Z170 Extreme7+ Sleep Issue
Printed From: ASRock.com
Category: Technical Support
Forum Name: Intel Motherboards
Forum Description: Question about ASRock Intel Motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=3786
Printed Date: 22 Jul 2025 at 1:32am Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: Z170 Extreme7+ Sleep Issue
Posted By: Sparatan117
Subject: Z170 Extreme7+ Sleep Issue
Date Posted: 12 Nov 2016 at 11:52am
Alright, here's hoping someone on the forums knows enough to maybe help me with this issue. I've been through the RMA department and tech support department, had messages sent to "HQ" only to come back with "Idk what you're doing, it works on our end", despite using the wrong RAM after I told them exactly which type of RAM to use.
Here's my current build:
Intel Core i7 6700K @ 4.0GHz 32GB 3200MHz Corsair Vengence LPX 16GB 3200 http://www.newegg.com/Product/Product.aspx%3FItem=N82E16820233859" rel="nofollow - http://www.newegg.com/Product/Product.aspx%3FItem=N82E16820233859 Nvidia GTX 1080 8GB (EVGA ATX 3.0) 600W Corsair PSU.
Current problem:
When putting the computer into sleep mode, it will not recover from sleep. Upon trying to restart from sleep, the board flashes memory errors and after going through the loop a few times states that there is no memory present on the board.
https://www.youtube.com/watch?v=RePIFa35bSE
However, booting the computer from shutdown works just fine. Startup and shutdown work as expected.
I've talked with support and told them my situation. After going through all the testing steps to isolate the problem, we finally decided to RMA the board. I sent it off. Around 2 weeks later, they sent me back a new one. No dice. Same problem remained.
I told the rep that the problem remained. We went through the steps to try and replicate the problem again... (which I had already done and tried other configurations to no avail). I read online that overclocking the memory worked for another guy with the Extreme6, and that he was having trouble with sleep also. Didn't work. Lastly, I simply asked him to send a note up to HQ with my board and my RAM and asked the development team for the BIOS to personally look at it. it came back that they used the same memory configuration but they used a differnet brand of memory to conduct the test. It performed as expected. We then decided to eliminate the bios chip as the source of the problem, and they sent me out a new bios chip after 2 weeks and checking back with them to see if they had sent it (they hadn't). Plugged that in and it wouldn't even post to the Logo screen much less post.
I'm not sure where I put the old Bios chip so I'm currently running off the B chip. Quite frankly I'm fustrated with this as I use sleep mode a lot.
I'm really hoping that someone else has an idea or that someone from the mod team will be able to say something to the BIOS team that it needs an update. I probably won't keep this board if this doesn't get resolved, and just buy an EVGA board so I don't have to deal with this. At least their customer support has a phone line I can call, and not just email support tickets. ( the person I originally spoke with and was emailing back and forth with, didn't even remember me or the situation, or the fact that we had spoken before )
I'm willing to try different things at this point, but I see no reason to toss out perfectly good working RAM in the name of keeping a motherboard that is at fault.
|
Replies:
Posted By: wardog
Date Posted: 12 Nov 2016 at 12:13pm
" rel="nofollow - A new build I assume by your words .......
Your two 16GB(2x8GB) Kits, are they all 4 consecutively serial numbered?
Using only two sticks from the same Kit, shut it down, change to only these two sticks, reset the BIOS, and fire it up. Post back the results please.
|
Posted By: parsec
Date Posted: 12 Nov 2016 at 12:58pm
" rel="nofollow - Your link to your memory at Newegg, has this on the Overview page:
Designed for High-Performance Overclocking on Intel X99 Motherboards.
That would be the CMK16GX4M2B3200C16 kit, of two 8GB DIMMs? So you are using 32GB, combining two 16GB unmatched kits?
Of course, Corsairs website magically states this memory is for Intel 100 series chipset boards. So what is correct, Corsair?
Corsair does not recommend any 32GB, 3200 speed memory kits for your board, according to the Corsair Memory Finder tool (sorry, the forum link too does not work with some URLs, just highlight the link text, left click on it, and select an Open option):
http://www.corsair.com/en-us/memoryfinder?type=motherboard&slug=Tx648d
Please use this tool to above to find compatible Corsair memory kits. Start with the By Motherboard option, then Manufacture, and finally board model.
The memory support list for this board shows a 16GB kit of similar memory, but that is a 4 x 4GB kit.
I also have an ASRock Z170 Extreme7+ board, which I use with G.SKILL 3200 memory. My PC wakes from Sleep perfectly every time, I just woke it after an eight hour Sleep. But then I only have 8GB of single sided memory.
Your memory is double sided, so more difficult to deal with by the memory controller in your CPU. Add to that the large amount of memory you have, from apparently two unmatched kits. You're fairly lucky that it usually works.
You know that all Skylake processors are specified only to work with DDR4 memory up to a speed of 2133? 32GB at 3200 is a pretty high OC, and any OC is not guaranteed.
We can try some tweaks to see if we can get your memory more stable when waking from Sleep, but it seems it takes the memory testing and training that occurs during POST (and does not occur when waking from Sleep) to make it stable.
No idea if you are using the XMP profile of your memory, or entered the speed, timings, etc, manually.
What do you see as the VCCSA and VCCIO voltages currently in the UEFI/BIOS? They are shown in the H/W Monitoring screen.
In the DRAM Timing Configuration screen, on the very last screen, there is an option called MRC Fast Boot. Try setting that to Disabled, which will cause more memory training to occur during POST.
The usual default DDR4 memory speed is 2133. Can your memory wake from Sleep at a speed below 2133?
Have you checked the Windows Event Log for any entries that might give you other clues regarding why the wake from Sleep fails?
I'm curious about this thing about the board "flashing memory errors"? If you mean the Dr Debug codes, those are simply POST codes, and are not error codes unless a code remains displayed when the PC fails to start/boot. That code indicates the POST test that failed. What single code remains displayed when the board fails to wake from Sleep?
Regarding the UEFI/BIOS chip you were sent that did not seem to work. If you put that chip in the A UEFI socket, and then run the UEFI Backup tool in the UEFI, you should have a working A UEFI again. If your board is like mine, you will have an early UEFI version in the B UEFI chip. You'll want to update to a later UEFI version that may work better with all your memory.
------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: Sparatan117
Date Posted: 12 Nov 2016 at 1:08pm
wardog wrote:
" rel="nofollow - A new build I assume by your words .......
Your two 16GB(2x8GB) Kits, are they all 4 consecutively serial numbered?
Using only two sticks from the same Kit, shut it down, change to only these two sticks, reset the BIOS, and fire it up. Post back the results please.
|
Hey thanks for the quick reply.
I've done everything from removing everything but a single stick, to RMA'ing the Power Supply, Changing out the Graphics Card (980 -> 1080) and also removing the Graphics Card and using the onboard GMA.
|
Posted By: Sparatan117
Date Posted: 12 Nov 2016 at 1:23pm
parsec wrote:
" rel="nofollow - Your link to your memory at Newegg, has this on the Overview page:
http://www.corsair.com/en-us/memoryfinder?type=motherboard&slug=Tx648d
Please use this tool to above to find compatible Corsair memory kits. Start with the By Motherboard option, then Manufacture, and finally board model.
|
http://www.corsair.com/en-us/vengeance-lpx-16gb-2x8gb-ddr4-dram-3200mhz-c16-memory-kit-black-cmk16gx4m2b3200c16
After selecting ASRock and Intel Z170, and then the Extreme 7+, that was the second result on the list (I did at least try and go look this up before putting it together initially )
parsec wrote:
Your memory is double sided, so more difficult to deal with by the memory controller in your CPU. Add to that the large amount of memory you have, from apparently two unmatched kits. You're fairly lucky that it usually works.
You know that all Skylake processors are specified only to work with DDR4 memory up to a speed of 2133? 32GB at 3200 is a pretty high OC, and any OC is not guaranteed.
|
It doesn't even work with a single stick of RAM, let alone all 4 of them. The result is the same and the error codes thrown. I've tried bumping the voltage to it just to see if that would power the dual sided-ness of it... to no avail.
parsec wrote:
We can try some tweaks to see if we can get your memory more stable when waking from Sleep, but it seems it takes the memory testing and training that occurs during POST (and does not occur when waking from Sleep) to make it stable.
No idea if you are using the XMP profile of your memory, or entered the speed, timings, etc, manually.
|
I currently have it set to AUTO. I've only got one other profile on there that was on there by default... and I manually overrode the MHz setting in there from 2333 to 3200 ... I haven't really seen a difference between the two though.
parsec wrote:
What do you see as the VCCSA and VCCIO voltages currently in the UEFI/BIOS? They are shown in the H/W Monitoring screen.
|
I'll have to get these for you after writing this.
parsec wrote:
In the DRAM Timing Configuration screen, on the very last screen, there is an option called MRC Fast Boot. Try setting that to Disabled, which will cause more memory training to occur during POST.
|
Did that already. No dice.
parsec wrote:
The usual default DDR4 memory speed is 2133. Can your memory wake from Sleep at a speed below 2133?
|
I've never tried underclocking it... I'll give that a shot.
parsec wrote:
Have you checked the Windows Event Log for any entries that might give you other clues regarding why the wake from Sleep fails?
|
The POST never really happens and the board never even manages to turn the monitors on, so I don't think Windows would even get a notification that the recovery from sleep was a failure. I could be wrong though.
parsec wrote:
I'm curious about this thing about the board "flashing memory errors"? If you mean the Dr Debug codes, those are simply POST codes, and are not error codes unless a code remains displayed when the PC fails to start/boot. That code indicates the POST test that failed. What single code remains displayed when the board fails to wake from Sleep? |
5d is shown initially. Following that is a repetition of 19 - 0 - 19 - 0 and then finally resets on 55 which is "memory could not be detected".
parsec wrote:
Regarding the UEFI/BIOS chip you were sent that did not seem to work. If you put that chip in the A UEFI socket, and then run the UEFI Backup tool in the UEFI, you should have a working A UEFI again. If your board is like mine, you will have an early UEFI version in the B UEFI chip. You'll want to update to a later UEFI version that may work better with all your memory.
|
I've updated the UEFI to the latest version via the internet backup. I'll try and reflash the other chip to see if that makes it work. However, the B and the original A act as the same and had the same effects so... I'm not sure there.
And yeah like I said, it boots up and runs just fine. It's run everything I've thrown at it and the board seems to do a great job otherwise. It just never manages to recover from sleep. I've tried disabling Windows Fastboot to see if that would cause an issue and even went as far as disabling hibernation mode and anything else that windows could have. Tech support had me install a brand new fresh copy of Windows 10 on a hard drive and boot from that to see if it would isolate the problem. It never did. I've also gone and thrown on a copy of Ubuntu, and tried to make that reboot from sleep. It too, failed.
I'm not sure if there are logs that are stored on the motherboard itself but I would love to know it's thoughts if they're there. Maybe that would give us some insight as to what it's doing... or rather... not doing.
|
Posted By: Sparatan117
Date Posted: 13 Nov 2016 at 4:33am
Alright Back with an update. I tried setting the frequency to 2133 or less throughout multiple settings all the way down to 1600 MHz... with no success
Here's a picture of the voltages for VCCSA and VCCIO http://i.imgur.com/jDPHogq.jpg" rel="nofollow - http://i.imgur.com/jDPHogq.jpg
|
Posted By: Sparatan117
Date Posted: 01 Dec 2016 at 12:47pm
Posted By: wardog
Date Posted: 01 Dec 2016 at 1:17pm
" rel="nofollow -
Sparatan117 wrote:
bump |
Yea. Bump.
Pull everything connected from the board that is not req'd to make it into the BIOS, mem (edit hdd/ssd) cpu+cooler case wiring to system panel header, and retry.
Too, check you wiring at the system panel header. Some case and MB combo's can get confusing here.
Did you list the PSU and it's age? Could be low voltage on the +5v(+3.3v?) rail
Failing those, that I'd be looking at here, pull the board out of the case and bench it.
Then the ugly. Pull the CPU and take a close inspection of the socket "pins"
|
Posted By: Sparatan117
Date Posted: 07 Dec 2016 at 3:20pm
I appreciate the taking a stab at it.
I've actually RMA'd the power supply right after I got it, due to coil wine when it was under load. It's a Corsair CX750 (which should be plenty). I've disassembled the board all the way down to the core components ( CPU / RAM / SSD with Fresh Copy of Windows 10 / Onboard Graphics ). It won't come back from boot following that. After we trouble shot everything, I RMA'd the board to ASRock and they sent me another one... same test conditions, same results.
The only thing I haven't tried doing is hooking another PSU into the board and trying to do it from that. I don't exactly have another brand of PSU to test it with at the moment so ... yeah. I don't think it's that though. Oh and it could be the CPU... but... if the CPU doesn't have any issues on boot, then it shouldn't have any issues on sleep recovery.
I think I mentioned it, but the last thing they did was send me a BIOS chip replacement. I did wipe that over and do an internet update on it and then switched to it... it's working fine off the chip now... but sleep is still in-op
I think the only options here are to do a RAM swap (brand new ram from GSkill or someone) or just start looking for another board (which would be cheaper than replacing 32GB of DDR4 3200MHz RAM).
The thing that really rubs me the wrong way is that ASRock acts like it's something I'm doing wrong or that I'm lying about the situation. I don't know that I'll ever buy another board from them after this experience though.
|
|