I also tried giving the power ok wire signal to mimick a post cycle, I still have no response. When its plugged in I can hear the coils humming, I just cant believe its fried.

Andrew

When the PS is plugged in, but without the PS_ON (green) grounded, do you have voltage on the +5 standby and the PS_ON line. Both should show around 5v.

runout74

I have gone through the same process with converting a PC power supply. With a 4amp load (charger) it works good for about 3 minutes. Then the soft hum turns into a soft clicking sound. Then the power supply output will flicker a couple of times and then it shuts down. I am using a cheap Dynamite Prophet 4A Fast Charger so I am unable to adjust the amps. The PC power supply is rated at +12v/16Amp. 350 watts. I have (2) 1ohm/10watt resisters in series on the +5v. It powers on perfectly with 12.15 volts. I don't detect any heat on the exhaust side of the power supply. Seems to be running cool. Any ideas? This is my second attempt at converting a power supply. I am starting to get frustrated.

How many amps can I hope to draw off of this PC power supply? What is everyone else getting?

vti-chris

Do a search on the net and there are 2-3 sites that tell you step by step about all three types of PS.

runout74

What are the three types of power supplies? I am only aware of two: switched and linear. Or are you referring to ATX or AT or what?

I have searched high and low on the net and cannot find what you are referring to. Here are the websites I referenced when making changes to my PC power supply.

[link=http://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supply]www.wikihow.com[/link]
[link=http://www.marcee.org/Articles/PCPowerSupply.htm]www.marcee.org[/link]
[link=http://web2.murraystate.edu/andy.batts/ps/POWERSUPPLY.HTM]web2.murraystate.edu[/link]


Is this what you are talking about? If you don't mind, give me an example link I can look at.

I appreciate the info.

Woody (Runout74)

Andrew

Your supply should easily carry the charger -- generally when the overload circuit captures, there is no sound and the PSU just shuts down. Can you be more descriptive of the "clicking" sound and possibly determine the general source. What is the frequency -- slow, 60 hertz, etc.

Some devices will chatter if internal insulation has broken down and you are getting internal arcing -- usually this does not last too long and almost always eventually ends with a POP and catastrophic failure. Capacitors will also whine occasionally, usually at a high frequency. It sounds either like an overload situation or component failure.

andrew

runout74

First off, thank you for your reply. I appreciate the help.

The "clicking" sounds like internal arcing. Very faint. The general source of the sound is either coming from the large capacitors or the big wire wound thing (transformer I think) in the middle. [sm=drowning.gif]

After 5+/- minutes of charging this is what happens. It happens pretty quick. The clicking will start intermittently for a few seconds. Then the clicking is in sync with my LED flashing off and back on indicating there is a break in power. Then the final click shuts the thing off.

Since I have not successfully converted a PC Power Supply I am thinking my method is the problem. Since I used (2) 1 ohm/10watt resisters in series, is the power supply latching off or not staying latched on. Also, I read a site that showed the lower the resistance on +5v the lower the amps available on +12v. I read that here [link=http://groups.google.com/groups?q=+%22The+power+resistors+will+generate+hea t%22&hl=en&lr=&ie=UTF-8&selm=386a8871.925894%40news.shasta.com&rnum=1]groups.google.com[/link]. What resistance should I be using?

Runout74 (Woody)

Andrew

The 2 ohm load is the minimum you should use -- there is no gain in reducing the resistance any further. I usually recommend starting with a 10 ohm load -- some PSU's will function well at 20 ohms. The only reason for lowering the resistance is to push the voltage up on the 12v rail -- however, many chargers will function quite well at 11 volts or less in some cases. If you feel the voltage level on the 12v rail is not high enough, you can drop to 5 ohms (two 10 ohm resistors in parallel) then lastly to 2 ohms.

Without the charger attached, power up your supply and see if it shuts off on its own after being left on. If it does, then you may have a faulty supply. Overload circuitry doesn't cycle -- normally, the PSU just shuts down and remains off until the load is removed or remains off until it is cycled thru a power down/power up sequence. Some supplies may need several seconds of power off time to reset.

Run it without the charger long enough for the components to come up to operating temperature -- it won't hurt it to run without an external load. Intermittant problems are hard to find and sometimes are heat related. From your description, it really sounds like a weak component or perhaps a solder joint that is opening up when it gets warm.

andrew

vti-chris

Here is another one:

http://www.wikihow.com/Convert-a-Com...b-Power-Supply

i didn't find the one i was telling you about.

http://www.qsl.net/9w2lw/power.htm

runout74

WOOOO HOOOO !!!!!! That was it. I swapped out the (2) 1 ohm/10 watt resisters and put in a single 10 ohm/10 watt resister and viola, IT WORKS !!!

Thanks for the help Andrew.

runout74

Thanks for the links. The QSL site is cool !!!

jaofos

I'm looking to do something similar to this and have a few questions.

I found an older server power supply being thrown out here at work, the specs are as follows:

425W

Outputs:

+5.1V -- 37A
+3.4V -- 30A
+12.05V -- 18A
-12V -- 0.5A
-5V -- 0.5A
+5V AUX -- 1A
+3.3V AUX -- 1A

Now I plan on using this with the A123 Sonic Charge system, my question is concerning the 18A on the 12v output. Does that means it's going to be constantly pushing out 18A, or is capable of up to 18A load? If it's constantly pushing 18A and I hook the A123 Sonic Charge system up to that kinda output, will it potentially damage anything? From the specs I read on A123's page it says nothing about the Amperage output of the power supply, only that it accepts 10-15V input and outputs 10A.

Any help is much appreciated.

Andrew

It means that the rail has an 18A capacity -- your charger will consume what it needs. Generally, when input specs are printed for chargers, the current requirements are for maximum charge rates. Unless you max out your charger on high capacity batteries, these current demands will never be realized.

jaofos

Unfortunately I couldn't get the PS from work to do anything, it was an older compaq proliant server power supply and probably wasn't wired like most standard power supplies.

I did have an old 450W enermax power supply from my old PC that does 24A on the 12v rail and I managed to convert it fairly easy (did I mention my solder skills a quite lacking?). Charged my two 6 cell packs at 4 amps using my current charger running off the dc output of my power supply, so I would say it's working. Now we'll see how it holds up under the 10A the A123 system will pull from it!

Also discovered the power outlet in my garage isn't grounded properly.. love that tingling sensation when touching the power supply :P

hunter306

I started doing a little conversion project of an OLD psu. It works great right now, but I have to have my load hooked to it before I power on, otherwise it shuts down... Is this a pretty normal reaction to adding a load to the PSU?

I'm wondering if i should just scrap this psu and get a little bigger one to try out.

jaofos

That's why the instructions call for the resistor on the 5v rail, to put load so it powers up when you flip the switch. Mine had two fan's already on the 5v rail so they provided some load, but not enough to keep it running to charge batteries. I ended up putting the 10W 10Ohm resistor in and it charged both my battery packs perfectly.

Andrew

INTEL Design Guide specifications state that an ATX12V2 power supply should be able to power up and operate with a 5000 uF capacitive load initially present on the 12V1 rail (3000 uF on the 12V2). This requirement is due to the heavy capacitive load imposed my most of the motherboards today. However, this load stabilizes quickly - capacitive charge rates are well known. Short circuit protection initiates latch_off when any rail senses an output impedance of less than 0.1 ohm. Latch_Off will remain in effect until the short has been removed -- some PSU's will recover automatically; some will require cycling the PS_ON (green) for a minimum of one second (I have seen some require 10) -- power down, then power on again.

Most of the newer smart chargers have significant ripple suppression capacitors on the front end -- when discharged, these capacitors can represent a sufficient current sink to cause the overload circuit to capture and shut the PSU down when the charger is attached after the PSU has stabilized.

Your power supply is behaving as designed -- attaching the charger before powering up is a workable solution.

hunter306

That was what I thought Andrew, I have been working building PC's for quite a while so I'm pretty familiar with it.

The behavior seemed normal to me, that's why I haven't worried about it. No problem hooking up the load prior to turning it on, just feels a little unusual for testing purposes (bench supply)

AND... I do have 20 ohms of resistance (2 10ohm 10w) in series on the 5v line.

runout74

Need more info:

Without a load does the PS turn on and then shut down? If so, how long does it stay on before shutdown?

Is this a dell power supply?

I am curious if you removed the 2 resistors if it would act the same way. You could remove the resistors and start the power supply under load. If it turns on with a load or will not turn on without a load, then I would say your resistors may not be connected properly.

hunter306

Actually I think I've already got this answered, but I'll indulge you.

It is not a DELL power supply. It's just an old 150w ATX psu. I know the resistors are connected because they show current flow when I power it up with the ol' multimeter connected in series with the resistors.

The power supply had no trouble starting up and working without the resistors, but I've been told that the switching transistors cannot operate under a no-load situation because they have a tendency to fail at a accelerated rate.

For my test, I was using a 27t 540 sized motor on the 5v line... I expect this could pull quite a few amps, so It could be what is causing the 5v line to initiate shut-down. I have not yet tested it with a smaller 3 amp Incandescent Bulb load... I should try that this evening.

mustang1964.5

hi i have a question about power suply. i converted one and i'm using it to power hub for internet and e-flite blade cx transmitter since i fly mostly indoor. i connected an rj-11 phone jack plug to the side of the transmitter and one to the power suply amd using 100 feet phone cable.

i'm trying to learn how to fly the blade cx so i can move to bigger heli but battery fly time is like 10-15 min a charge (guesstimating here) then i have to wait for battery to charge for about 1.5 hr wich is not enough time to learn for me, yup i need to buy more batteries. so i was thinking can i convert a power suply to give me 7.4 volts so that i can power the blade cx connected with phone cable and if posible to get 7.4 volts from power suply will it be safe to run the heli or will i fry it since i dont know how many amps the power suply will output at 7.4 volts and the li-po 2c battery pack is 800mah wich i dont know what that means. anyway, i connected the heli to 5 volts but did not have enough power to hover it barely wanted to take off.

i read this page (http://reckerclub.tripod.com/id105.html) and saw that by using different resister values the voltage changes so maybe i can get 7.4 v or come close to that value enough power to hover.

i have like 4-5 power suplies cuz i fix comps on the side and could mob one just to power the heli but i dont want to fly it since i dont know much about electricity so i'm hoping i get good advice here.

thanx in advance if anyone can help

p.s. sorry for my english. english is my second language.

Andrew

If yours is a standard ATX power supply, use the 5v rail (red) as ground and the 12v rail (yellow) as positive -- you will get 7 volts. If the voltage is down on the 12v rail, adding an additional resistive load to the 5v rail will usually push your 12v rail to 12.1 or 12.2 volts.

mustang1964.5

here's a drawing of what i'm talking about thanx