Author Topic: My First PSU to Lab Power Supply, Design and Resistor Question?  (Read 5990 times)

0 Members and 1 Guest are viewing this topic.

Offline mictasTopic starter

  • Regular Contributor
  • *
  • Posts: 127
  • Country: au
  • Well I am stupid.
As my tag says, I am new and want to know what people think, plus I have a question.

In the next few days, I am hoping to put this one together, its to test regulators that I have been taking off old TV board, mostly theses TV boards have been from the old LCD TV's that have broken down. (most TV's had logic board errors, and this is what people have been dumping at out hacker space -_-)

The power supply is an old P4 unit, with:
400w
+3.3V 20A
+5V 20A
+5VSB 2.0A
+12V 10A

These are the only voltages, I'll be needing for this project, if it comes to me wanting more outputs, I have a step-up regulator that I'll be testing and might get myself a hobby box and do a whole rework, but for now I'm just wanting four out puts and two ground.

If you look down in the attachments, you'll see a quick and dirty schematic for what I'll be adding on, I'll also say that any wires, I will not be using, will be desoldered and removed. I also plan to place the fuses housing on the top of the power supply, I would put it in front, but that just is not going to work. Also all connecters are going to have there own wiring, like the power okay led will have its own ground, as will the Power On LED will also have its own ground wire, the 5w resistor too will have its own ground, each 12v connector will have its own 12 wire, etc.

So now for the question, it has to do with the 5W resistor, the photo I have attached with this post, are three of what looks to be the closest to what I have seen used online, can anyone tell me if and which one I should use on the 5 volt rail as a dummy load?
I think these are 10 ohms, the others I have are all 4.7ohms, 1N, 15ohms, etc... (Now that I think about it, I also have a 20 ohms resistor in the photo as well.

These resistors have all be recovered from TV power supply's and not tested, I am wanting to use recovered parts as much as possible, next time I plan to try and reuse some of the fuses, if they are still okay.
« Last Edit: March 17, 2014, 11:35:25 pm by mictas »
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 5139
  • Country: ro
  • .
Re: My First PSU to Lab Power Supply, Design and Resistor Question?
« Reply #1 on: March 18, 2014, 04:20:12 am »
It that power supply says 400w but it lists 10A for 12v, then it's really a 200-250w power supply AT BEST.

Also, keep in mind that 3.3v and 5v are usually coupled together. I mean it may do 3.3v @ 20A and it should do 5v @ 20A but both together won't exceed  about 100-120 watts. 

Considering it says 10A on 12v, I would bet that the power supply actually uses 2 x 3A diodes as a 12v rectifier - if you want to use that 12v extensively, you may want to switch the diodes on 3.3v with the diodes on 12v, so that the 12v would be capable of more current.

Keep in mind that the labels LIE.

Here's for example an album with pictures I made of a 400w psu some time ago : http://imgur.com/a/YFWcA

Notice it says : 3.3v @ 28A, 5v @ 40 A, 12v @ 17A  but you look on the chips on the second heatsink, you see there SBL1640CT ( 16 A , 40v rectifier) used for 5v, SBL1040CT (10A, 40v for 3.3v) , and two 3A diodes for 12v in the center.  Now in practice, as they're reasonably cooled, they may do a bit more current than rated, but still far away from the 40A or 28A values written on the label, not to mention the 17A for the 12v.

So you really have to open it up and inspect the parts carefully and determine the power supply's capabilities before taking those ratings as true.

Most power supplies already have some resistors on the output, but they're only meant to keep the power supply from going nuts without any load.  These older power supplies, as they were designed for systems that used a lot of current on 5v.  Without anything on 5v, the output voltage on 12v may be wacky.
As you may not use 5v when turning it on, you need to cheat and use a resistor to consume some current on 5v all the time.

You use Ohm's law :  Voltage = Current x Resistance.

So let's take 10 ohm and 5v ...  5v = ?  x 10 ohm =>  Current = 5/10 = 0.5 A

The power dissipated in a resistor is Power = Current x Current x Resistance , in your case Power = 0.5 x 0.5 x 10 = 2.5 watts which is below the 5w rating of the resistor so you're good (even though the resistor is rated for 5w, you normally don't want to dissipate more than 3.5-4 watts).

 

Offline mictasTopic starter

  • Regular Contributor
  • *
  • Posts: 127
  • Country: au
  • Well I am stupid.
Re: My First PSU to Lab Power Supply, Design and Resistor Question?
« Reply #2 on: March 18, 2014, 06:28:21 am »
It that power supply says 400w but it lists 10A for 12v, then it's really a 200-250w power supply AT BEST.

Also, keep in mind that 3.3v and 5v are usually coupled together. I mean it may do 3.3v @ 20A and it should do 5v @ 20A but both together won't exceed  about 100-120 watts. 

Considering it says 10A on 12v, I would bet that the power supply actually uses 2 x 3A diodes as a 12v rectifier - if you want to use that 12v extensively, you may want to switch the diodes on 3.3v with the diodes on 12v, so that the 12v would be capable of more current.

Keep in mind that the labels LIE.

Here's for example an album with pictures I made of a 400w psu some time ago : http://imgur.com/a/YFWcA

Notice it says : 3.3v @ 28A, 5v @ 40 A, 12v @ 17A  but you look on the chips on the second heatsink, you see there SBL1640CT ( 16 A , 40v rectifier) used for 5v, SBL1040CT (10A, 40v for 3.3v) , and two 3A diodes for 12v in the center.  Now in practice, as they're reasonably cooled, they may do a bit more current than rated, but still far away from the 40A or 28A values written on the label, not to mention the 17A for the 12v.

So you really have to open it up and inspect the parts carefully and determine the power supply's capabilities before taking those ratings as true.

Most power supplies already have some resistors on the output, but they're only meant to keep the power supply from going nuts without any load.  These older power supplies, as they were designed for systems that used a lot of current on 5v.  Without anything on 5v, the output voltage on 12v may be wacky.
As you may not use 5v when turning it on, you need to cheat and use a resistor to consume some current on 5v all the time.

You use Ohm's law :  Voltage = Current x Resistance.

So let's take 10 ohm and 5v ...  5v = ?  x 10 ohm =>  Current = 5/10 = 0.5 A

The power dissipated in a resistor is Power = Current x Current x Resistance , in your case Power = 0.5 x 0.5 x 10 = 2.5 watts which is below the 5w rating of the resistor so you're good (even though the resistor is rated for 5w, you normally don't want to dissipate more than 3.5-4 watts).

I wasn't able to understand any of what you said.

Anyway, what I have done is post a few photo I have just taken of the power supply I plan to use.

But I am guessing that using one of those resistor, like the RWS-5W 10R0JSMT just one should allow it to work as a lab power supply?



 

Offline mictasTopic starter

  • Regular Contributor
  • *
  • Posts: 127
  • Country: au
  • Well I am stupid.
My First PSU to Lab Power Supply, Update.
« Reply #3 on: March 24, 2014, 01:25:32 am »
Well I have had this sitting aside for a few days, but so far I have done this:

I'm thinking about buying a larger power switch, at one point there was going to be two switches, but after I had cut the box, I didn't want to go ahead anymore with two and went back to just one.

I added the banana plugs and added wires to them, before I put the whole thing togather I want to get my hands on a bottle of Liquid Electrical Tape and use that one the ass end of those banana plugs. And I need to get a hold of a lower wire gauge for the 3A 12v out put, or I will just look at linking the -12V and a +5V rail to get 17v output.

The two LEDs are at the moment floating about, the power works great, but I need to change the risistor on the power on LED, the LED hardly lights up.



You can see below that I still have a mess of wires, but a quick test some days ago showed it was working. I added a mains line filter, there was a few that had been recovered from some large TV's and I am wanted to remove the mains lines and shorten them, I have to put a small cover over where the bottom output mains plug use to be.

I went with this 5W 10ohms resistor, as you can see under the mains filter.

As for the power switch and banana plugs, that is what I want to use the Liquid Electrical Tape unless the pack of heat shrink arrives in the post before I get to it. But I am wondering if I need to solder the banana plugs, before covering them up


And the fuses and left over wire and banana plugs, just encase something happens to any of the others before I am done. I am going to install the fuses around the heat sinks on the power board, but that is still a little away off yet.

And as you can see my sad attempt at noting down the pin layouts on the mains plugs.
 

Offline Huluvu

  • Regular Contributor
  • *
  • Posts: 91
  • Country: de
    • ECM Home
Re: My First PSU to Lab Power Supply, Design and Resistor Question?
« Reply #4 on: March 24, 2014, 07:44:22 am »
How about this solution from Elektor : ATX to bench top lab power supply adapter

I think its more safe and you can use different power supplies if needed.

Rgds
"Yeah, but no, but yeah, but no..."
 

Offline mictasTopic starter

  • Regular Contributor
  • *
  • Posts: 127
  • Country: au
  • Well I am stupid.
Re: My First PSU to Lab Power Supply, Design and Resistor Question?
« Reply #5 on: March 24, 2014, 11:42:59 pm »
How about this solution from Elektor : ATX to bench top lab power supply adapter

I think its more safe and you can use different power supplies if needed.

Rgds

But dose that not take the fun out of it?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf