asrock b450/1600x/blue screens help
Hello. In October, a friend gave ryazhenka and a Geil bar, bought an ASRock B450M Pro4 motherboard for it, and there is a patriot screen bar, I did it at 3.9, the voltage was 1.37000, I didn’t touch anything, I threw the memory into hmp 3200 everything started, everything worked until yesterday. Blue screens and stop screens flew, the computer went off, made 3 squeaks, turned on without overclocking and with 8GB of memory, with an additional reserve of 8GB, poked the bars back and forth, 2133 started on the drain, then it seemed like with overclocking everything turned on again, a little bit drove the tests and left to sleep, today the same picture sat down again, stop screens and blue screens, google codes the first reason is something with memory, while like aid and a meme test, I turn on nothing, I’ll literally play for 5 minutes and immediately catch blue screen. Tried to turn on with 1 bar of a patriot, went into the game, flew out of it and again blue screen, either I ditched something, or I don’t know what to do and where to go. Periodic attempts to set memory above 2133 end with 3 squeaks, a hint of problems with geil? It starts at the lowest frequency, but lasts no longer than 10 minutes.png)
.png)
.png)
.png)
So I have and so after resetting the BIOS, everything is on the standard memory 2133, 3.6 percent did not work out
Monartov
Monartov wrote:
...I turn on the meme test, nothing pops up... The
bars are different, the problem may be because of this. As an option - one of them started to fail. Was it tested by MemTest (86 / 86+) not from under the OS (there is MemTest64 that works in the OS), separately each of the modules, at least one "run" (completely all tests), without XMP enabled?
SonyK_2 This
morning I spun two sticks on the drain in 1-3 slots, on the 3rd round I got 244.016: Error at 4F74D70
244.032: Error at 4BDDEF8
244.063: Error at EB6950 these errors, today I plan to throw a memtest on a flash drive and spin
ClearSkyForUs
Threw 1 bar in 1 slot, both started up, how much time do you need to drive them and which program is better?
later I put it in 1-2 slots and just sat for 2 hours, caught a crash with a blue screen, returned again and spent another 2 hours off the computer
Tested further put it in 1-3 slots threw overclocking, in the memtest on the 2nd loop the stop screen occurs, I do a reboot 3 squeaks - overclocking reset and so on in a circle, on a clean Windows it turned out the same.
Initially, there were 2-4 on the included hmp
Another question in my board, the slots go a1 a2 b1 b2 - paired a1 b1 / a2 b2, for example, when the memory was in a2 b2, the drum information shows slot3 / slot4 to choose from, it’s not logical
.
Monartov
I did not delve into the mat part, but the pairing of slots through one is some kind of feature of the AM4 platform. According to the instructions, it is recommended to install memory in A2 / B2 (slots 2 and 4).
If you put memory sticks one at a time, just to check the stability of each one individually, then it's better to put it in the A2 slot anyway.
You can look at the Asrok website for instructions for your motherboard, it says exactly which slots and which RAM to install (you have it peer-to-peer if that).
For tests, I personally used TestMem from 2017 from the TechPowerUp website and Prime95.
Monartov
Different bars and different timings, you need to clearly set the same timings at a certain frequency, it's not a fact that they will take it!
Yes, right now, the question of system stability at stock frequencies is in question
Monartov
Monartov wrote:
...twisted two bars on the drain...
For starters, it would be nice to "twist" each of them (separately) on the drain. If some module will give errors - most likely the problem is in it.
... I think I'll try to check each one for all slots
. This will be the best - longer, but more reliable. If there are errors in some slot, but not in others, first check if the contact has disappeared. Blow it out of dust, but not with your mouth (so that drops of saliva do not get there) but with a special can of compressed air, or a clean and dry syringe bulb:
Before blowing, make a few clicks, pointing the tip down (away from the computer). If there is something in it, it is more likely that it will fly out.
When the contacts in the slot are oxidized, you can first try to insert one of the modules there several times. If there is a faulty one, use it so as not to erase the coating from the normal one once again.
I drove yesterday in the second slot (a2) from under the bios with a memtest flash drive for 2 hours on each 4 tests without errors, so I’m thinking today to try to set the bar in b2 and drive both already, yesterday I went through the eraser again, walked through the connectors with a brush
I gave it yesterday to the lads who checked it on Intel, they say they did not reveal any problems
On the path of least resistance, you can check one of the bars in b2 in order to eliminate problems with the slot, I'm already tired of pulling them back and forth (
Monartov
If you have one bar in the A2 slot, everything is fine, then the RAM itself is in order. And I highly doubt that the B2 slot is damaged, although putting one bar into the B2 slot and running the test with it alone is a good idea.
When you set the XMP profile in BIOS (on AMD it is DOCP), it automatically sets the frequency / base timings / voltage for both bars, and due to the fact that they are different and work with different timings and most likely with different voltages, you get: such blue screens of death and freezing. Therefore, you need to sit for a while at this stage.
The first thing I would do: in the BIOS I set the RAM to work not "DOCP", but "Manual" (or manual control). After that, I manually set the speed of the RAM to 2666/7 (according to your screenshots from the first post, this is the minimum speed of one of the memory sticks), after which I set all the timings to "automatic" mode.
And similarly with the voltage in the "auto" mode.
I still think to run them together in stock through a flash drive on a2 b2 because the last test for Windows on slots a1 b 1 ended with errors on the 4th loop and the system freezes on 5, judging by the information from the Internet in general, it is not worth putting a full test from under Windows memory
It’s also unlucky that another die with filthy chips, but according to general tests, they both should take 3000 at least
Well, they gave me the geil with a process, and I bought the viper out of inexperience, I saw 3200, which corresponded to the geil and took it, well, they still worked fine for half a year, I'm afraid that I ruined something at the hardware level, but the percentage in general is higher and takes it, the only thing for the future is to give him more loyal tension