Dropping PSU voltage

LED Flashlights – General Info
1

I have a desktop power supply that measures 13v, 5.6v and 3.5v for the respective 12v, 5v, and 3.3v rails, all of which are out of spec. Is there an easy way to drop the voltages for the rails? Based on googling it appears that I can splice in some IN5817’s to drop the voltage but is there an easier way to do that? And is there a limit on the current passing through the IN5817?

Btw I have a 10ohm 10 watt power resister hooked up to a 5v wire. And voltages were measured with 2 different DMM’s which haven’t given me cause to doubt their accuracy.

Thanks

2

I think resistor is unsuitable because it’s voltage drop will depend on the current that goes through it, so for very low load you would still have overvoltage, and as load increases, you would get bigger and bigger voltage drop.

Perhaps though the PSU lacks some more load on it so it would then stabilize within spec? Perhaps you can try increasing the load ? Minimal consumption of a PC configuration should be in 50-80W range, so perhaps that’s what’s lacking.

if it is still out of spec after you load it more, I wouldn’t waste time on it. But it is strange that it’s got higher voltage on all rails.

3

I assume you have used a DMM for checking the voltages. Has the DMM been confirmed that it is reading the correct voltage?

4

I’ll load it up some more and see if it makes a difference. Thanks for the suggestion.

5

The 1N5817 is a diode and will drop between 0.2 and 0.5 volt depending on current, maximum current is 1A.

The best would be if you could do some adjustment inside the power supply to fix all the voltages.

6

I measure it with 2 DMM’s (Fluke and a cheapie Harbor Freight one). Although the Fluke hasn’t been calibrated in many years it has always seem accurate given things that I’ve measured with it (fully charged car batteries, car alternators, charged 18650’s, new out-of-package primary lithiums, etc). Unfortunately I have no way to really measure its accuracy against a known voltage or calibrated DMM :frowning:

7

Yes I hear you and your not alone.

8

A known voltage is fairly cheap.

You can get different types here: http://shop.voltagestandard.com/product.sc;jsessionid=9535E124E172E23769F782A57C50DA32.m1plqscsfapp04?productId=1&categoryId=1

Or you can get this from ebay:

9

Are you trying to use it for a desktop computer or just a power supply?

10

Thanks for the link. I might pick one up so that I can calibrate my DMM’s.

11

It came out of an old desk top computer. I’m making it into a power supply for a friend and I was checking the voltages of the different rails after I connected them to binding posts.

12

1A would be too low for the chargers :frowning:

I was looking for some way to adjust the power output from the inside but nothing jumped out at me. Mostly likely it’s not knowing what to look for :expressionless:

13

the link is not working for me :frowning:
searching ebay for “voltage reference” didn’t yield anything similar

14

Sorry, I copied the link when I was in their shop and that included a session id. Try this without any session id: http://www.voltagestandard.com/

15

It’s weird how the 5V rail is out of spec even though there is a load resistor attached.

Did you try checking the voltage under say a 30 or a 50W load?

16

I haven’t done so yet mainly because I ran out of resistors. I’ll make a run down to Radio Shack to get some more this weekend. However I connected my Accucel 6 to it and started charging a 18650 at 1A to load it up some. The 5v rail went up by .02v and the 12v rail decreased by .2v. What it all means I have no idea 8)