LED electrical question + short circuit question

[plopke]

Sidenote : I am very very very novice when it comes to anything with V/I/R/W,A,Ohm... ;p

1)I did install a custom start button some time ago and plugged that into the motherboard , one of following:

http://www.aquatuning.nl/waterkoeli...mm-edelstahl-blau-ring-beleuchtet-6pin?c=4930


Now it did work before but the LED stopped working so I tested it on a 12V -5A battery which resulted positive in a shining LED. Now I noticed the LED was actually 12V specification from the beginning so hooked it up now to a Molex 12V line and works again. Apparently most front panel motherboard connector only supply 3-5V for LED?

Any reason a 12V led working on 5V before is not working now , weather , led wearing out,...? If I swap the button with a newer one that works still from the motherboard connection.



2) By accident I short circuited either 12V or 5V pins for the button which resulted off course in the power supply shutting down but then it did not boot for a while, I know in a house there safety switches basically jumping into a off position but how does a power supply do it. Because after waiting for a while everything is working again. Could it have dmg any system parts?

Regards Ward.

PS not sure In which forum to post this.

 

[aigomorla]

There can be a lot of things with the LED.

Most front panel only take 3V led.
Rarely have i seen one handle 5V, and 12V is just no.

As for why its not working and working, maybe the header blew out on the board?
If its working in one port and not in the other, its either the current on the board is not supplying power to the LED, or its deactivated.

Since your not that electrically knowledged, its very difficult to tell you how to diag it.
Most likely would require a multi meter and testing the voltage at the LED header on the board directly, to see if there is current.

As for the PSU question, you probably kicked the fuse on it.
And it took some time to reset the fuse internally.

As for did you short something out? No one can answer this question.
You may or may not have.
The only way to find out is to test the components and see if you have any problems with them.

 

[plopke]

The board header still works and is still giving around 5V , connecting it to the case old power LED works and a other push button(also 12vDC LED-power supply) also works.


The one I installed is shown as
LED - power supply: 12V DC
max. switching load: 5A / 250V

It works on a 12V Molex line just not any more on the motherboard 5V which just feels odd to me.

 

[aigomorla]

well i have those exact switchs also.
they are made by lamptron, and i remember them always being 12V and not 5V.

:\

if u say they worked on 5V, are you absolutely sure it was 5V?
the 5V line on the molex should look something like this:

its very easy to confuse for the 12V rail.

 

[plopke]

measured the voltage on the motherboard frontpanel led connector which was around 4-5V.

The led is working on 12V of the molex , if I put it on a 5V molex , it works only on the closest one to the power supply , so I guess any little dip in voltage is enough for the LED to not work.


On a fun side note , bought 2 sharkoon molex adapters (one 12V split off , other 5V), which where both color coded wrong .... (gets me kinda grumpy lol)

Example of a 5V split off ,....

Now if I understand this right , this is wrong color coded?

Lookingat MALE side with 90 degree corners upwards , that should normally be 5 V on the right , and they do sell it as a 5V split off but yellow is used,.... My antec power supply was color coded the other way around ,...

 

[aigomorla]

yellow + brown = 12V
its always been that voltage spec.

 

[BonzaiDuck]

I have a similar chore before me, and found it fortuitous that this thread was initiated by Plopke.

I have an old CoolerMaster Stacker 830 case, and I'm fixing it up for more contemporary usage. For instance, I discovered that I should be able to install a 280mm radiator in front if I merely modify or change out the bundled drive cages. Or -- what I've currently done -- I can use 140mm fans as opposed to the 120's intended in the CM Stacker design.

I expect that with a case this old, there won't be any eSATA ports on the case front-panel. So I fixed that with some art-board construction, a 3.5" front-panel cover-plate from the 1990s, and an SATA/eSATA converter plug I bought for $7. I installed the eSATA ports with this art-board construction in the recess at top-front of the case. The recess had originally been intended as a case "carrying-handle."

Because of the fans, I can only manage to use two 5.25" drive bays at the top of the case-front. So I installed the DVD-burner and a 3.5" 2-port USB 3.0 front-panel device.

No problem, but there's an "LED" dimension to this.

The original Power-LED of the Stacker is part of the power-switch assembly, and to replace it, I have to remove the metal box in the case top-front which includes the USB, 1394 and audio connections. I figured instead that I'd simply install a new LED connected directly to the mobo pinouts.

I have small cello-packages of LEDs I'd purchased at an electronics warehouse some years ago. They are all apparently marked "5V," but I'll have to test them for "color."

So -- thanks for the thread. It gives me more "insight" while I try and figure out how to attach the flexible extension I bought for my Dremel.

 

[Paperdoc]

A LED typically operates properly on 2 to 5 VDC, with blue ones being often in the 4 to 5 VDC range. Each also has its own proper current flow when supplied with the correct voltage - often blue LEDS's are in the 10 to 12 mA range. So circuits that use them (including your Phobya pushbutton switch) include a dropping resistor in series with the LED. For example, IF your LED was specified to operate properly at 4 VDC drawing 12 mA, the series resistor would have to reduce the voltage from 12 VDC (supply) to 4 VDC across the LED only with that 12 mA current flowing. So the resistor could be 8/0.012 = 667 Ohms with a power rating of 0.1W, and it would be built into the switch case.

Given that circuitry inside your switch, supplying it with only 5 VDC means the LED MIGHT light up weakly, or it might not light up at all. The LED is a diode. Among other things, that means that it allows current to flow through it (and thus generate light) as long as the voltage across it is above a certain minimum value, but for a lower voltage the current flow (and light output) is virtually nothing. So with whatever mysterious aging effects have happened, your under-supplied LED finally has got to NOT allowing enough current to flow to generate any light.

Now, the Phobya unit you linked has six contacts on it. Two obviously are for voltage to the LEDs. I suspect you have these two contacts connected to the "PLED" pins of your mobo's Front Panel header to supply the 5VDC to the LED. Of the other four (two pairs), I cannot read German but I suspect one pair is for a "normally open temporary contact" switch. That is, with no external action, the switch contacts are open, but when you push the button the switch contacts close and stay connected as long as you keep pushing; when you release, the switch contacts open again. Those are the contact pair you would have used for your front panel's On / Off connections. The other contact pair, I suspect, is the opposite action: "normally closed temporary", so that those contacts are closed until you push the button and open them. If I'm right about that, there is no simple way I can suggest for you to make this switch and its LED work the way you want. The closest thing I can suggest is that you could buy a LED designed to operate at 5 VDC, mount it in your front panel, and connect that to the PLED pins of the mobo header.

Your tale of the "short circuit" probably is nothing to worry about, and I'm sure it did NOT blow a fuse - if it had, it would not have started to work again. Here is how the On / Off function of that pushbutton switch works. The supply to the switch is 5 VDC, but little current, and the switch is open normally. The PSU actually is "on" for very limited functions at all times. When you push the button it temporarily shorts out those particular the mobo header pins and that is the signal to the mobo to start up everything. At some later point a brief push would do nothing. But Windows has a special setting for another function of this same pushbutton, and it is set by default to allow you to force a shut-down if necessary. If you push the button and HOLD it in for at least 4 to 5 seconds, the mobo will shut down immediately, not an orderly shut-down by Windows. So that is what a "short circuit" of the 5VDC signal to the On / Off pushbutton will do -force an immediate shut down, which is what you saw. Now, it the short circuit condition were maintained for a while after that might cause some self-protection feature in the system to block a quick re-start. But as you've seen, it corrected itself and it all works again. This incident has nothing to do with the failure of the LED to light up.

Regarding colors on 4-pin Molex output connectors from a PSU, here is the standard way:

http://pinouts.ru/Power/BigPower_pinout.shtml

If you look at the FEMALE connector from the PSU so that the open holes are toward you and the beveled two corners are at the TOP, then Pin #1 is on the LEFT. The connections are:
1. YELLOW +12 VDC
2. BLACK Ground
3. BLACK Ground (also)
4. RED +5VDC

Note that for fans and some other parts it is common to used Red as + and Black as Ground, but the Molex output does not put out the 12 VDC supply on Red. On that output, Red is 5 VDC, and Yellow is 12 VDC.

 

[BonzaiDuck]

Paperdoc: "You're the man."

My understanding of this, from a retired elec-tech friend, is that a circuit with excess voltage for these LEDs without the appropriate resistor(?) would eventually damage the LED.

I wanted some red LEDs for drive controller activity lights, went to the local electronics-warehouse, and casually asked for them to "work with PC motherboard pinouts for LEds" (which probably wouldn't differ from the pinouts on a controller.) Got red ones, and they seem to be working fine.

I suppose that "I'm good." Comment, if you wish.

A LED typically operates properly on 2 to 5 VDC, with blue ones being often in the 4 to 5 VDC range. Each also has its own proper current flow when supplied with the correct voltage - often blue LEDS's are in the 10 to 12 mA range. So circuits that use them (including your Phobya pushbutton switch) include a dropping resistor in series with the LED. For example, IF your LED was specified to operate properly at 4 VDC drawing 12 mA, the series resistor would have to reduce the voltage from 12 VDC (supply) to 4 VDC across the LED only with that 12 mA current flowing. So the resistor could be 8/0.012 = 667 Ohms with a power rating of 0.1W, and it would be built into the switch case.

Given that circuitry inside your switch, supplying it with only 5 VDC means the LED MIGHT light up weakly, or it might not light up at all. The LED is a diode. Among other things, that means that it allows current to flow through it (and thus generate light) as long as the voltage across it is above a certain minimum value, but for a lower voltage the current flow (and light output) is virtually nothing. So with whatever mysterious aging effects have happened, your under-supplied LED finally has got to NOT allowing enough current to flow to generate any light.

Now, the Phobya unit you linked has six contacts on it. Two obviously are for voltage to the LEDs. I suspect you have these two contacts connected to the "PLED" pins of your mobo's Front Panel header to supply the 5VDC to the LED. Of the other four (two pairs), I cannot read German but I suspect one pair is for a "normally open temporary contact" switch. That is, with no external action, the switch contacts are open, but when you push the button the switch contacts close and stay connected as long as you keep pushing; when you release, the switch contacts open again. Those are the contact pair you would have used for your front panel's On / Off connections. The other contact pair, I suspect, is the opposite action: "normally closed temporary", so that those contacts are closed until you push the button and open them. If I'm right about that, there is no simple way I can suggest for you to make this switch and its LED work the way you want. The closest thing I can suggest is that you could buy a LED designed to operate at 5 VDC, mount it in your front panel, and connect that to the PLED pins of the mobo header.

Your tale of the "short circuit" probably is nothing to worry about, and I'm sure it did NOT blow a fuse - if it had, it would not have started to work again. Here is how the On / Off function of that pushbutton switch works. The supply to the switch is 5 VDC, but little current, and the switch is open normally. The PSU actually is "on" for very limited functions at all times. When you push the button it temporarily shorts out those particular the mobo header pins and that is the signal to the mobo to start up everything. At some later point a brief push would do nothing. But Windows has a special setting for another function of this same pushbutton, and it is set by default to allow you to force a shut-down if necessary. If you push the button and HOLD it in for at least 4 to 5 seconds, the mobo will shut down immediately, not an orderly shut-down by Windows. So that is what a "short circuit" of the 5VDC signal to the On / Off pushbutton will do -force an immediate shut down, which is what you saw. Now, it the short circuit condition were maintained for a while after that might cause some self-protection feature in the system to block a quick re-start. But as you've seen, it corrected itself and it all works again. This incident has nothing to do with the failure of the LED to light up.

Regarding colors on 4-pin Molex output connectors from a PSU, here is the standard way:

http://pinouts.ru/Power/BigPower_pinout.shtml

If you look at the FEMALE connector from the PSU so that the open holes are toward you and the beveled two corners are at the TOP, then Pin #1 is on the LEFT. The connections are:
1. YELLOW +12 VDC
2. BLACK Ground
3. BLACK Ground (also)
4. RED +5VDC

Note that for fans and some other parts it is common to used Red as + and Black as Ground, but the Molex output does not put out the 12 VDC supply on Red. On that output, Red is 5 VDC, and Yellow is 12 VDC.

 

[Paperdoc]

Yes, overvolting LEDs could damage them. I suspect the time it takes for them to fail depends on how large the excess voltage is. Basically too high a voltage would force too large a current through the LED causing it to overheat and wear out faster.

What you did if very likely correct. I'm pretty sure that the outputs to case LED indicators from mobo are based on 5 VDC. After all, almost all of the logic circuits on a mobo are based on 5 VDC TTL logic components. I'm sure the electronic shop people know that and would have provided you with the right LED's. Come to think of it, since the plain red LED's normally operate at less than 5 VDC and the mobo makers know that, I expect the circuits sending power specifically to red LED indicators include their own small dropping resistors on the mobo. I suspect that if one were to be careful and connect a DC voltmeter across the leads of a lit common red LED in the case, it would show the voltage across the LED around 2 VDC.

 

[plopke]

Was on a holiday but very interesting read , thx for the information.

 

[BonzaiDuck]

Yes, overvolting LEDs could damage them. I suspect the time it takes for them to fail depends on how large the excess voltage is. Basically too high a voltage would force too large a current through the LED causing it to overheat and wear out faster.

What you did if very likely correct. I'm pretty sure that the outputs to case LED indicators from mobo are based on 5 VDC. After all, almost all of the logic circuits on a mobo are based on 5 VDC TTL logic components. I'm sure the electronic shop people know that and would have provided you with the right LED's. Come to think of it, since the plain red LED's normally operate at less than 5 VDC and the mobo makers know that, I expect the circuits sending power specifically to red LED indicators include their own small dropping resistors on the mobo. I suspect that if one were to be careful and connect a DC voltmeter across the leads of a lit common red LED in the case, it would show the voltage across the LED around 2 VDC.

Well, actually -- maybe I said earlier -- I have a collection of wires and LEDs extracted from computer recycle discards. One of them was green, but some of these old computer front-panel LEDs used colored plastic. That one went to the motherboard replacing the blue LED on the case push-button. The second LED was the red one I spoke of -- in clear plastic. That one is now cabled to this:

http://www.newegg.com/Product/Product.aspx?Item=N82E16816101792

It has two banks of 4x pinouts for "activity" and "fault," one for each SFF-8087 port. So I hooked up to the activity pinouts for the port connected to what is so far two HDDs.

Very Christmas-y. What enthusiast doesn't like a few LEDs and flashes?