Asus Maximus XI Code - after delidding 9900k won't POST with 3200mhz DDR4 (unchanged hardware)

[Kalm_Traveler]

So here's a funny one... finally got around to delidding my 9900k to use der8auer's direct-die frame and temperature-wise all seems well. Playing games and in benchmarks I haven't seen any core pass 62c (prior they could reach 90's at 5.2ghz on 1.33v).

 

However, I let the machine sit on overnight and woke up to it on but not running (display wouldn't come back up etc).  on reboot the LED would stop at b4 and never finish POST.

 

I've found from troubleshooting so far that it just won't POST anymore with the default XMP 3200mhz settings on this G.Skill Trident Z RGB 3200 CL14 pair of DIMMs which it's been using for months. If I drop the RAM to default DDR4 settings everything boots and runs fine.

 

I've OC'd the CPU even higher and that's all fine too - it only hangs if I increase the RAM speed (even tried setting the timings to 14-14-14-34 and voltage to 1.35v per the XMP profile but leaving the actual speed at Auto, which again is fine and stable).

 

At this point I'm really not sure how just removing the IHS and solder from the 9900k would have changed the RAM stability like this, I didn't even remove the 2 sticks until troubleshooting why it stopped POSTing.

 

Do you guys have any ideas? I'm almost at the point of ordering a different motherboard just to see if somehow the socket is unhappy without its CPU retention mechanism or something.

[Lord Nicoll]

19 minutes ago, Kalm_Traveler1 said:

So here's a funny one... finally got around to delidding my 9900k to use der8auer's direct-die frame and temperature-wise all seems well. Playing games and in benchmarks I haven't seen any core pass 62c (prior they could reach 90's at 5.2ghz on 1.33v).

 

However, I let the machine sit on overnight and woke up to it on but not running (display wouldn't come back up etc).  on reboot the LED would stop at b4 and never finish POST.

 

I've found from troubleshooting so far that it just won't POST anymore with the default XMP 3200mhz settings on this G.Skill Trident Z RGB 3200 CL14 pair of DIMMs which it's been using for months. If I drop the RAM to default DDR4 settings everything boots and runs fine.

 

I've OC'd the CPU even higher and that's all fine too - it only hangs if I increase the RAM speed (even tried setting the timings to 14-14-14-34 and voltage to 1.35v per the XMP profile but leaving the actual speed at Auto, which again is fine and stable).

 

At this point I'm really not sure how just removing the IHS and solder from the 9900k would have changed the RAM stability like this, I didn't even remove the 2 sticks until troubleshooting why it stopped POSTing.

 

Do you guys have any ideas? I'm almost at the point of ordering a different motherboard just to see if somehow the socket is unhappy without its CPU retention mechanism or something.

What was the VCCSA voltage? I super doubt it got degraded at 3200MHz, however it could also be uneven mounting pressure in theory, 1151 pins is a lot of pins, even one not making good contact could cause a fault, the B4 is a USB hot plug error usually though. 

[Kalm_Traveler]

41 minutes ago, Lord Nicoll said:

What was the VCCSA voltage? I super doubt it got degraded at 3200MHz, however it could also be uneven mounting pressure in theory, 1151 pins is a lot of pins, even one not making good contact could cause a fault, the B4 is a USB hot plug error usually though. 

on auto I believe it is 0.975 but I tried setting it to 1.0v manually, also tried setting VCCIO to 1.0. Yeah I really doubt it suddenly degraded right at the same time (this machine is less than 4 months old IIRC - the RAM is less than a year, I bought it when the 8086k came out last year). MB and 9900k were purchased in November last year.

 

Definitely could be uneven mounting pressure but I'm not sure how to account for that. I tried removing the waterblock and die frame, removed the CPU and cleaned off its pads with alcohol just in case something was dirty, reinstalled everything and as far as I can tell it's as even as I can get it to be.

 

Very weird problem though eh? and yes I also saw that b4 is a USB hot plug error... all that's plugged in now are the keyboard and mouse. Works fine as long as the RAM speed isn't raised above stock 2133mhz

 

*EDIT*

It POSTs and runs fine if I drop the RAM speed to 3000, won't POST at 3100 or 3200 though.

[Kalm_Traveler]

In case anyone else searches and finds this, another user on reddit had the exact same issue with the frame, reinstalled the stock retention mechanism and IHS, proper RAM speeds work again so this definitely is a mounting pressure issue.

 

I'm going to try a few things tomorrow to try and get mine to work as is but may need to do the same.

[HoneyBadger84]

I was just about to say this sounds like a mounting pressure issue to me when I saw your latest post.  Pretty crazy that's the ONLY issue it caused, you'd think it would be more wide-spread if one of the CPU pins wasn't making proper contact eh?

[Kalm_Traveler]

8 hours ago, HoneyBadger84 said:

I was just about to say this sounds like a mounting pressure issue to me when I saw your latest post.  Pretty crazy that's the ONLY issue it caused, you'd think it would be more wide-spread if one of the CPU pins wasn't making proper contact eh?

yeah I am still surprised too - had the thing running now for 3 days, played BF V, World of Warcraft, ran a bunch of Cinebench and Time Spy Extreme - no issues whatsoever aside from that I'm having to do so with the RAM at 3000 mhz instead of its rated 3200 mhz.

 

I tweeted der8auer to see if he has any thoughts on it as I'm not sure if the mounting pressure is too much, too low, or the PCB just isn't being held down evenly enough. When the reddit user first suggested this from his, I had thought that maybe I tightened the 3 motherboard mounting screws too far.

 

That's basically what I want to mess with today - try loosening them and see if that helps. Right now they're as tight as I dared go turning 1-handed with the short end of the provided star wrench.

[Kalm_Traveler]

In case anyone sees this from a search - the 3 frame-to-motherboard screws were too tight.

 

After re-watching der8auer's video about it, I noticed that he did not tighten them very much (he stops when they are tight enough to move the free-standing motherboard). After seeing that I pulled my water block, cracked all 3 loose and re-tightened them using only my index finger and thumb to gently twist the short end of the included star wrench just until there was resistance.

 

Remounted the block (thank you liquid metal for not having to redo TIM), fired her up, changed RAM speed in BIOS back up to 3200, save and reboot. Voila, problem solved.