Wont boot when warm - new AMD Ryzen 7 5700G - Gigabyte B550M Aorus PRO-P

BlindedByScience

More Human than Human
Joined
May 26, 2000
Messages
9,225
Really scratching my head over this one. I wanted a machine for my office, doing some audio processing and circuit simulations, nothing too expensive so the integrated graphics were appealing. Put together a system with a AMD Ryzen 7 5700G, on a Gigabyte B550M Aorus PRO-P with 32Gb of Team Teamforce Dark-Za 3600MHz memory (two sticks). Flashed to the latest BIOS (thanks to Gigabyte's Q-Flash), no OC. System powered up and ran great for a couple weeks. Thought the CPU was getting a little warm, so I put a Noctua NH-U9S on, replacing the 'stock' AMD heat sink.....and the damn thing wouldn't boot. No fans, nothing. Ugh.

So, I looked everything over, figured I'd pulled something loose, something dumb....and I couldn't find anything. Finally, quite frustrated, I pulled the power cord, pulled the CMOS battery, and put the CMOS memory jumper on the thing and called it a night. Next morning, it fired right up. Went into the BIOS and re-enabled XMS. Warm booted like a champ. Ran perfectly. Temps, everything looking very good. Being cautious, after running it for an hour, I powered it down. Guess what....no boot, dead as a hammer. Did the same trick; pulled power cord and battery, shorted the reset CMOS pins, called it a night. Of course, in the morning it started right up. This time, no XMS enabled. Ran great. Ran for a couple hours. While out enjoying a cigar, the system went into standby. Tried to come out of standby.....nothing. DOA. This time, I just pulled the AC power cord, left the battery and CMOS alone, and let it sit overnight. This morning, it fired right up.

Because it runs perfectly when it runs, my thinking is there must be a MoBo issue..?? Any thoughts would be really appreciated..!!
 
Seems obvious....haven't tried it yet. But I will.

I cannot imagine why and how the HS could cause this issue.......kinda like changing your tail light bulbs and having the transmission fall out.....:eek:
 
Seems obvious....haven't tried it yet. But I will.

I cannot imagine why and how the HS could cause this issue.......kinda like changing your tail light bulbs and having the transmission fall out.....:eek:
odd pressure maybe.
maybe try reseating the chip, keep good pressure on the center while closing the arm.
 
I really appreciate the suggestion. Seems obvious, but I can't imagine the failure mechanism with the Noctua....and I run Noctua's in all the machines here, and like them a lot. I was careful taking the OEM HS off, let it run for a while to get it warm, came off without any violence. CPU looked fine, felt around the edges and the install seemed as flush as it could get, so I left it alone.

Let me pour a cup of coffee, I'll swap the HS's back. And I'll let you know.
 
Brand new Corsair CX650M.

Typing this on the system. Tried something; removed the HSF, removed the CPU, made sure everything was CLEAN (it was). Pulled and reseated the RAM just because it was easy. Put the CPU back in there, applying pressure to the center of the heat spreader while I latched the lever. Put the Noctua back in there.....and it fired right up. Huh. Well, let me fiddle with it for today. If it acts up, I'll swap the HS's and put the OEM back in there. But, for now, it's so far so good.

Bizarre......
 
Brand new Corsair CX650M.

Typing this on the system. Tried something; removed the HSF, removed the CPU, made sure everything was CLEAN (it was). Pulled and reseated the RAM just because it was easy. Put the CPU back in there, applying pressure to the center of the heat spreader while I latched the lever. Put the Noctua back in there.....and it fired right up. Huh. Well, let me fiddle with it for today. If it acts up, I'll swap the HS's and put the OEM back in there. But, for now, it's so far so good.

Bizarre......
ive seen several people have issues like that. so when i built my sig rig with a 2600x i made sure to do it just in case and again when upgraded to the 5600x. its like there is a deformity in the socket or something and the center pins dont get grabbed properly and are just making enough contact that temp affects it(shrink/expand?). thats my theory anyway...
 
Thanks for the suggestions, guys. Really appreciated..!!

I'll power cycle it a few times today, and let you know what's up.
 
Still power cycling, still starting like it should. Maybe, just maybe....this sobered it up. Thanks again for your suggestions...and, maybe I got it. I'll cycle it a few more times tonight, and again tomorrow, and post the results.
 
How's that 5700G working out? Been considering a build with one for my garage office machine to do book keeping and accessing repair data information systems.
 
After multiple power cycles and recoveries from stand by, it's working as expected. REALLY appreciate the suggestions..!! One more tool in the toolbox, one more thing to check when things go bozo.....

learners permit - I'm really digging the 5700G. It's fast as can be, and the graphics are very good. Good enough to play COD2 perfectly. In spreadsheets and other work, it's a fast, capable processor. Since decent graphics cards at decent prices are basically not-existent, this makes this a pretty easy to recommend solution. I did go for 32Gb of RAM, easy to recommend that too. Very happy overall with the performance of the machine.
 
Just started having this problem. I'll reseat the CPU and see what happens.
 
Just started having this problem. I'll reseat the CPU and see what happens.
I didn't reseat when I last posted. I'm using a 3600 and Giagbyte B550. I unplugged the PSU and held the power button to bleed off any caps or static. Booted fine, until today. Now nothing I do will get it too boot. Also, taking out ram and using a beep speaker produced no beep codes. I just sent a B450 back for the same reason, no post, no video using the same 3600. The B450, however, lasted two years. The new one, about a month.
 
I messed up my trouble shooting steps, but I reseated the same card, same result, RAM swap, yadda, yadda. It wasn't even posting. Reseated the CPU (while putting pressure in the center of it, and it posted, but without video. I know it posted because all of the ARGB lights came on, and came in in all default mode, which means the BIOS did get reset, when I reset it. I reset it, and then tried booting, and same thing, no post. I also tried a beep code speaker, not no beeps at all, even with the card and ram out.

After I reseated the CPU it, as stated, it posted, but without video and I know it wasn't hitting Windows because the lights stayed at default. When it boots to windows, it changes. So I swaped in an old card, and in booted into windows. Swapped back in the old card, and it booted to Windows.

So at this point, I have no idea. Obviously something is wrong, and my bet is the MB. It wouldn't have mattered if I hadn't fucked up my trouble shooting steps because something is wrong, imo, with the MB, but it's not anything that is linear. My best is some sort of manufacturing defect that I will never track down. This is the second Gigabyte ITX board I've had similar problems with, with different cards in both.

Note: Yes, I tried diff cables, monitors, RAM swap, etc.

I'll follow up if I have any other problems. I've read about similar problems with the B450 and 550 boards, not all Gigabyte, either.

I really have no idea. It might even be a bad solder trace or connection on the board itself.

UPDATE: Booted twice, then no video again.
 
Last edited:
Back
Top