ABIT KD7 - No video/post
I have a KD7 that I am building with all new components. I put everything in the case and powered it on. No video and no beeps. After an hour of trying different things, I removed the motherboad from the case and tried it with just CPU/heatsink, ram and video. Still no video. I have tried clearing the CMOS several times to no avail.
The fans spin and the light flashes on the CD-Rom drive.
I'm too tired right now to take a video card out of another machine and try it, but I will tomorrow (later today actually).
Any ideas on what I might be overlooking?
Specs:
KD7 - non raid
xp2500+ Barton
MSI ti4200
350watt psu
TIA!
The fans spin and the light flashes on the CD-Rom drive.
I'm too tired right now to take a video card out of another machine and try it, but I will tomorrow (later today actually).
Any ideas on what I might be overlooking?
Specs:
KD7 - non raid
xp2500+ Barton
MSI ti4200
350watt psu
TIA!
0
Comments
I had a P3 that did the same thing (wouldn't post but powered up) and it was to do with the ram or the the header connections. I'm thinking in my case it was the ram but it can't hurt to check the other possibilities.
Kanez: I am not sure the mobo was good before I got it. It is used, but was supposed to be tested.
Shivian: It's not the reset/etc connectors as it does it with nothing connected at all. I will try ram in different slots though if removing the battery doesn't help.
Thanks guys! I'll let you know what I find out.
Turn it on, the mobo should beep if the CPU is good and no memory is present.
Next add one memory stick. See if it posts then.
Omega: I have reseated the CPU, but that didn't help either. However, I have not tried booting up without the ram installed. We'll see what happens when I move the new components one at a time to my KX7.
Man I hate tearing down a working computer....
Remove ONLY the CPU and power-up for a few seconds, put it back and try again.
If that doesn't work, I'll show you a trick later.
Too bad Abit Forums are down otherwise I'll give you links to many of my threads on the issue.
Here's an update to what I have tried thus far....
Put the MSI ti4200 in my KX7 - works okay
Put the PNY pc2700 in my KX7 - works okay
Put the xp2500 in my KX7 - gets to post screen and freezes at ram OK (says unknown CPU).
Also, I tried my 1700tbred and radeon 8500 in the KD7 and had still had no video/post.
I put all my old stuff back in the KX7 and I'm back on it now.
Then, re-install CPU and try again.
What we're trying to do is reset the OTP circuitry.
I have, since reading your post, put the memory and video back in, but left the CPU out. I turned it on for about 10 seconds and shut it off. I then installed the CPU and turned it back on.
I nearly shit myself when it POSTED! If I never told you before EQ, I'm telling you now. I LOVE YOU MAN!!
Now I can put it all back in the case and hope it still works.
THANK YOU EQ...YOU ARE THE MAN!!!!!
Bill
In fact, he put everything back in the case and it doesn't boot anymore. It's not case related or anything else, the board's OTP circuitry is acting up.
The fix for that is to solder a jumper to the board but you'll need good soldering skills. The other choice is to use conductive paint instead of soldering but doesn't always work.
1. Guy brought in this brand new mobo.... Had nice used case, bought from friend. Friend forgot to give him risers. He thought-- ok, screw mobo to backplane. Well, three problems resulted:
CMOS battery was totally drained first time power was applied as backplane was bond-grounded to ground wire of power cord and battery instantly got drained when power cordplugged into wall and it had a ground route. I took MB out, put on insulated surface, put in new CMOS cell and it powered up. Stuck standoffs of normal height in and srewed mobo in-- no power up. Took out of case, powered up after resettting CMOS, POST time. Scratched head, pulled some hair, put pink foam antistat pad with holes for studs under MB, screwed it on, silly thing posted right up. Dismounted, put pink foam only under fan mount braket on underside of mobo (nice metal thing, um,conductive), remounted, POST. Repeated sans pad-- NO POST. Removed screwon backplane for mobo with mobo on it, and found the heatsink mounting had warp\stressed mobo. R&R'ed with pink foma under mobo where heatsink bracket was under mobo. A year has passed, no problems.
Look for a warped mobo, CMOS Cell that was jumpered to on prior to mounting and got drained almost all the way by being shorted to case after "first good post" worked, and test ground-bond the board to a ground that you know is good from one of its ground points to the ground connect on a surge strip outlet that does have a good polarized connect to a known good outlet.
2. Had a friend with a new PSU, new mobo, old case. He jury-rigged the PSU, and did not replace the switch. Key words here were old case, and the OLD switch that was not a momentary contact switch( ever tried to control an ATXPSU with an AT switch and satisfy its power-good hunger?? try every other boot you get a post if lucky!) Leson two, make sure the power good circuit works and someone did not try to use an old PSU(non ATX 2.03) or old switch or did not connect power good jumper and tried for a second boot attempt and got zilch. In our case, we figured out that the second boot always happened after the PSU had been unplugged for 10-15 minutes, and I remembered an AOPen PSU's liking to cold reset and needing 10-15 minutes to do so(thermal cutoff, and PSUs shorted that way thermal quickly). One reason I like to keep a retail Antec PSU around is that uaully they come with switches pre-connected. I also used a little PC Power And Cooling ATX tester to cross-check this-- PSU stayed bad if plugged before 10-15 minutes was up, tested good after that.
3. Guy brought in a mobo with a nice post-mounted heatsink he had bought used with no posts. No POST after his posts were put on board, while board was mounted in case. Post out of case on insulated surface FINE (every time, we repeated boht ways twice). Looked on underside of board. NICE METAL posts. Hmmmm..... Artic silver paste.... Hmm.... Copper heat sink.... Hmm....Metal heat spreader on topof CPUIS NOT supposed to be grounded, but looked like a nice ground route to me.... Lesseee-- metal plate, silver conducts. heat sink conducts, Post conducts, case is safety bonded to ground, did he just manage to ground short CPU from the TOP SIDE??? Yup. Double stick foam tape under bottom of metal posts, no more boot problems(cork or rubber would have worked here also, had foam tape handy and did not want a lesser non-conductive paste as it would not have conductively cooled as well). Also, nylon bottomed post sets woudl have worked....
Lesson--ground *not* thy heatsink directly or indirectly to chassis\case\wall outlet when using metallic content heatsink paste, nor carbon impregnated conductive pads (carbon conducts electricity too-- in fact, most of the best heat dissipative things will conduct electricity at least SOME, heat is ions jumping around and electricity is ion flow).
4. This one was more "fun," took a magnifier cuz I have iffy vision, but someone had cross-shorted his CPU by sticking a tib too much heatsink compund on (actually,gobs too much) and managed to short transistors to core. Cleaning excess paste off actually let the thing work (rubbing alcohol and qtip works in an emergency).
5. This one was not a normal short-- what are the chances of one corner pin on a CPU getting bent and shorting against another on a P4 chip between mountings???? Bend pin carefully up with very finetipped tweezers and tip of Xacto knife blade and bent-tip dental probe-pick, runs like a champ. Oh, WiHa (WILLI HAHN Corporation) makes great fine little tweezers and one of the little dental picks they and others have available made a fine "starter" as the pin was squashed flat ( Did I hear "who needs a dental pick??").
Times 6-20 were variants on the first two, so will not bore you more.
In this case, think from what you said the mobo probably needs RMA'ing, but any other issue you may have noticed that seems very much like any of these might need to also be fixed. Fixing them first might save an RMA.
John Danielson.
This behavior is known to occur on all KD7 models (R/G/S/E) and has nothing to do with PSU's, RAM, case mounting or any other component for that matter.
Simply put, sometimes the CPU fails to send the right signal to the OTP circuitry (Over Temperature Protection) or the circuitry fails to read it properly and the board shuts down to protect the CPU from heat damage.
And it is not something caused by a slight undervoltage into PSU from supply causing an overread, but instead is a mfring or design problem???
Live and learn and do not buy that one thank you, IMO.
John Danielson.
We cannot call it a mfring/design flaw as most of the boards do not exibit this behavior.
I had my friend solder a wire per EQ's picture and it posted on first try outside the case. I'm getting ready to put it all back together AGAIN and see what happens. So far, so good....
I'll report back later.
Put everything in the case and zilch. Back to square one. I cleared the CMOS and it posted. Booted to the MaxBlast Cd and formatted the HDD. I don't have an OS to load yet, so I shut it off and just for kicks turned it back on. NOTHING!
I give up.....ordering a new NF7-S from Newegg tonight.....
Bill
I had a similar problem with mine, it would post after clearing cmos but not on reboot, i have pc3200.
I cleared the cmos and then went to bios and I changed the DRAM clock from SPD to 166, and didnt have a problem after that.
Dont know if that is your problem but it was for me.
The first time it showed DDR333, then it showed DDR200. The bios shows 166, but it's picking it up as 100 for some reason. Sometimes it shows the processor at 1100mhz even though the bios says xp2500 (333).
It is booting up repeatedly now, it just doesn't show the right DDR speed.
Cheap ram? Defective mobo?
I'm confused.
So.....I'm back working on the KD7. So far tonight it has booted repeatedly at the proper CPU and DDR settings at POST. I'm installing Win2K now and should have it up and running shortly. I guess when the NF7-S gets here, I can start all over again.
Thanks again EQ. The OTP mod has made the board usable.
//edit: I'm breaking the xp2500 in at 2162mhz running F@H...