Thunderbolt Display turns off randomly

My problem ended up being a problem with the display cable. Had it replaced and it's as good as gold now.

My NON Thunderbolt display will also go black ramdomly! It happens when it's on all day. Not in a hour or so.

I have two EIZO displays connected to Mac Mini late 2012, the first through HDMI->DVI adapter, the second connected to the Thunderbolt AJA IO XT device, via Thunderbolt->DVI adapter. So the display is last in the chain. Pulling the Thunderbolt cable out/in will take the screen back.

Have all updates installed, and yes, I have 16 gig of OWC memory. I thought that was the better ones for mac Mini...

Hmm... yeah, okay, this is interesting. Is that RAM CL9? It's been mentioned before that this is likely a timing issue, so the CAS level could be the problem. If it's not CL9 then you could get these problems -- nothing to do with the display, per se. The RAM would test fine, since it's not a question of it failing in any way...

[EDIT] Okay, just looked around OWC's site, and it look like it's CL11, which might be the problem. I wonder if they're testing it with TB displays? (That is, since they certify it for these machines...)

Well, I don't know what CL9 is, and I can't find it in the specs, but you can see the RAM I have on the following website:

http://macupgrade.eu/catalog/owc-geheugen-16gb-2x8gb-sodimm-pc12800-1600mhz-p-84 4.html

Yeah, looking at that link it is CL11, which is... well... 2 slower! ;-)

The CAS level is a latency rating. I don't honestly know the gory details about it, but I'm sure someone else here can fill you in on that. But being related to latency, it definitely involves timing! I believe it's an attribute of the controller chip on the DIMM.

...oh, and just to contextualize this; I had the problem too, but got rid of my TBD a long time ago. A few months later I discovered the CAS level thing, so I looked into my own 16GB kit -- it is CL11. Could be a coincidence, but probably not. I'd be interested in trying CL9, but I don't have the TBD any more, so it's a bit of a moot point.

But jonmichaels above, posted a link to Crusial memory on Amazon, and that is also CL11...

And he has no problems since that change...

Yeah, CL9 is harder to find, and usually a bit more expensive. But if it works, well...

[BTW - I seem to recall that mine is Crucial CL11 as well. In on a 2012 MBP.]

Yeah, my computer has been rock-solid ever since I installed that Crucial RAM. The display hasn't turned off once since May 16.

I can't speak to the technical differences between the OWC 16GB kit and the Crucial one, but if you're going to use 2 monitors on your Mac Mini, it's probably a good idea to go with the Crucial (or possibly the Kingston that was mentioned in an earlier post).

Have a late 2012 Mini connected to a Thunderbolt Display and a Dell Untra Sharp 2408WFP. I upgraded to 16 GB of OWC memory and started to have blackouts of the Thunderbolt Display. Removed the OWC and installed Kingston and have not had a problem since.

I can reproduce the blinking screen by running memtest application like "Rember"

Can everybody who "hasn't had a problem since" post the CAS latency on their RAM? The brand doesn't mean much at all, but if it can be determined that everybody who has been "cured" is using CL9 RAM, then that would (probably) be useful.

If you can reproduce the flicker by running memtest application then you can pinpoint your issue. In my case, I have had no proplem since switched to 16G CL9.

We've chased on this issue a couple times. Mac mini 2012 flicker/black screen issues have been an issue out of the gate with certain display setups even with stock, factory ram.That said - updates - including a couple earlier this year appeared to eliminate these for pretty much all. Not a cop out, but these systems have not been perfect in terms of video stability for a lot of users right out of the gate.

The factory design specs for the memory in this model are 1600MHZ DDR3 CL11 1.35V (Low Voltage). 1.35 vs. 1.5V memory uses a little less power and generates a little less heat. Relatively small difference, but it all adds up.

We have a couple dozen of the Mac mini 2012s in our testing labs and have not had success in replicating the issue. With the information in this thread, going to look at it in a new way + we'll try to get access to a system and accessories that combined create the problem as well.

Since the current issue is with two displays... and with a report that replacing a mini-Displayport cable was a solution in one of these two display configs - I am having our team look at noise today from the video side. The actual mini-Display port cable we use for the one display witht he Thunderbolt unit being the other display may be related to our non-detection of an issue.

And... noise would give reason why 1.5V memory (which is what the majority of memory ships operating at vs. the 1.35V low-voltage Apple has moved to), with the higher power draw could potentially reduce that aspect. 1.5V memory does operate just fine in all Apple's systems with the LV spec, just not what was factory intended.

CL9 is the CAS latency for 1333MHz. CL11 is the cas latency for 1600MHz. CL9 vs. CL11 I do not believe has anything to do with this. 1.5V vs. 1.35V is what we're going to focus on. Further, as a side note - 1600MHz programmed with CL9 latency is CAS overclocked memory. That would be you don't see a lot of it as there is little to gain with overclocking - unnecessary and tends to increase long term fail probability.

From testing and a lot of experience, this two display issue is limited to certain Mac minis and potentially further limited depending on what two displays/cables, etc in use in the config. And in this situation - 1.5V may be the solution to overcome it. From the reports here, that appears to be the case and we're going to be on this very quickly.

Any definitve results to share?