did i kill my 'puter

Nov 16, 2012 92 Replies

i had a CMOS checksum error on my CNC mill. So, I installed a new motherboard battery. Now the CPU fan just comes on and it makes no attempt to boot. Is the computer DOA? I checked connections and wiggled stuff.



Karl



Remove the new battery & short out the socket for a few seconds, then reinstall the battery. You may have some corrupt data in the CMOS scratchpad RAM.

tryed that, No Joy Karl

Try another power supply, if you have one. You may have a dying or dead motherboard. Unplug all the drives, and cards except the video card, if it isn't part of the motherboard. If you can't get the bios screen to flash, look at the electrolytics on the motherboard, and the ones near the processor in particular.

If they are bulging, they need replaced, or replace the motherboard. They are a special low ESR 105° C type capacitor, and not always easy to replace, because they are soldered to internal layers. If you can find the exact same type motherboard, and move the processor to the replacement board you shouldn't have to reinstall any software.

Maybe re-seating the memory will work. I have had to replace memory sticks , too.

How did you determine that the unspecified cnc mill with the unspecified controller running an unspecified OS with an unspecified user interface made no *attempt* to boot?

If it's an IBM-PC compatible controller with a keyboard and display and you don't get any beeps and you see nothing on the display and entering the BIOS setup pre-boot keystrokes at the proper time don't do anything, yep, you've probably borked it.

If it's the typical coin-cell non-rechargeable CMOS battery, there's a diode in series and shorting the socket does nothing. You have to WAIT, I'd give it an hour before you give up and replace the motherboard. There may be a jumper to reset the cmos, but it may only open the circuit and also require you to wait. Shouldn't take an hour, but what do you have to lose by waiting longer.

Also, did you unplug the computer while you waited? Many computers have a 5V supply that's hot when the power is off and may prevent the CMOS from resetting. It's also possible that there's enough power coming back thru an I/O port to keep the CMOS alive.

I had a situation where there was enough juice coming back through a GPIB port to keep the GPIB fully functional with the power to the microcontroller removed. I unplugged it and it just kept on running. Thought I'd been possessed...

This is a P4 motherboard with an ISA slot for my Galil card. Rare and expensive to replace. I happen to be in Ocala Tuesday. Are you interested in looking at it?

Karl

I just learned that a severly disabled freind of mine died last night & I'm just not up to it tonight. I only slept a little over three hours last night, and I'm not able to concentrate on anything right now. I was just getting ready to post a message that I would be off the group for a few days to a week, to see if I can help her family. My email is good, so you can contact me off group. I should be in better shape by Tuesday and more than happy to take a look at it. I may even have a spare motherboard that would support the card. Email me and give me the details, I probably won't be back on the group before Monday or Tuesday.

To everyone else, this is the second disabled freind who has died in the last couple months. I wasn't able to do anything to help the first, because they can't locate a next of kin. She had a sister, but wouldn't talk about her or even tell where her sister lived. In this case, her yuongest son is disabled and she was all he had.

Anybody with more than one computer should get a power supply tester. They are just not expensive at $23. Its the first thing I do on a problem box and often the only thing I need to do.

formatting link

A good article on bad (bulging) electrolytic capacitors, causes, and good photos can be had here:

formatting link

Looks like one major cause of many failed units started roughly around the early 2000's, and traces back to a botched industrial espionage episode. Details are in the the above article.

Erik

I have restored function to computers by unplugging them from electricity and unplugging and re plugging every connector in the computer. Doing the unplug/re plug routine gives all the connectors a new fresh connection. Corroded memory board contacts will keep the computer from booting.

"Michael A. Terrell" on Fri, 16 Nov 2012

20:29:11 -0500 typed in rec.crafts.metalworking the following:

Memory Eternal.

My prayers go up for you.

-- pyotr Go not to the Net for answers, for it will tell you Yes and no. And you are a bloody fool, only an ignorant cretin would even ask the question, forty two, 47, the second door, and how many blonde lawyers does it take to change a lightbulb.

You may need to reset the CMOS. Pull the new battery, and short the pins of the holder for a few seconds, then re-install the battery. You will probably have to reset all the options after this, of course.

Good Luck, DoN.

I had a similar problem. Can you get to the boot menu? I found that the order of devices used to load the operating system had changed and it was trying to load from the floppy drive instead of the hard drive.

Dan

^ This. Whenever strange issues start to crop up with a machine that's fairly old, powering it down and unplugging/replugging every connector a few times wipes oxide buildup off the contacts and usually solves the problem.

JOY, This was it but shorting it didn't work. leaving the battery out all night did.

Karl

That was particularly true back in the days of the "model T" computers - where all of the ICs were socketed - and only the best used high quality screw machine? sockets. Every memory bit was a separate chup - with tinned legs.

On today's computers, with gold plated instead of tinned connections, it is EXTREMELY rare.

I'd say, in order of likelihood- bad power supply (average ATX P/S up to about 5 years ago only lasted 2 - 3 years) Bad caps on the motherboard, or bad motherboard. What model board is it???? ASUS, A-Open, etc and actual model designation???

Lucky!!!. Back when the RTC was a separate chip we would remove the battery and put a peice of tinfoil over the clock chip to short out all the pins. The clock chip contained the C-MOS in most of those older machines .

Glad it worked. I had forgotten about the likelyhood of a diode in the battery circuit, but it makes sense. (Now, if you could identify the actual chip with the CMOS memory in it, and which pins were power and ground on the chip, shorting across there would have been quicker. But you have it done, so no worry.

Good Luck, DoN.

[ ... ]

Easier than in the older Sun workstations, which used a chip by Toshiba which had the clock, the CMOS RAM, a crystal, *and* a coin cell potted in a single package. No way to short out the battery from outside. *And* -- it contained the host-id and the ethernet MAC address inside it, with no easy way to reset those. (There was an around-the-corner way, but it was not easy.) And if you had any licensed software, the host-id was very important. If you did not have it recorded, you had to go to Sun for a replacement, with the barcode from the label on the chip.

If you left your computer on 24/7 you would get a really long life from the built-in battery. If you stored it powered off, you would be lucky to get five years out of it.

And -- there were (are) instructions on the web on how to dig into the potting and wire an external coin cell to replace the now dead one inside it. :-)

Enjoy, DoN.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required