Mushkin Blue errors

I have a set of 2x512 Mushkin Blues (990992) that I had been running on my DFI at 257MHz 2-2-2-8 3.5v stable. I decided to see how they would run on an Intel rig, so I tried them in my AI7. They wouldn't run stable at any speed (up to 234MHz) or voltage (up to 3.2v) at 2-2-2-8, but I wasn't surprised as I don't think they were really designed for Intel mobos. This is when I retested my old Mushkin Lv2s on the DFI - stable at 265MHZ 2-2-2-6 3.5v (they only get better with age). Since the Blues wouldn't work on the AI7, I switched the sets back. The Lv2s were fine, but now the Blues throw Memtest errors in (at least) tests 3 & 5, at any speed, voltage, timings, individually & together. The errors are reported in exactly the same spot each pass for each stick. I retested them individually in 2 other Intel rigs at 200MHz and a variety of timings - errors in the same spot. The sticks are done for - can't even get into Windows.

Now, what to do. I really don't know what caused this. I didn't mishandle the set - I've been moving my Lv2 set from rig to rig for years with no problems. They've always had active cooling. Thing is, I removed the heatspreaders shortly after getting the set, once I knew they were OK. I assume this voided the warranty, though if I can still get support from Mushkin please speak up (Duonger?). If Mushkin can't help, I can understand - I knew what I was doing when I removed the heatspreaders.

So I may just have a set of bad memory on my hands, and have to live with it. But if anyone has any suggestions, about what might have caused this or what I might do about it, feel free to say so.

Well you did void the warranty as I think 2.8 or 2.9 is max allowed. what mobo allowed you to run intel with vdimm over 3.2volts? Basically if you push them and break them you are out of luck. go over to their forum and ask if maybe you can trade them in but I don't think they will do it if you over did it.

doesnt hurt to talk it over with muskin support.secondly unfortunately at any time a set can go bad.. maybe this happened to u.. although it seems strange, usually the sticks just degrade but keep working

Yatta said:

what mobo allowed you to run intel with vdimm over 3.2volts?

Click to expand...

As I said, up to 3.2v (didn't say beyond) on Abit AI7. Up to 3.5v on DFI LP nF3 250Gb.

I've never RMA'd anything before, not that this set is even eligible. I don't believe in running a product till it breaks, then expecting a new one. I know the high vdimm and heatspreader removal may leave me with no warranty support, but I'm pretty sure something else killed them. It seems to have something to do with the switch from one rig to another, but that's just a guess. No ESD that I know of, I handled them carefully. When I loaded them back into the DFI (when the repeatable errors began), I booted them up at 200MHz and 3.0v. Then went up gradually in speed and voltage. I tried to get into Windows at a previously stable 255MHz 3.4v. When Windows wouldn't fly due to file corruption, that's when I went to Memtest and found the errors.

Anyway, I was just hoping to get some suggestions from those here in the memory forums that have more experience than I do, especially with product support after overclocking. If I had done something with this set that was well outside of what others here do, I would at least have a better idea of what killed them. So even if I'm out of luck with support from Mushkin, I'd still like to know what killed this set so I don't repeat it.

Is it possible that the sudden jump from 3.2v on the Intel to 3.5v on the DFI was the culprit?

I've read, with respect to the OCZ branded Value VX to not just shoot them right away with volts, but I don't know if (1) there is actually any validity to that or (2) whether that concern would carryover to the Blues, as well.

The only thing that you might consider, if you haven't done so already, is to clear the CMOS on the DFI and start over with the Blues, at default settings and try to work your way back up. It may not get you anywhere, but it will allow you to verify that it isn't some obscure setting that is messing you up.

In any event, it really hurts to loose a nice clocking set of Blues so I hope you can get it worked out.

Try 255 and see if it helps? Might also want to look at your 3.3 rail and see if the adjustments you made got messed up when you took out your mem.

KrisMCool said:

When I loaded them back into the DFI (when the repeatable errors began), I booted them up at 200MHz and 3.0v. Then went up gradually in speed and voltage.

Click to expand...

3.0v is what the set needs to boot at 2-2-2 at 200MHz, so that's where I set them after a swap. I tried clearing CMOS - no change. The errors exist at ANY speed, voltage and timings, in the exact same place on every pass. The board, PSU and the CPU are fine, so you know - writing this at 10x260 with Mushkin Lv2s at 3.4v, 2-2-2-6. The Blues are out of the rig for good. Not dead, but consistent errors bad enough that I can't get into Windows. I'll try Mushkin support soon, but I'm not holding my breath. I wouldn't feel so bad about the "cost of experience" if I at least knew what to learn from it (what killed them, do not repeat). Thanks for the replies. This thread is just one of my sad tales - just got OCZ PC3200 Gold to replace the Blues and they're not working either.

Have you tested each stick individually?

KrisMCool said:

the Blues throw Memtest errors in (at least) tests 3 & 5, at any speed, voltage, timings, individually & together.

Click to expand...

Yes.

Surprise, surprise! The set still lives. And at previously stable settings. I read a couple posts (at several sites) about people getting errors or killing memory with a variety of UTT sticks. After being out of use for several days, they were put back in and worked like before. Thought I'd give it a shot, as my set hasn't been used for about 4 days. Well, they both work. Really odd, but this worked for others with problem UTT memory. Whatever the reason, I'd say that if others out there have errors or think they killed their sticks, let 'em sit for a few days and test again. Anyone else experience something like this?

I thought most of the users experiencing that phenomenon were using twinmos modules and that it maybe attributed to the SPD chip on the modules. This is the first case I heard about the blues.

I recall the previous problems of winbond ch-5 with intel i875/i865 mobo's: devider and instability issues. Well the utt ch-5 seems to be no exception.

I've ordered a Gig of Corsair twinx pc3200 c2 rev. 1.5 with utt bh-5 chips. I hope they perform as well as my good old kingston pc3000 sticks!