Putting Load on 5V Line???
 

Ok so I just got this 550w power supply and I'm wondering how much load i should put on the 5V line to increase the power on the 12V line. I've been told around 5 ohms, will this suffice or should i put more????


Edit:

1. I've heard the term "power resistor" being throw around, and i'm not familiar with this term. Is this different from a normal resistor??

Power resistors are able to dissapate more power than their regular old counterparts. AllElectronics.com sells some 50W ones.

As far as the ammount you need to use for loading the 5v line, I can't help there, sorry!

You decided not to go with the meanwell?

No meanwell :(
 

Ya my budget on this limited me to stick'n with the 550w supply and also the price dropped $10 so for $36 shipping and tax i just went ahead and ordered it, got it today.

http://cgi.ebay.com/ws/eBayISAPI.dll...&category=3670

Would have loved to have a Mean Well, yet with the waterblock, pump and such I couldn't afford it. Poor College student here. I'm going to start putting it together later on this week, maybe on fri., I'll report back with some measurements, hoping for the best.

Hey PKR,

No guarantees, but it's worth a try... :)

Placing a 5 ohm resistor across your PSU 5 v rail will pull 1 amp (I=E/R). A 1 ohm resistor then will pull 5 amps. For what you are doing, I would get two 1 ohm 50 watt (or better yet, 100 watt) wire-wound power resistors to place a 5 or 10 amp load on the 5 v rail. Depending on how your PSUs regulator circuits are designed, this may or may not help stabalize the 12 v output.

I use a bank of 1 ohm 50 watt resistors in a home made PSU tester and they will get very warm - almost hot (going with the larger 100 watt units won't get as hot).

Good luck - let us know how you make out... :D

CAUTION: Some PC switching power supplies will self destruct if operated with no load on the 5 v rail ! To be safe, you should ALWAYS have at least a 1 amp load on the PSU when powered up. The little PC power supply testers you can buy will take care of this for you - as well as an easy way to turn it on without having a mobo hooked up or jumpering wires... :)

Ok, here's a noobish question:

I recently powered up an ATX PSU, with the paperclip trick, but I didn't put a load on the 5V line.

The 5V line read 5.2, and the 12v: 11.71.

Should I have put a load on the 5V line?

I have a 50W 10 Ohm resistor, would that do it?

You would only be drawing 0.5A if you did that, where apparently you need at least 1A to stabilize things.

Hum... thanks for the reply, but I thought 1 amp was needed to power up an AT PSU. Is the same true (paperclip trick notwithstanding) for ATX supplies?

Hey Ben,

I don't have a definitive answer for you. My understanding is that different PSUs handle a no-load state differently depending on their actual circuit design. This applies to both AT and ATX switchers. It is my opinion that a 1 amp load on the 5 v rail is a generic rule-ot-thumb. Do it and you're probably safe, don't and maybe pow!

I had direct personal experience with this about 3 yrs ago. A sparkle 350 watt ATX. Powered it up by grounding the green pwr-on lead in the ATX connector (paperclip trick). I was playing with some different fans. It had been on about 10 seconds and I was reaching for a fan to connect, when pow sparks, smoke and game over!. Since that time I have read countless references from numerous sources warning of not powering up a switcher with no load. Sometimes you get away with it and sometimes you don't.

Sorry, I can't be more specific about the exact current draw - my guess is it varies from one unit to another. I have never had a problem since when using a 1 amp load.

Both of the PSU testers I have use two 2.5 ohm resistors in series (thats all a tester is - a 5 ohm load resistor across 5v line, jumper from pwr-on to grnd and an LED on the pwr-ok line).

I also had a EE at PC Power & Cooling tell me that the 5 and 12 volt rails in most PSUs share a common regulator circuit and pulling a heavy load on one with little or no load on the other can cause the unit to loose regulation. Case in point: Epox 8K7A mobo pulled all power thru 5 v rails - some PSUs had real trouble with this and one recommended solution was to try puttin MORE load on the 12 v rail to bring up the 5... :) It apparently worked for some people but not all.

Wow, I wish I had known that!

The PSU was on for an hour, without a load, and everything was fine.

I'll try my resistor, and see what happens. In the meantime, I'll make a note to pickup a couple of 1 Ohm power resistors.

Thanks!

VERY interesting RoboTech.............

I never hear about the 1A load on the 5V line before and as such have used the paperclip/tweeser/whatever trick for some time without ever having a PSU blow on me before. I was aware that not having a "true" load on the lines caused the PSU to not put out the correct voltages on all lines though. When I go to test PSU I usually see if the voltages are somewhat within specs or just way out in left field when doing that test.


Hmmmmmmmmmm.......

Also (if I understand what you are saying) is that I can potentally get better voltage ratings off my lines if I increase the load on some of the other lines? For example, I have an Enermax 650W PSU that I recently replaced with an Antec True550W because the Enermax was not producing the correct voltages on the 5V line. I knew that the 5V line was heavily loaded and that the 12V line was not. Could I have added some load to the 12V line and then gotten better voltages from my 5V line? IF SO, does this also apply to the TruePower PSU then?

Very very interesting.......now I am going to have to get some (2)x2.5ohm resistors or (1)x5ohm resistors and make up a real PSU tester... BTW what Wattage resistors should I use when making up this PSU tester? 1W? 2W? 20W???????

Use P=VI and V=RI to calculate the required power. for the 5 V line @ 1 ohm, you must have 5 Watts. For 1 amp on the 12v line, you must have 12 Ohms at 12 Watts.

I would go with a lot bigger (higher wattage) resistors than that... :)

Several sites (PC Power & Cooling, Antec, etc.) sell pre made load resistors, which consist of a 5 ohm, 25 watt resistor attached to the 5 v lines of a 4-pin Molex. These will get hot after a few minutes!

I know... P= IE =1 amp x 5 volts = 5 watts... => VERY hot!

This little PSU tester I made has (7) 1 ohm, 50 watt resistors in the center cage (puts load on 5v and 3.3 v rails and (1) 1 ohm, 200 watt across the 12 v rail... :)

http://www.leesspace.com/images/Quie...ester_rear.jpg

Notice also they are air-cooled... :D (and they still get warm)

wow...learned something here. i always wondered why there was something called a 'load resistor' over at directron.

i see what cal-centron has in stock and see if i can level out the volts on the true430 im using for my pelt. im not exactly sure what i measured it at, but i think its was in the low 11.XX volts

as for exploding psu's, hasnt happened to me (yet!) and ive started a bunch of pos AT's once with no load at all. hell, my bench-top psu that i use for testing out the things i make hasnt blown up yet either (250w generic ATX from about '94 or so) and i got a nice switch on the green-black lines to power it up

UPDATE!!
 

UPDATE!!!

OK so i have my 550W power supply, look at my 1st post for more details. An I put an 8ohm resistor on it rated at 20w. This resistor get HOT!!! wooo. So what I plan on doing is getting another 8ohm resistor and put it in parallel with the one i have now giving me 4ohm. Maybe the added resistor will cool it down. I had to get an 8 ohm cause Radio Shack here didn't have anything lower. :(

Anyway running setup tests now and hopefully by tomarrow i'll have the rig together.

Question: How do I measure how much current is being drawn????

A typical ohm meter can measure a little current, but I wouldn't recomend it, as if you exceed the rated limit, you could fry the ohm meter.

I used a current sensor and scope, borrowed, of cource, I couldn't afford anything like that!

I think you can get current sensing probes for normal ohm meters, but I don't know for sure. Start looking around fluke's website and you might find something ;)

Hey PKR,

Glad to hear it's coming together... :)

Two 8 ohm resistors in parallel won't cool them down - you will just have TWO very hot resitors. Are they the type that can be mounted to a heatsink? If not, just mount them so they don't touch anything and maybe get some air blowing over them.

Most typical Volt-Ohm meters can't read the high current values your TEC will most likely be pulling. I use a FLUKE 36 clamp-on ammeter 0-200 Amps (But $$$)

Good luck,

Ok, I hooked up my 10 Ohm 25 Watt resistor to the 12V line of my 300W ATX PSU, and with the paperclip, and no load on the 5V line, I got 11.58 v on the 12v line. (tweak time!)

The resistor warmed up rather quickly:D

I'll still try adding a 1 Ohm power resistor on the 5V line, to see if it changes anything.

hey how much is everybody getting these 'power resistors' for? just a quick and dirty estimate is ok if you cant remember

www.allelectronics.com

1 ohm, 75 Watts (Way over rated, only 5 Watts is needed, but you have to go a smidge higher, for safety and accuracy).

$3.50 a piece.

Of course they have to be out of stock on it, + discontinued!)

So 2 * 2 Ohm, 10 Watts each, 0.33$ a piece. Hook them up in parallel for 1 Ohm.

Antec Truepowers have independant monitoring of the voltage rails (+3.5V, +5V & +12V), with no max ampage except the PSU~model (350w, 400w, ect) max for all three rails combined. So loading the 5v rail will do nothing for the 12v or any other rail with these PSU's...

The independant 12V rail is a boon if you're buying a Nforce2 mobo. Some, or all,(Abit for sure) use the 12v rail to power the CPU instead of the 'traditional' 5v rail...