The Cellar  

Go Back   The Cellar > Main > Technology
FAQ Community Calendar Today's Posts Search

Technology Computing, programming, science, electronics, telecommunications, etc.

Reply
 
Thread Tools Display Modes
Old 04-03-2012, 09:25 AM   #16
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
Quote:
Originally Posted by tw View Post
So far, good. The black probe could also be connected to the chassis.

Now follow those directions. Measure the green and gray wires both before and when the power switch is pressed. Note the behavior on each wire as the power switch is pressed.

Obviously the system will not load and execute tasks. Whereas that third step is valid for BrianR, it is different for you. What does any one yellow, orange, and red wire do when the power switch is pressed? For example, do each start to rise and then fall? Does any one not move? Or do all not do anything?

Diagnostic LEDs are not inside. From one Dell service manual: "your computer is equipped with four lights on the back panel labeled "A," "B," "C," and "D." These lights can be yellow or green. When the computer starts normally, the lights flash. After the computer starts, the lights remain green. If the computer malfunctions, ..."

To say more unique to your system requires the Service Tag number. Often a label somewhere on the side or back of a Dell Desktop. And often found on the underside of laptops.
Not sure how I'm going to test the voltages plugged in yet, I guess I'd have to contact on the solder ends on the other side of the board, whihc would mean a more major dissassembly job. I'll put some more thought into this.

I don't think there are external indicator LEDs on this model, maybe older or higher end models, think of the cents they save by not including them.

I see some models used the keyboard caps lock, num lock light as indicators, but not mentioned in the manual for this one.

At this base model it's barely worth changing anything but the power supply, certainly not to pay a professional to do it.

Dell Service code: JJ8FJG1
Beest is offline   Reply With Quote
Old 04-03-2012, 01:51 PM   #17
Undertoad
Radical Centrist
 
Join Date: Jan 2001
Location: Cottage of Prussia
Posts: 31,423
Maybe the notification LED is the power switch itself. It means different things if the lamp in the switch is off, blinking or steady amber, blinking or steady blue.

http://support.dell.com/support/edoc...g/MT480A01.pdf

Quote:
Blue light — Blinking blue indicates that
the computer is in sleep state; solid blue
indicates power-on state.
Amber light — Blinking amber indicates a
problem with the system board. A solid
amber light when the system does not
boot indicates that the system board
cannot start initialization. This could be
an issue with the system board or power
supply (see "Power Problems" on
page 35).
Undertoad is offline   Reply With Quote
Old 04-03-2012, 06:09 PM   #18
monster
I hear them call the tide
 
Join Date: Dec 2005
Location: Perpetual Chaos
Posts: 30,852
No light.......
__________________
The most difficult thing is the decision to act, the rest is merely tenacity Amelia Earhart
monster is offline   Reply With Quote
Old 04-03-2012, 06:41 PM   #19
tw
Read? I only know how to write.
 
Join Date: Jan 2001
Posts: 11,933
Quote:
Originally Posted by Beest View Post
Not sure how I'm going to test the voltages plugged in yet, I guess I'd have to contact on the solder ends on the other side of the board, whihc would mean a more major dissassembly job. I'll put some more thought into this.
The previous post said:
Quote:
Measure a purple wire from PSU where it connects to the motherboard. It should read about 5 VDC.
How did you measure the purple wire (5.02 volts)? Only one place to measure it. Push the meter probe into that nylon connector to touch a wire inside that connector. Do same for all other measurements.

Again, if not clear. Disconnect, remove, or disassemble nothing (other than the cover plate). Disconnecting or removing can make a problem more complex. If also not stated earlier, disconnecting sometimes can create more damage.

The entire task should take as little as 60 seconds. Touch the probe to each wire where it connects to the motherboard. Record those numbers and behavior. Post numbers. That's it. Most time should be spent posting those numbers.

Was this Inspiron 518 purchased in Best Buy? If those critically important Diagnostic lights are not on the front panel, then that MBA attitude says why. Best prices for a Dell means purchasing direct from Dell. Have seen too many products (even HP printers) missing essential functions to meet Best Buy's 'increase our profits at the expense of the customer' attitude.

From the Technology Guide for your machine:
Quote:
Some desktop computers have lights labeled with numbers on the front panel. When the computer starts normally, the lights flash, and then turn off. If the computer malfunctions, the sequence of the lights can be used to help
identify the problem.
BTW, two updates have been available for Inspiron 518 after you purchased it in Mar 2009. One was a recommended update for the video display. The other was a critically important updatda for serial ATA. Enter that Service Tag into the Dell website (under Product Support) to learn more.
tw is offline   Reply With Quote
Old 04-04-2012, 11:27 AM   #20
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
Quote:
Originally Posted by Undertoad View Post
Maybe the notification LED is the power switch itself. It means different things if the lamp in the switch is off, blinking or steady amber, blinking or steady blue.

http://support.dell.com/support/edoc...g/MT480A01.pdf
Yes, the power button is an indicator light, blue solid or flashing for asleep or awake, never seen it be orange, as said now nothing.

System was purchase direct from Dell.

I measured the voltages I did by pulling the plug and putting the probes in that end, I'll see if I can slip them down the side to get a reading while still plugged in.

The only other light on the front in normal operation is the drive light.
Beest is offline   Reply With Quote
Old 04-05-2012, 12:31 AM   #21
tw
Read? I only know how to write.
 
Join Date: Jan 2001
Posts: 11,933
Quote:
Originally Posted by Beest View Post
I measured the voltages I did by pulling the plug and putting the probes in that end, I'll see if I can slip them down the side to get a reading while still plugged in.
Don't slip a probe. Saying again. Jam that probe inside the connector hole for each wire. Disconnect, disassemble, or remove nothing. Do not even move a connector. Jam that probe in with aggression. If the probe is bigger than the hole, then wrap a paper clip onto the probe and stick the paper clip end into the hole. But, if not clear before, absolutely do not move, partially move, disconnect, disassemble, or remove anything (but the already removed cover plate).
tw is offline   Reply With Quote
Old 04-05-2012, 08:49 AM   #22
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
That was pretty straight forwardm ny $4 Harbor freight meter has noce pointy probes.
So with everything plugged in as in the wroking state

All the red wires I measured were 0.01V power button on or off,

all the single orange wires were 0.03V power button on/off

There is a single socket with both orange and brown wires in it, this read 0,01 power button on/off.

The grey wire was 0V power button on/off

The purple wire was 3.38 V but dropped to 3.37 when the power button was pressed.


The green wire, the first time I measures it at 2.23V power on or off, when I measured it again after looking at the others it was 3.21 then dropped to 3.19 when the power button was pressed

about an hour later, with the system left plugged in while I was on a call it is reading 3.24 power on/off, this small difference could well be variation in my chepa meter though I guess.
Beest is offline   Reply With Quote
Old 04-05-2012, 12:07 PM   #23
tw
Read? I only know how to write.
 
Join Date: Jan 2001
Posts: 11,933
Quote:
Originally Posted by Beest View Post
The purple wire was 3.38 V but dropped to 3.37 when the power button was pressed.

The green wire, the first time I measures it at 2.23V power on or off, when I measured it again after looking at the others it was 3.21 then dropped to 3.19 when the power button was pressed
OK. First, a PSU is actually multiple power supplies. Purple wire connects a special voltage constantly and only to certain motherboard functions. That voltage is why all computer work must be performed with the AC power cord disconnected. Otherwise damage can result. That LED on the motherboard was also making the same warning. Only work on a computer when that LED eventually dims out.

Purple wire voltage (5VSB) is defective. Therefore a power controller cannot power on the system. However, the question is why? Is the motherboard shorted or is the PSU purple wire power supply defective?

Well, you measured the purple wire when disconnected from the motherboard at 5.02 volts. That only says it may or may not be defective.

Subsystems powered by the purple wire. Anything that might 'wake up' the computer including the keyboard, modem, and NIC. Also the Front I/O Panel is powered by the purple wire. It can be temporary disconnected where it connects to the motherboard.

In your BIOS setup would be a more complete list of those devices that can wakeup the machine. So, with AC power cord disconnected, remove any peripheral (ie USB device, modem card, keyboard, etc) that might be excessively loading the purple wire 5VSB. Reconnect the PSU to AC mains. See if that purple wire voltage rises up to about 5 volts. If yes, one of those removed devices might be defective.

We clearly see the defect. 5VSB (purple wire) well below 5 volts means the power controller cannot tell its power supply to power on. And cannot even let the CPU execute (power controller tells the CPU when it can work).

The green wire is the power controller telling the PSU to power on. But your numbers on the green wire confirm defective numbers from the purple wire. Therefore the green wire cannot go to near zero volts when the power button is pressed. Therefore the main PSU power supply is never told to power on.

Some will recommend jumpering the green wire to any black wire with a paper clip. In your case, the PSU will probably power on. So they will conclude the PSU is good. They would be confused. The meter says 5VSB - the other supply - has failed. Their 'paper clip' test would never see the defect. An example of why so many get confused when using observation rather than using a meter. An example of why the meter so quickly cuts through confusion.

Now, moving on to fixing it. The fault could be on the motherboard side or inside the PSU. If removing those other devices does not restore 5VSB. Then let's try any other test. After removing the AC power cord, then disconnect all PSU power connections to the system. Connect another PSU to that 24 pin connector (where you made measurements). Any PSU (even an undersized one) that has a similar 24 pin connector is sufficient. The second PSU need not even be in the system (if its wire is long enough to connect to the 24 pins on the motherboard).

Connect that second PSU to AC mains. Measure the purple wire voltage. It must be greater than 4.87 volts. If yes, then the first PSU has a defective 5VSB. If no, then the defect is somewhere else in your system (probably on the motherboard).

A second possible test. Again with the PSU disconnected from the motherboard. Buy some 100 ohm resistors from Radio Shack (maybe $1). Stuff maybe all four resistor leads into the purple wire connection hole. Connect the other end of those resistors to any black wire hole. Connect the PSU to AC mains. Measure the purple wire to black wire voltage. If not 5.02 (as before), then the 5VSB supply is defective - without doubt. If it is about 5volts, then the 5VSB supply 'might' be good. This 'load test' is a best option.

Or a third option. Buy a new supply. Connect it. Confirm the new PSU can provide sufficient purple wire 5 volts. Let me know if you choose this third option so that details can be provided.

Apparently the main supply inside the PSU is good. But a tiny 5VSB supply has either failed or has been shorted out by something on or attached to the motherboard. (Shorting out any supply does not harm that supply.) Above should help define which of two is defective. And says why the blue front panel light stays on constant.

Appreciate how much information was in those few numbers.
tw is offline   Reply With Quote
Old 04-10-2012, 10:45 PM   #24
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
No external devices have been connected, keyboard etc.

So I disconnected everything I could from the motherboard, optical drive, disk drive, video board, memory card etc.

Plugged in mains, green wire tested 2.23 and purple wire 2.34, no lights on the front power button.

Plugged all those peripherals back in, voltages still the same.

I don't have a 24 pin PSU too try, i do have a 20 pin, the green purple and grey wires are all in the same place, almost all of the other wires are the same. one of the red wires on the 24 is white on the 20.

I saw the paper clip test early in investigations, it does make the fans power on.


I found some 600 ohm and 1K Ohm resistors lying around and tried plugging them into the purple wire, also 2 in parallel, no voltage drop.

I look for some 100's

Last edited by Beest; 04-10-2012 at 11:11 PM.
Beest is offline   Reply With Quote
Old 04-10-2012, 11:49 PM   #25
tw
Read? I only know how to write.
 
Join Date: Jan 2001
Posts: 11,933
Quote:
Originally Posted by Beest View Post
I found some 600 ohm and 1K Ohm resistors lying around and tried plugging them into the purple wire, also 2 in parallel, no voltage drop.
As you may have already seen, a defective supply can act good when not connected to a load. That is why the paper clip test says so little useful. That is why power supply testers are almost useless. That is why a power supply is best measured by not disconnecting a single wire.

The 5VSB should output maybe 2 amps (its label may say more). A full load is 5/2 ohms or 2.5 ohms. Radio Shack may still sell 10 ohm 10 watt ceramic resistors. One of those might be easier to use than four 100 ohms in parallel (25 ohms).

5 volts divided by 10 ohms is a half amp. If the 5VSB is defective, then a half amp load should cause some voltage reduction.

If those 10 ohm resistors are sold in pairs, then two ten ohm resistors means a full 1 amp load. A great test of the 5VSB. Those two 10 ohm resistors (that will get hot when held) should not cause the 5VSB to drop. If 5VSB does drop, the PSU's 5VSB (and not anything in the PC) is defective. That is a definitive and "you have found the problem" conclusion.

5 volts divided by 25 ohms (four 100 ohms in parallel) is 200 ma. Not a great load. But enough that can be purchased cheaply. The ten ohm ceramic would be better but I don't know if it remains available.

600 ohm or 1k ohm resistors are only 10 milliamps (only one LED) or 5 milliamps (less than 1 LED). Virtually a near zero load.

If the PSU 5VSB does not drop with that load, then the problem (excessive load) is probably located on the motherboard.
tw is offline   Reply With Quote
Old 04-11-2012, 07:22 AM   #26
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
OK,got it.

Thanks for all your help with this.
Beest is offline   Reply With Quote
Old 04-16-2012, 06:12 PM   #27
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
Quote:
Originally Posted by tw View Post
Radio Shack may still sell 10 ohm 10 watt ceramic resistors. One of those might be easier to use than four 100 ohms in parallel (25 ohms).

5 volts divided by 10 ohms is a half amp. If the 5VSB is defective, then a half amp load should cause some voltage reduction.

If those 10 ohm resistors are sold in pairs, then two ten ohm resistors means a full 1 amp load. A great test of the 5VSB. Those two 10 ohm resistors (that will get hot when held) should not cause the 5VSB to drop. If 5VSB does drop, the PSU's 5VSB (and not anything in the PC) is defective. That is a definitive and "you have found the problem" conclusion.

.

Yes they do.
Purple to black
no load 5.02V
10 ohm resistor 1.26 V
2x 10 ohm in parallel 0.82 V

The green light on the back of the power supply dims out when the resistors are connected.

I tried the same test on another PSU I have borrowed (20 pin connector), minimal voltage drop with this test.

So is that definitive on the PSU ?
Beest is offline   Reply With Quote
Old 04-16-2012, 06:28 PM   #28
zippyt
LONG LIVE KING ZIPPY! per Feetz
 
Join Date: Mar 2003
Location: Arkansas
Posts: 7,661
So its the power supply like i initially said ???
__________________
"Success is getting what you want. Happiness is wanting what you get. "
Brother Dave Gardner
zippyt is offline   Reply With Quote
Old 04-16-2012, 08:17 PM   #29
Beest
Adapt and Survive
 
Join Date: Jan 2007
Location: Ann Arbor, Mi
Posts: 957
Quote:
Originally Posted by zippyt View Post
So its the power supply like i initially said ???
LOL, yeah , thought you'd get akick out of that.

Beest is offline   Reply With Quote
Old 04-16-2012, 09:07 PM   #30
classicman
barely disguised asshole, keeper of all that is holy.
 
Join Date: Nov 2007
Posts: 23,401
lol
__________________
"like strapping a pillow on a bull in a china shop" Bullitt
classicman is offline   Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

All times are GMT -5. The time now is 03:47 PM.


Powered by: vBulletin Version 3.8.1
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.