PDA

View Full Version : No Power on First Boot



geojo2
08-01-2009, 01:14 PM
My Computer specs.
Case: Antec Sonata 3
Power Supply: 500W Antec
Intel Core2 Quad 9550 2.83GHs LGA 775 95W
Gigabite EP45-UD3P Motherboard
G Skill 4GB (2X2 GB) 240 PIN DDR2 SDRAM 1066
WD 1TB HD
Zalman CNP 9500 Cooling Fan
GeForce GTX 260 Video Card (temp removed)

I have a problem. My computer did not boot up on its first boot. I have no power, no lights and no fans or anything.

I have the 24 pin power supply connected to the Mother Board. There is also another 8 pin connector power supply which is connected. I went over the connections to the front of the case and all looks OK + and - etc.

When I go to power-up and push the front switch, the led lights on the mobo flash on(the top 2 in red and the rest green. Then I have to reset the power for the lights to go on again.

I would really appreciate any help!!!!!

zburns
08-01-2009, 02:23 PM
If your earlier comments about looking for a mobo ground, and the 2 prong wire from the GTX have not been resolved, is there a chance you hooked a non-ground wire to a ground? This would result in the PSU acting in a "self preservation" manner by shutting down.

It sounds like you grounded something which in turn grounded out the mobo and the PSU is acting normal!

If you are still having the no boot problem, pull out the video card and just see if your lights come on and stay and the fans run.

Do not disconnect any of the fans.

Then disconnect SAtA and IDE other external devices, HD, Opt drive. Last pull our each front panel connectors to the mobo one at a time. Do one thing at a time until hopefully the problem clears.

Did you use the Antec supplied spacers and screws for the mobo mounting. Did you have to force anything to get the back panel connectors to go thru the Antec template in the back?

No washers under screw heads on the top of the mobo? Correct?

geojo2
08-01-2009, 04:19 PM
:mad:I had the video card out of the pc. I will check out everything you said, If that does not work I will remove the Mobo, the cooler and the processer and check to seen if I put too much Artic silver between the processor and the cooler. When I installed the cooler it felt like I was putting at of pressure on the processor. This time I'll try starting it up on the bench and see what happens. I think the vidio card problem was solved thru EVGA.com but I won't be sure until I can get this up and running.

Thank You

zburns
08-01-2009, 05:39 PM
If it "oozed out" and ran down the sides of the CPU chip, that could cause a problem since the Artic Silver has metallic silver in it and would act as an electrical conductor.

When you purchased it, you should have got some Artic Silver cleaner or solvent with it. Use a mag glass and make sure you get it all off all side surfaces. When reapplying a pea size blob about 3/16" for an AMD dual core (my chip).

I do not recommend trying it on a bench because of static hazard. When you mount it in the case, that is probably the grounding source for the mobo; you would have to simulate this outside of the case which may be confusing (or not) plus you have to use wrist strap static control and ground the strap to the mobo, but the mobo should be grounded first.

Just be very sure you cleaned perfectly any artic silver off any "side" that it ran onto!!

geojo2
08-01-2009, 09:20 PM
Right now the mobo is out of the computer and I jumpered the power supply to the computer case button,
cleaned up the processor,and reinstalled the cpu fan. All I have connected is the 24 pin supply, the 8 pin supply, the SonataIII case fan and power supply fan which are connected together and then connected to the system fan. The cpu fan is connected to the cpu fan pins on the mobo. I have a feeling it has something to do with the Sonata III case power buttons, BUT WHAT, I do not know..If nothing gets solved by Monday, I will call the Antec support..

BUT STILL NOTHING

zburns
08-01-2009, 10:01 PM
Do your led lights still flash on momentarily or does nothing light up even for a brief moment? Providing the same problem is showing itself the same way, means nothing has changed which is good for now. If the lights don't even flash, that means something else has changed. So do they behave as they did initially (as in your first post).

I have the same case; I can take a look at it (the wiring) in the morning.

This is a side question. If it is possible, look at your mobo front and pack where the case mounting screws go thru and into the spacers. If there is only green circuit board showing around each "used" mounting hole, front and back, this would indicate there is no grounding via the screw or spacer mounts, which is probably the way it should be. I am reaching here; do not read anything into this request. I have no idea whether this is relevant or not; but since the PSU looks like it is shutting down quickly, a grounding of a voltage point would do that.

zburns
08-02-2009, 10:15 AM
Wise Monkey, please confirm or not what I am saying in this post. I have "sort of" confirmed it using "google" but I do not want to contribute to "geojo2's" problem.

Point # 1: Without knowing for certain, I would assume at least one metal screw in standoff would purposely sit on a ground plane lan on the mobo when mounting the mobo. The mounting screw goes thru the mobo and pulls the standoff tight against the ground plane lan on the underside of the mobo. The mobo in an ATX case is grounded to the case this way.

If you pull the mobo out of the case but still hook the mobo to the PSU while in the case, you must ground the mobo to the case with a single wire to the hole and the ground plane lan where the metal standoff would normally go.

For static control you hook the cable from the wrist strap to this single ground point where the case wire connects to the mobo.

Here is a link from 2006 that confirms part of the above for ATX: http://biorobots.cwru.edu/server/howto/buildcomp/mountmobo/ There is a Tom's Hardware Link three posts down that is a better explanation than this link.

Point # 2: I think it would be ok to remove the CPU and fan from the mobo and still turn on power while leaving the fan hooked to the mobo, but the fan just laying loose on the bench. The reason to do this is to see if the green led that verifies PSU voltages present will light up and stay lit, thereby, checking out whether the CPU is partially or grossly shorting out the PSU.

Say it another way: Can you take the mobo out of the box and power it up with the PSU 24 and 8 pin connector with no added components, ie RAM, CPU, Video Card, HD, etc.. You could plug in CPU fans. Again a "check" to see an OK mobo w/o installed components, one of which may cause a short to the PSU.

That's it. I am just looking for a test to get the green led on full time with only mobo powered up with no components.

geojo2
08-02-2009, 10:23 AM
I would like to call antec(the Company who made the case and power supply and make sure I have the leads connected from the case the right way before I go any further. The way antec has the leads labeled and the way gigabite has them labeled are alittle conflicting.. I have six led lights. When I turn the power on at the back of the case and hit the power button in the front, the top 2 lights flash red and bottom 4 are green and go out. To get them to flash on again I have to turn the back switch off, and then on.
My mobo shows white circles around all the mountings for the screws. Right now my motherboard is connected, sitting on some bubble wrap outside the case.

zburns
08-02-2009, 11:11 AM
I have six led lights. When I turn the power on at the back of the case and hit the power button in the front, the top 2 lights flash red and bottom 4 are green and go out. To get them to flash on again I have to turn the back switch off, and then on.
My mobo shows white circles around all the mountings for the screws. Right now my motherboard is connected, sitting on some bubble wrap outside the case.
___________________

I am just commenting here, not asking or recommending you do anything; I hope Wise Monkey will jump in and confirm and add to, perhaps, what I say; this course of action so that I am not giving you wrong or misleading info.

To get your attention, I would say you potentially have a "static discharge" disaster awaiting you, conditions being correct. If your mobo when installed in the case is grounded to the case by one metal standoff delibertly connecting with a mobo circuit board lan, then you must connect that same "mobo grounding point" with screw, nut, and a wire back to the case attached by a screw or standoff to the case to simulate the condition when the mobo is installed in the case -- only one wire.

Side comment here. Your bubble wrap is an insulator; any static charge built up on the mobo will not necessarily disapate thru the power connection.

Your wrist strap static control device should still be hooked to the case if you want to touch the mobo with power on -- But you have to exert a lot of CAUTION if you do this. You only use the hand, the wrist strap is attached to.
THIS IS AN ELECTROCUTION CAUTIONARY WARNING!

NEVER USE BOTH HANDS ON YOUR SET UP WHEN IT IS HOT!! NEVER TOUCH THE MOBO ON THE BUBBLE WRAP WITH ONE HAND AND THE CASE WITH THE OTHER HAND WHEN THE MOBO IS TURNED ON AND OUT OF THE CASE. EVEN IF YOU TOUCH SOMETHING WITH ONE HAND, DO IT WITH A FINGER, DO NOT "GRIP WITH A WHOLE HAND"

THE ABOVE PRECEEDING STATEMENT IS AN ELECTROCUTION CAUTIONARY WARNING!!

Okay, what I started out trying to say, is to set a senario where you test the mobo in the case but without the CPU installed (but you are doing it outside the case which may work but I do not recommend it. Too dangerous, potentially, for a novice. I would keep things in the case.)It is my understanding that at least one green led will light indicating the PSU voltages are ok -- your mobo inst. book will tell you which one, maybe. If the CPU is shorted due to Artic Silver contamination, say where the pins are, then with no CPU installed, the short is removed, etc. or for example, the CPU is shorted. The point being your lights come on and go off; but with the CPU out, and the correct green led comes on and stays on, it means you removed the short.

Assuming all this was correct and you were working with the mobo in the case, you could then add components one at a time, power up (still with static control and CAUTION), and track down the cause of the short, etc..

Summarizing: I have only built one computer; I think you could power up the mobo w/o anything installed to it and not cause any harm, but I just have not done it; I do not want to suggest to you that you do it without some confirmation of my recommendation. Wise Monkey has tons of experience, and I would trust his answer as to whether you can power up the mobo with no components installed and not cause harm. You or anyone can probably google this senario and find the answer; I briefly tried this but found too much extraneous discussion that made it difficult to sort thru.

I hate to through in the ELECTROCUTION WARNING but doing what you ( or anyone, for that matter) are doing is potentially very dangerous if you use both hands and work on a hot setup such as you have set up. I can from personal experience tell you a horror story that directly relates.

zburns
08-02-2009, 03:26 PM
Here is a Tom's Hardware link that explains the mounting of the mobo, brass standoffs and there purpose plus some explanation of "ground plane" within the mother board construction. They also tell you why the ground plane is necessary. Page 5 of there multi page instruction on PC assembly. Link as follows: http://www.tomshardware.com/reviews/how-to-build-a-pc-part-3,1382-5.html

This is me talking, my explanation. The motherboard is constructed of multiple layers of "circuit board fibreglass" . Internally one of these layers is either a solid conductive layer (like a layer of tin foil) inside the flat motherboard; it is either the size of the motherboard or broken up into smaller rectangles or shapes thruout the motherboard surface.

The screws used to mount the motherboard to the spacers go thru a hole (some do, maybe not all) that has a silver (conductive) ring around the hole. This silver conductive ring is connected to the internal ground plane, therefore, the connection to the screw head and, therefore, the spacer. Since it is a ground plane or multiple pieces thereof, they are all connected, probably. To be sure a continuity meter would confirm by checking one "silver ring" to another, etc.

The point is that this ground plane is connected to the case when mounted in the case. You have your mobo out in the air with the PSU 24 pin and 8 pin cable connected. These two cables have ground connections in them but they do not connect to the ground plane I just mentioned from the article. Technically, with your mobo out in the air, you should hook up a connection between Antec case ground and the ground plane; it may matter, then again it may not.

The PSU 24 and 8 pin have grounds on certain pins, however, these grounds are part of the power supply circuit for the 5 v, 12 v power supply voltages, etc.. Under no circumstances, even tho they are at ground potential, they should never be used for "case" ground or ground plane "ground". Again, the ground plane is connected to case ground via the mounting screws into the spacers.

The case ground and the ground plane are separate "grounds" from the PSU grounds; they might be connected at some point somewhere but the method of connection would be such that no current flows between the PSU grounds and the ground plane. By not allowing current flow between the "various grounds", this prevents "ground loops", the cause of hum in audio systems, and noise in the computer circuits. They are connected at some point so that the electrical potential (voltage) of both grounds is always the same meaning zero, no voltage.

Recommendation: The simplest thing to do to be certain you are not causing a problem by having the mobo outside the case to work on it is to put it back in the case.

zburns
08-02-2009, 03:57 PM
This post will be about my Antec front cables if I can find my Antec book.

I am starting it now to mention something you can do if you have a continuity meter. You can unplug the cables one at a time and check the front panel on / off push button as well as the reboot smaller button, above left, of the on / off button.

For example, the on / off button and for all cable checks unplug each cable connector from the mobo. Check each lead separately from the plug to any part of the case (metal ground of the case). Should be an open circuit, no continuity. As long as there is no continuity, these are not shorting the mobo or anything else to ground; do the same thing to all the plugs from wires or cables from the front panel. Check each pinout, one meter lead to case ground, the other meter lead to the pin you are checking -- should always be no continuity as long as the plug is removed from the mobo and the other end is connected to the front panel.

If you have continuity to ground at any point, that indicates a problem.

For the on / off and the reboot push buttons, stick both meter leads into the respective plugs, push the correct push button, and you should get a short circuit -- only these two pushbuttons on the front panel.

The Wise Monkey
08-02-2009, 04:12 PM
Have you tried a completely basic boot i.e. just one stick of RAM and the CPU?

The LEDs you are referring to are related to how much load is put on the CPU, so would just flash temporarily at system startup just to show they are working.

I guess that there is a problem with the power getting to the motherboard, either because of a problem with the front panel connectors, or an underlying issue with the motherboard.

Have you tried something as simple as reversing the way power button leads are attached? If they are the wrong way around, it won't affect your computer at all, it just won't turn on, but considering that is what is happening now, there is no downside. But it would be a good idea to check it out. :)

You could also try and short the connection between them using some normal wire, as this is what the power button would do anyway.

zburns
08-02-2009, 06:06 PM
Here is an interesting comment from the Tom's Hardware Link on the previous page, it is a direct quote.

"It's important to observe the motherboard's mounting hole positions and place a standoff in each corresponding tray location. A mistake made here could potentially damage the board, but the most likely result is just a system that simply refuses to "power on". "

I know everything was in the case at one time. Still I was suprised to read the above. I do not why this would happen but I am suprised at the writer's words "....the most likely result ... refuses to "power on".

geojo2
08-02-2009, 06:49 PM
Thanks for the info and saftey precautions. It is well appreciated. I'm going to put the mobo back in the case and do the cpu test tomorrow..

zburns
08-02-2009, 08:48 PM
I'm back!! I guess you had the mobo out of the cabinet since you cleaned the cpu of the first Artic Silver coating. So that means the author's comment from the Tom's Hardware article may be applicable. I hope you have used static control as suggested by Rob when you first built; the Tom's H,ware article is long and on the first page talks about the same kind of static control as does Rob. So please follow their advice as you put the mobo back in; you are minimizing the opportunities for mistakes by doing this. Good luck!!

geojo2
08-03-2009, 01:39 PM
I did a continuity test The power (shorted when the button was pressed and so did the reset button. The others were all clear. I called antec and he said there was nothing he could do about testing the power supply itself.. ***Do you know any tests for this??***
I am going to do the CPU Test and may that will just narrow it down to the mobo and the power supply. I would prefer to just send back 1 item to Newegg..

I have been booting with 1 stick mem and nothing else

zburns
08-03-2009, 03:31 PM
The following info may help you. I have an asus mobo and the Antec Sonata III or 500 case -- it is called both. In my asus mobo book it describes the "System power LED (2-pin PLED) as follows, exact quote:

"System power LED, This 2-pin connector is for the system power LED. Connect the chassis power LED cable to this connector. The system power LED lights up when you turn on the system power, and blinks when the system is in sleep mode."

Several things: (1) your mobo book should have similar language about this LED. (2) The sentence "Connect the chassis power LED cable to this connector." refers to a cable from the Antec front panel. (3) "System power" as used above refers to the incoming power plug to the back of the case, meaning the LED stays on as long as the power cord is plugged into a live outlet and the on / off switch on the back of the Antec case is in the "on" position.

Later on in my Antec book, it states that the purpose of this LED being on is to warn you that power is "live" inside your case and "that you should not unplug or plug in, remove any components etc. without shutting off power" -- this is an important warning, in that you can really damage on board (mobo)components should you not shut off power and mess around.

So if you turn on your rig, the lights momentarily come on and then go off, and you have a "power on LED" as I describe above, this would definitely mean that your PSU is "shutting itself down as a protective measure", is "malfunctioning" or "damaged".

It could also mean that the mobo has a problem, or the CPU, if the CPU is the only component installed, or the RAM if one or more sticks are installed.

I assume I am correct when I say all ATX mobo's must have the 'system power LED'. Assuming you have a system power LED (your inst book should tell you), I would remove all installed components from the mobo but leave the mobo power cables installed, at least the 24 pin cable. Remove the 8 pin one.

With everything removed from the mobo except the 24 pin power cable and if you do have a system power LED, the LED should come on and stay lit. If this happens, then the CPU is probably defective. If the lights flash and the system power LED blinks on and then goes off, the mobo or the PSU is defective (assumes no components and a system power LED as I have described).

I am trying to figure out how you could tell if it is the mobo or PSU with what you have. If you access to a simple digital voltmeter, you could check the voltages from the PSU; you could take two 6v auto headlight small bulbs and check the 5 v and 12 v rails in the PSU. Would require some more explanation.

Anyhow this is enough for right now; I need to reread this post, check my logic and confirm back to you that it is correct. What I have told you essentially follows the way I think I know my Asus mobo to work; it is an ATX mobo as is yours. Both being ATX means that the system power LED has to follow the same rules for all ATX mobos. Does not matter whether Asus or Gbyte.


I will reread this post several times to be sure my logic is correct. Everything above looks ok. You checked the front panel switchs with a continuity meter; is it a digital voltmeter, even a cheap one? I can or you can find the voltages based on generic ATX power supply specs. I will look, tell you what I find. Worst case if you cannot find the Antec "output" circuit, you can find the pinouts for the 24 pin power cable plus the other smaller connectors from the power supply. Checking the open circuit voltages is probably good enough; however, with a little more effort you could load down the PSU voltage outputs one at a time and measure them. Difference being that the "loaded" circuit test is absolute in accuracy, whereas, the open circuit test would not show a bad "solder" joint in the PSU itself and you would assume the PSU was ok, when it really was not.

This is a late edit about 10:45 pm: Something I meant to add earlier but forgot it. The "system power LED" circuit board traces go directly from the incoming mobo 24 pin PSU cable to the LED itself; this means the "system power LED" is in effect directly wired to one particular output of the PSU. The wiring to the LED does not go thru any mobo circuits, chips, etc..

zburns
08-03-2009, 06:01 PM
Here is a current website link that looks like a good pinout diagram with some explanation for the 24 pin power supply cable for ATX rigs.

http://pinouts.ru/Power/atx_v2_pinout.shtml

You only want to know if it works. Read the voltages from the various pins to a ground pin. Not sure if the ground pins are common to each other at the plug (mobo end) but as long as you get a reading at each pin marked with a voltage to a ground, that may be good enough to tell it is working.

A test with a load (a correctly sized resistor for example) is simple and easy to do but you need a few properly sized resistors from Radio Shack and some more time to do this -- really not sure you need to do it. I can calc the resistor values and provide more explanation for the pinouts and how to use resistor loads if you need the help.

Do you have a voltmeter. No insult intended, do you know how to measure dc voltages with it? If not, it is really simple, sort of like eating breckfast. No big deal!

My PSU is Earthwatts 500 watt. DC outputs of +5, +3.3, +12, +12, -12, +5VSB. You will see the same voltages repeated, but these are separate dc voltage outputs, each with its own current limit, and each for a specific separate circuit.

I hope you have enough info now to isolate the source (reason) causing the lights to go out. Any questions tho, do a post!

geojo2
08-03-2009, 09:55 PM
There is one other thing I would like to mention before we go further. On my Zalman Cooler which I am using instead of the intel cooler that came with the cpu.. There is an upper plastic(which the cooler mounts to, and a bottom bracket which mounts under the mobo. The 4 screws that hold this mounting bracket extend thru the upper to the lower mounts "DO" contact with the metal of the case. I had seen this when the mobo was out of case and I did check them with the standoffs. They were extended just enough thru the mounting brackets to touch the frame. I did not think that this was important because it was in what I thought was a safe zone around the cpu. I was talking to my grandson who had computer training at a university and he said that is what could be the problem He lives far away but I will be see him next week and he will go over it with me.

I did a chat with Intel support today and they said the only way I could check if the cpu was was to put it in another computer which could support the Intel core 4

Everything that was about what was disconnected and what was not etc in the last post I probably did wrong. I will probably end up sending the mobo back to newegg.
I called antec, went over the connections which seem to be right. I do have a mulitester but antec said I needed different one to test the psu.

Since I live in nothern Maine USA and the nearest computer shop is over 150 miles away and it creates a problem. Because at this stage of the game I would just take it there and say "here, you fix it"

I am 70yrs old, retired, and this is really keeping my mind occupied "which is good"

I really thank everyone in this forum for being so considerate

zburns
08-04-2009, 09:54 AM
Read your post; I am sorry you are having these problems. I do have a few comments. I give long winded answers in order to explain things carefully; I did research everything I said and I feel good about the accuracy of what I said. For example I was concerned about powering up the mobo with no cpu/cooler on it; never heard that you can or should do this on this forum. However, it made sense to me that you could do it, so I called Asus and asked a tech support rep if I could do it and he immediately replied, I could; but that the mobo just would not do anything without any components on it (obvious). However, what you and I both wanted was the mobo "system power led" to stay lit. This would mean that neither the mobo or the psu was causing a short circuit to the PSU.

The intel rep was telling you that you must put the CPU into another computer to tell if it was OK; meaning that it was not only not shorted out but was working according to specs, etc..

If your mobo has only cpu / cooler installed and the lights come on and go out; then you remove the cpu / cooler, and then power up the mobo and the system power led stays on, then you have "tested" the cpu (including the zalman), for the absence or presence, of a short circuit, only.

Your comments on the Zalman mounting screws. If the screws, washers, nuts are all mounted on plastic and go thru holes in plastic, then the fact that they touch the metal case is irrelavant. Any chance the Zalman instructions say that the screw heads are supposed to be on the bottom side of the mobo? That could be a game changer, but there instructions would say so!

Zalman provided the hardware; Zalman and Intel had to agree on compatibility; Zalman had to write the install instructions, knowing full well the combo would go primarily into an ATX case. They would tell you which direction the screws go in thru the mobo.

If you have a multitester, non digital, but if it has an old fashioned meter with pointer, selection of ac volts, dc volts, ohms, etc. it will check your PSU voltages. You simply got wrong information from a young tech support person with "digital knowledge" only. (In the absence of a multimeter I already said you could check the PSU with "small auto light bulbs"; I would have to tell you which light bulbs to use by volts and watts, and which PSU voltages to test. What would the Antec guy say about that!)

I can rewrite my last two posts in numbered step by step instructions and separate the "explanations" part from the "do this" parts in order to minimize the confusion. Last I can assure you that putting together the computer is relatively simple even for a beginner as I was; the trick is knowing when someone like Rob tells you that you should do "this or that" -- he does so for a reason. I would encourge you to continue; this really can be simple stuff. There are just steps you have to follow in a certain order when you are troubleshooting.

One last example, static control. While it is not a big overriding problem, conditions being right, you can screw up a board in a microsecond with static electricity; Rob's, Wise Monkey's instructions about wearing a wrist strap, hook it to the case evidently are sufficient to protect the components, but you need to do this "everytime" you handle a component. And you only handle components by the edges.

Also, when you have a board just laying loose on the workbench, no connections, no power, it should lay on "anti static" foam or be put back in the anti-static bag it came out of -- but you would not put a "live" board on these "semi conductive" materials. The bubble wrap you used is an insulator, as would be clean cardboard; this is what you would put out of the case "live components" on.

I cannot write a short post!! Anyhow, if you want to continue I am glad to help. I live in north metro Atlanta; if you wanted to call and it did not violate forum rules, I don't mind. I would email you my phone #. Regardless, good luck!

geojo2
08-04-2009, 05:38 PM
Thats a good thing that I can check the cpu and the cooler like that, and will do so Thurs.

The mounting screws on the cooler have to go down thru the top and catch the nut so they can only go in one way.

I have 3 rows of pin connectors on the MOBO that connect to the case, the bottom row which is semi inclosed with the second row. Labeled in manual as F-Panel

The bottom row contains 5 pins from left to right
HD Led(2 pins marked in mobo manual as: +HD drive activity LED -
Res (2pins marked in MOBO manual as: - RES +
The one single pin in the manual is NC


The 2nd row up has a total of 8 pins from left to right:
first 2 pins are vacant (marked as - MSG + in the mobo manual)
pw (2 pins marked in the mobo manual as -pw+)
-speaker+ 4 pins

The 3rd row there is a 3 pin connection right above the PW pin. This is off by itself. The case manual says: +Power LED-. (The mobo manual says
PWR_LED (System panel LED Header) This is a 3 pin connection with only 2 wires at the connector

and then there are the audio, usb and sata

Because we have been referring back to this from time to time I thought I would be more accurate with the discription.

zburns
08-04-2009, 06:16 PM
Specific question here before you put power back on the system. Do the screws that fasten the Zalman fan to the mobo "rest on plastic and pass thru plastic to the nuts and any washers underneath the mobo". If you do not know this for certain or it is not obvious by looking at them, I can probably call Zalman with your part number and ask them. Logically, this cannot be a problem because Zalman and Intel or Zalman and Gigabyte had to approve the design, but since yourson picked up on this, it needs to be answered.

OK I am not thinking. With the mobo, cpu and Zalman installed in the case, just take your multitester, set on ohms (continuity) and check from the steel head of each bolt to the case where there is "clearly" no paint. Also check to a screw head that you know goes into one of the brass standoffs that in turn go into the case. On all four screw heads, you want an "open" reading, no movement of the meter pointer at all; your meter has an ohms scale, zero ohms is a short circuit, the other end of the meter scale is an "open" circuit. "megohms or infinite ohms" means an open circuit. (I am assuming you have a meter readout, not a digital readout. Correct me if I am wrong). If I am not clear, please ask me to explain again. Check all four of the Zalman mounting bolts; does not hurt to do it several times to be certain. All four must show an open circuit.

geojo2
08-04-2009, 06:57 PM
I was talking to EVGA where my video card came from and I was asking about my 500W psu. They said it would be fine now but if I started to add more things down the road like media devices, TV etc. It could give me problems + the grandkids are going to want to play thier video games on it.. According to "my super pc.com," what I based my config on, the system should be fine.. But the PSU calculater on newegg said I needed 550W. The other PSU calculators on other were lower..but maybe newegg just wants you to upgrade your PSU.

Well anyway,I ordered a CORSAIR CMPSU-650TX this morning $99.00 + a 20.00 rebate=$79.00 FROM Newegg.

If I would have kept th 500W and later had to sell it on Ebay, I would get nothing for it. This way, I can have it bench tested in NYS next week and sell it "as new" and not "used" on Ebay.
I should be able to get $30.00 or so on a "buy it now"

I got an excellent deal on the case and PSU so it is only costing me a few$$ and piece of mind.
I think Antec is in chapter 11 or going to be???? and that also could have created a warrantee problem down the road..

geojo2
08-04-2009, 07:30 PM
The NUTS that catch the screws are recessed about an 1/8 th of a inch at the bottom part of the bottom bracket. The bottom plastic mounting bracket is set approx. 1/16th from the case. It also looks like the screws also protrude 1/16th. When I took the mobo off I had noticed the protruding screws from the mounting bracket. When I looked closer at the case I could see where 1 screw put a slight mark on the case. I also could just put a couple of washers under the screw heads..(no washers came with the cooler).
As a reference I have a Zallman CNPS9500 AT. The mobo is : gigabite EP45-UD3P

zburns
08-04-2009, 10:21 PM
The way I understand your description above is that the bracket on the underside of the mobo is 1/16" off the case, the nuts recessed in the bracket 1/8" and the screws sticking out the bracket 1/16". This means the screws are sticking out the nuts 3/16"; if all four bolts do this, you could buy 4 new bolts 1/8" shorter than what you have. But the washers would work also; it would be better to have only one flat washer under the bolt head. Zalman did not mention or supply lock washers did they? I would try to get more than 1/16" clearance between case and end of the bolts if possible.

If your bolts thru the Zalman can touch any metal surface that is "live" including the Zalman heatsink, and the bolt also touch the case, that could be your problem.

From a "equipment design or engineering design" point of view, a mounting bolt such as we are talking about should never come in contact with the case or get "accidentally" close to it. No designer would allow this; so I wonder why you have that situation. It could be as simple as the wrong length bolt was supplied to you.

Anyway, try to get the ends of the bolts at least 1/8" off the metal case.

Regarding the Corsair PSU. I assume it is going in the Antec case. Dimensionally, it will fit; you checked this out, correct? That case is a little crowded up in the PSU area with the Earthwatts 500 watt. Just a comment!

zburns
08-05-2009, 10:38 AM
Please disregard and do nothing regards my, Zalman, post immediately above this post. It is not that the above post contains wrong information; I was able to find the Zalman installation manual, which will allow me to ask you some better specific questions. Thanks.

zburns
08-05-2009, 03:16 PM
This post is to try to clear up any confusion over the Zalman fan and how it is installed. I assume you have socket 775 mobo for your intel quad cpu. Your part # on your Zalman box is CNPS 9500 (not CNP 9500).

Please use this link to go to the Zalman website: http://www.zalman.co.kr/Eng/product/Product_Read.asp?idx=161

This website has two small blocks to the right of the large cooler fan picture. One block says "How to install" and this is a animated video. The other block "Download now" is a PDF file which you may or may not be able to open on your present computer; you must have Adobe Flash Player (I think) installed to open this. If you do not have Adobe, just use the video -- How to install.

Step by step here: 1. There are four holes in the mobo to accept the Zalman backplate (underneath the mobo). The Zalman backplate has "round bosses" on the top side that engages the bottom (underneath side) of the mobo. The holes on the mobo should be just large enough for these "round bosses" to go thru.

2. The Zalman plastic Clip support that goes on top, the corners with the screw posts, should have top holes a little larger than the diameter of the screws. The bottoms of the four posts on the clip support should have recesses to accept a portion of the four "bosses" that poke thru from the "backplate" underneath the mobo.

If para 1 and 2 are as I describe, then the four screws are forced to go thru the two Zalman plates in the exact center of the Clip support posts and the "backplate" bosses. Since the mobo four holes have to accept the large diameter of the bosses, the screws cannot touch the mobo holes; therefore the four screws cannot touch a "live" surface and become "live" themselves; therefore, if the end of one of the four screws touch the metal case, nothing happens.

I am giving you a description of what I see on the Zalman webpage and saying what must happen. You need to tell me if you agree with my descriptions.

The little video is pretty good, it shows the bosses on the backplate and shows them penetrating the mobo holes to show a small protrusion of each boss coming thru the mobo. This means the top post must have a slight recess to go over the lower "boss". etc..

The above is the way I see the permanent installation of the Zalman; based on the pictures I see, I cannot see how it can cause a short.

If I am right so far, the only other possible cause is the two screws that go thru the "spring clip" and screw into the "cross members" of the thick plastic Clip bracket but this bracket in the video appears completely "up in the air", nothing under it.

geojo2
08-05-2009, 11:55 PM
Hi, did you look at the post on 8-4-09 6:38 pm?? let me know if this seems correct..... My quick replys are not posting..The cooler was mounted according to specs.
I have just e-mailed gigabite to see if the screws going thru the mobo would cause a short..The mobo is not recessed, and the screws could be touching the sides of the mobo . I am going to take the zallman screws out, insulate them, and reinstall,and see..........let you know

zburns
08-06-2009, 08:07 AM
Good morning! Yes I did see your post. My comments in my Zalman "assembly" post were based mostly on the little video from the Zalman website for your cooler part #. When I used the word "recesses" or recess, the reference was to the large plastic mounting brackets you referred to: ie, the Clip bracket on the top side of the mobo and the "backplate" on the underneath.

In my writing in my post "recesses" refers to the brackets not the mobo surface. Zalman's instructions show the backplate two ways: in the video the backplate is flush tight against the back of the mobo. In the instruction manual (the PDF version I downloaded), page 2, section 8, it shows the backplate not flush with the back of the mobo but offset like you said. I did not make that "connection" or observation as I was looking at what I downloaded -- I used the video and that led to my post the way it was written. I was just happy to have found the instructions and pictures!

The screws all go thru plastic, they are already insulated from everything else -- I am going to stop at this point and continue in an hour or so! I need to write out what I am going to say so that I am not confusing.

zburns
08-06-2009, 11:08 AM
About 11:40 am EST. I found what I was looking for. I will go step by step.

1. Use this link from an Amand Tech review of your mobo. The link will take you direct to the page with the picture I want to discuss with you. Here is the link: http://www.anandtech.com/mb/showdoc.aspx?i=3508&p=4

2. Go to the second picture on the page. The third picture may be even better. They are enlarged pictures showing a corner of the mobo that includes the CPU mounting area. You can clearly see one or two of the mobo "mounting holes" that go thru the mobo; the area around each of the holes is outlined in white. The visual dimensions of the space mentally equal the size of the plastic clip bracket that goes on the top side of the mobo. These are the holes the screws go thru. Look closely at the surface around the holes and the surface of the hole "thru" the mobo. You do not see any silver.

3. Look at the right hand corner of the mobo; you see a mounting hole with "short dashes of silver" aligned radially around the top surface of hole but nothing on the surface of the hole depth or hole thickness. These short dashes of silver on the flat surface are the ground points of the mounting hole to the screw head, the ground plane and hence to the standoff underneath.

4. If you look at the wall thicknesses of this mounting hole with a magnifying glass, I will bet you see black fibreglass board, not any silver. The two mobo mounting holes for the mobo that you can see do not have any silver on the wall thickness -- (no engineer would design this way. You would not depend on silver coating the wall thickness as a ground connection with a bolt just "slipping thru". So all you have here is another insulator in my opinion).

5. Final point: The top plate you use to mount the CPU is a rectangular piece of black plastic with four corner holes that line up with the picture link mentioned above. The picture of the board where the mounting holes go thru has no conductive material in the area of the mounting holes. The CPU mounting holes have no conductive material around them. The underneath "backplate" that the screws go thru is plastic and is an insulator. The screws go completely thru plastic insulating material.

Therefore, it does not matter if the screws touch the case metal. Nothing happens.

I am making my statements based on color pictures of your mobo. You can physically look at what I am saying above and see if I am making any incorrect statements.

If you are still not convinced, you can use Scotch Electrical tape and put a 4 inch solid square pattern of the tape on the area of the case where the screw pattern may impinge. Cut the tape with sissors or knife so it is not stretched.

I just think you are fine like it is. If you want to confirm something, just look at the holes thru the mobo; if they look like apx 1/8 inch holes drilled thru black plastic; and there is no "silver" substance around these CPU mount holes, my analysis is 100% correct.

Changing the subject: I am going to look thru your instruction book (downloaded) to find reference to System Power LED; what I have seen so far is a two wire pinout "header" which implies the wires are there but no LED was installed. Something more than "header statements" may be in the book; will prevent assumptions and guessing.

zburns
08-06-2009, 12:27 PM
System Power LED. I will tell you what I think I know about this so far; not sure I am done researching it as this post will indicate.

1. This is about the shaped "chrome" template with three connector ports, two leds and two audio female plugs on the front of the Antec case. It is the right hand LED that lights up and stays lit when you turn power on.

2. The right hand led is the System Power LED and should be lit when power is on. However when I shut my system down with the Vista software, "Click on Shut-Down", the light on the front panel of the Sonata III case goes off. BUT I HAVE AN LED ON MY ASUS MOBO THAT STAYS LIGHTED UP. I AM NOT SURE YOU HAVE THAT PARTICULAR LED ON YOUR BOARD. This is not a problem; I made "earlier" posts here that assumed your Gigabyte mobo had the same LED; it may not have it. Does not effect anything.

3. I do have a PDF download of your manual. Pages 26 and 27 deal with the System Power LED.

My question to you: When you previously say that the lights come on and then go off, have you looked at this front panel led on the right to see if it is going on and off as other leds inside the case are doing?

It is now about 3pm. The following in brackets is from your post dated Aug 4, 2009 6:38 pm.

[The 2nd row up has a total of 8 pins from left to right:
first 2 pins are vacant (marked as - MSG + in the mobo manual)
pw (2 pins marked in the mobo manual as -pw+)
-speaker+ 4 pins

The 3rd row there is a 3 pin connection right above the PW pin. This is off by itself. The case manual says: +Power LED-. (The mobo manual says
PWR_LED (System panel LED Header) This is a 3 pin connection with only 2 wires at the connector]

My comments: On my Antec Sonata III500 case, a flat ribbon cable with 4 conductors has the wires that go to a plug similar to what you describe. Two wires, green to + MSG (left pin, 2nd row) and black to - MSG (2nd pin, 2nd row). My lead color is green for the plus +, and black for the minus -. The Antec book, referring to these leads says that colored leads are plus +, and black or white leads are minus -. Both leads go into a two pin female plug which match the first two male pins left side, 2nd row. Same configuration as yours apparently. The Antec book says to turn the plug over if the light does not light. Row 2, on your case, the first pin is +, the second pin - . (just repeating myself).

Final comment on this, the three pin connector on the 3rd row is described in the Gigabyte manual (page 26, section 10 and page 27, section 11) exactly as the first two pins from the left, row 2. Same configuration. Same explanation but it seems to be for a LED like the one on the case front panel, but located inside the case. I have an LED inside the case that stays lit even when I turn off my system with the software; if I turn off by switch at the back of the case, my "inside" LED goes out. Not the same as yours unless it is one of the other leds you talk about blinking on and then going out.

My earlier test suggestions for the mobo / CPU should still work.

zburns
08-06-2009, 09:33 PM
Well congratulations. I did say it was simple but I did not say it would be quick!! You learn a lot doing this stuff.. Good luck on the rest of it.

zburns
08-07-2009, 09:27 AM
Clear me up on what exactly happened!! From what you write, the rear 120mm fan was hooked up incorrectly and did not run at all. This, presumably, would allow excess heat to build up inside the case, the CPU run too hot, and the unit shut down because of the excess heat.

Is this the way you see it? Thanks

IMPORTANT, PLEASE HELP!! (New comment, little later than the first sentence above) Go back to when all this started. You did not have the rear 120mm fan connected properly. Here is what I would like to know-- Did the system shut down in seconds or did it take a couple of minutes or even more!!

The reason is that your mobo may have circuitry that orders a shutdown if that rear fan is not properly connected. Or the CPU was shutting it down; this could take just a few minutes or longer. Not sure. I have done some searching; plenty of answers out there, but nothing declarative from a CPU or Mobo manufacturer, mainly forum opinion.

So whether your system shut down in a few seconds (a type of fan warning and shutdown) or several minutes or longer (CPU thermal shutdown) will tell us which it was. Thanks!!!