I went to a junkyard in town today and bought three Dell Precision T3500s for $35. Only two have CPUs, those being Xeon E5520s. I put some RAM in them and a GPU, because they didn't have any and booted them from a live Fedora USB. Both worked until suddenly one shut off while testing, and would not power on after. I would press the power button, and it would be on for like a quarter of a second. I'm not sure what happened. It just randomly quit, I did notice, however that before booting from USB, the computer said low battery voltage, so I tried swapping out the CMOS battery with one from the other working T3500 to see if that was the problem. No luck. I thought maybe the power supply died, swapped it out. Still nothing. Maybe the DIMM slot went bad. Moved the DIMM. Nope. I tested the RAM in another system to ensure it still works and it does. I'm not sure what to do. Does anyone have any suggestions? I'm going to bed soon, so I probably won't reply till tomorrow, so thanks in advance for any input.
If you can make one work out of the three you should be happy.
Generally workstations like these are from enterprise environments and enterprise companies recycle their old hardware, not throw in in the trash. I would imagine someone like yourself has already gone over these and determined they are beyond (economical) repair. I wouldn't waste to much time, let alone money, trying to fix these up. They can be had for $40 on eBay in working order.
I know, but I just find it strange that one just turned off out of no where. I'm certainly not going to put money into it to try to fix it. Maybe I'll think of something I didn't check, but thanks for the input.
Random power offs could be so many different things. Obviously you would want to look at the power supply, is the PSU fan moving? If the fan died or its really dusty the PSU could be overheating. Check/clean the HSF, change thermal paste, make sure the CPU fan is moving properly.
I had a old Dell enterprise machine that had a problem, believe it was memory related, BIOS update fixed it, so that's something you want to look at too. Dell has one click BIOS updates for these machines so check out Dell's website and get the latest BIOS.
Could also be a deeper issue - CPU, memory, motherboard. You could spend all day on it and get no where. If you have the time and just want to tinker with something than by all means go for it.
Let it die after a poweron attempt and look for the service lights (should be 1234 or ABCD). Then look up the manual and see whats up.
If there are no service lights on:
Take the bios battery out and somewhere in there is a jumper for RTSC_RESET. Pull that out and put it on the other available jumper that SHOULD be next to it called bios_config or something like that. Do all your tests again. On the mobo there should be a max of 3 service check lights that could be green, yellow, red, or alternating between all of them. 1 should be by the IO panel and the general back of the mobo, one should be under those jumper spots somewhere, and the last one could be by the sata, but could also be by the CPU sockets, fan headers, RAM, or area of the ram. At least ONE should be green. If 2 are yellow that means theres a bios fault that probably should be updated to fix. If any are red or not lighting up at all then the board is bad. If all are green and the service lights aren't on and your testing hasn't done anything it should be working fine and not doing anything out oy the ordinary and something might just be loose.
I have been fucking around with dell shit for way too long to not share info.
For the price you paid, as @IRAGEDandQUIT says, if you can make one from the three and call it a day /._./ . If you have a multimeter and some time to waste, you can measure the caps and resistors on the good one and spot the failure on the fucked one and replace what went bad. If you're lucky enough, it could be a blown or working below threshold cap which usually are easy to replace. I have done that with several P4 era optiplex & dimensions, those early foxconn boards where terrible but usually the caps were the most common failure, I've seen some explode like popcorn.
I've seen a cap open slowly and just a tube of orange start growing out of if. After about 5 minutes it hardened, started to glow, then caaught fire and was shot out of the machine via fans.
P4's are great from 2003.
................
Yeee. I remember at elementary school we had P4 dimension 2300-2400's, & Optiplex 260's. Funny how those things caught on fire or just did weird shit and break from time to time and no one would give a shit, teachers blamed us on "Playing computer games" (Halo, mainly) but the school's tech was a crazy dude and he would snatch a lot of e-garbage and replace the caps on those things (!). And now ppl are making the Note 7 exploding a big deal. Back in the day we were for real lol #noexcuses.
I'll look into that. I did notice some LEDs on the board.
The service lights are on the front top or back of the case fyi
I didn't notice that yesterday, so I'll be sure to make note if that too when I have a chance.
By my count, that leaves either the CPU or the mainboard.
You could be all professional-like as Aremis suggested and check the service lights, but I would just try the other CPU, especially for a dumpster-dive system. (Props for accepting that challenge from Wendell and then tripling down, btw.) If that doesn´t work, I would say process of elimination leaves only the mainboard, so eliminate it. :)