Contribute
Register

[Solved] Hackintosh Killing RAM

Status
Not open for further replies.
WOW a lot to respond to. Thank you. First of all, this is my first time building a PC, and I've never run windows so I can rule out any Windows malware, the SSD seems fine but I will keep an eye on it, the RAM is non-ECC, which the system is supposed to be able to handle, but I decided to order registered ECC here: https://www.amazon.ca/gp/product/B01BW535O4/?tag=tonymacx86c0c-20

The Dell board required a small mod to get the brand new Intel heatsink to fit, basically extensions for the screws to extend them into the holes since Dell uses a proprietary heatsink but the layout is the same more or less. I didn't know this before I bought the board. It seems to fit fine, and the CPU fan works well and doesn't give me any errors, but if the RAM I ordered doesn't fix it then that I believe is my next culprit.

I just have the one PSU, as a note I have an extension on the PSU to extend it to the motherboard, since in the DELL case the PSU was at the top next to the connector, but in the modded Mac G5 case I have the only place I could fit the PSU was in the bottom of the case where the Mac one used to be. This PSU uses Dell's proprietary pinout, but the extender clicks on fine and connects to the right pins. I should have a good look at it and see if any pins have come loose just in case. I was a bit worried that extending the cable might reduce the voltage on the rails a bit, losing some to magnetism or whatnot, do you think this is possible?

I have some PCIe cards (a SATA card based on the Marvell chipset) and an NVIDIA EN210 graphics card which I am replacing with an ATI 7870 soon, but unplugging the PCIe card (except the graphics card for obvious reasons) didn't make a difference.

Now I am trying to decide if I take the CPU out right away to check the pins (I'm pretty sure it's fine, unless the heatsink is somehow squeezing it too hard), or test the new RAM right away when it arrives in a few days and see if it fails. What do you think I should do first?

I haven't made too many demands on the system, and the 635w PSU is I believe the middle one (they have a 4-hundred something and a 700-something), the graphics card is really conservative with power (until I get the 7870 running), so I'm not too convinced there is a power issue. I'm going to pray that it's the RAM (which I bought used and was supposed to be all the same, but maybe it was worn out, or maybe it was just wrong for my machine?) and that new RAM will fix it. Or else I will be out $60, which is what it cost on Amazon. Or I could check the CPU first, which is a pain because I'll have to buy some thermal paste, but maybe it won't be as hard as I think. What I don't want to do is damage it while removing it.

Also, the CPU is an Intel E-2640 6-Core first generation. I don't think they ever shipped a system with this particular CPU, and I've only seen this CPU in dual-processor (12-core) configurations. There was no reason reading the documentation that it would not work solo, so I got it cheap on eBay where there are a bunch of them flooding on from old servers. It ran fine and passed the built-in Dell diagnostic.

Perhaps I'll run all the tests including Memtest again when the memory arrives. I have one (sometimes) working stick of 4GB for now, but I don't know if there's any point testing it again with the bootable memtest if I'm going to replace it soon.
 
An update on this. I gave away all the RAM except for one stick, thinking it was probably bad. Then I switched graphics cards, from NVIDIA to AMD, and for some reason the AMD card wouldn't work in the slot the NVIDIA had been in. I tried it in a slower slot, and the computer runs fine, and i haven't had any RAM issues since. It makes me think that something funny was happening between the graphics card (or slot) and the RAM in the computer. It's been a stable system for a couple days now, just waiting to see if it holds.
 
Status
Not open for further replies.
Back
Top