Out of graphics card, I don’t doubt that, but same can happen on motherboard too. This can be caused by a few things, first by incorrect attachment (ie wrong cables to wrong pins in the programmer or board), or by incorrect software version for that chip, and sometimes by using incorrect chip ID (Sometimes you have to pick close, not exact, or similar)
Is this on the motherboard? If yes, what is the exact chip ID, and I will test if I have it, and let you know which software version works best.
I realized that the problem is my grip thing. It has 2 broken cables. I can solder that but didn’t found the time to do it. Chip is MXIC MX 25U12873F
That is 1.8V chip, you need 1.8V adapter before you can read/write to it, like this one - https://www.ebay.com/itm/253208845481
Oh Soo I miss that. And I still need to repair the clip… Is that compatible with the CH341A USB programmer?
Yes, it’s compatible and that is what it’s designed for. The bottom is just like asus/asrock type BIOS (8 pin DIP), it will connect right into the programmer like those socketed BIOS or like your SOIC cables
It’s like these SOIC adapters for soldered on type BIOS but used without solder during programming, these below are drop in spring holder/adapter for programming SOIC8 style flash roms (150mil GPU, 200mil motherboard)
Only the 1.8v adapter is larger due to the components on top of the PCB - https://www.ebay.com/itm/263278959546
I was hoping to make a request for all of the zen 1 gaming k7 users. F10 supposedly gives the best memory results for zen1 chips but lacks several of the features and options a mid range mobo bios should have. Mainly p state and several excluded memory options. Every bios I have tried since for pinacle pi agesa has had either bugs that make it unusable or worse performing. Secondly I was wanting to ask if the most current version is based f23d or the finalized version of f23? I ask because the official bios for f23 final was either uploaded wrong or has major problems. Because without fail causes uefi corruption for me. And unfortunately something in the modded k7 bios here performs much worse for me on the memory front on b die dimms.
In all honesty i bought this board refurbed because of how crappy the asrock b350 bios i had was, and because it had supposedly the best memory support out of gen1 boards and I’m at my wits end with it. I think its possible that both my bios chips are corrupted at a cell level but havent confirmed yet.
Make just one ROM chip active that should stop the firmware corruption it has for everyone I’ve suggested it to who had the same problem, myself included. Firmware I modded here is based on F23 final and with some work got up to 3466MHz stable with some Sammy B-Die so can’t help regards the issue you’re having there.
I’ve tried both bios, in single bios mode and they both fail regardless. IDK my boards been acting really weird since i switched to b die. I could run 3400 stablish on 2666 hynix mfr kit. but i run it on this b die it will crash almost immediately. My kit has been acting weird ever since i purchased this b die. I think I might try and rma it. The most I can get stable on them is the stilts 3333mhz fast on 2t gdm off. Damn things refuse to run with gdm off and 1t with every timing strap ive checked except 2133.
Or maybe, adding to what ket mentioned, does flashing both roms to same version stop the corruption issue too?
Check that your SPD hasnt been corrupted in the ram
I have a 1st gen Ryzen (1800X) and an X370 Gaming K7.
I’m running 4x 8GB b-die modules (two kits of GSkill F4-3200C14D-16GTZR) with XMP enabled (3200, 14-14-14-34, 1.35v) and even tightened up some timings further (thanks to Ryzen DRAM Calculator) and the system runs wonderfully. I’m on BIOS F7aM, and was running F7a official before, and have had no problems. I did have one soft brick situation very early on, but I chalk that up squarely to Gigabyte’s half-baked automatic dual BIOS functions. I’ve since moved to single BIOS mode and manually flashed both main and backup to the same version and then done a complete cold system reset (power off, unplug, clear CMOS, then pull battery for ~10 min). No BIOS weirdness issues since.
Anyway, gave F10 a try when it first came out, but didn’t like how it fed my CPU quite a bit higher voltage on auto, so I returned to F7a. Then after reading about all the issues that have plagued later releases, I decided to just hang out on F7a until I actually need to update, which probably won’t be until I think about putting a Zen2 / Ryzen 3xxx chip in the board down the line at some point.
I don’t overclock the CPU, as frankly I’d just be wasting power and hammering the chip with constantly high voltage for ~200MHz higher performance. Precision Boost plus XFR for 3.7GHz base and 2-core 4.1GHz boost is fine for me if it means keeping the great power efficiency features of the chip enabled.
Anyway, with a 1st gen chip I’d absolutely give F7a or F7b (if you need/want Pstates) a try. It was the last version with the 1st-gen focused AGESA before support for Raven Ridge was added. Since disabling dual bios mode and completely clearing things out it’s been nothing but great for me.
@redtopracer it sounds very possible that one of your memory sticks is bad, run HCI memtest for at least 3hrs to see if you get any errors. Which slots do you have the RAM installed in? If you put them in the same banks as your Hynix modules that there could be the problem swap your B-Die to the other set of unused DIMM banks. Have you completely erased the ROM chips from an EFI shell and reprogrammed them with my modded firmware? I’d consider that worth doing as well (instructions in first post) as a last resort you could also get a soft eraser and clean both sides of the gold contact pins on the memory modules, they shouldn’t be dirty anyway but sometimes even pins that look clean just need an eraser over them to clean them up where their extra sensitive to grime.
Well Ive been screwing with them for 2+ weeks now and can only get 3400 14,15,14,30 stable with some atrocious secondary and tertiary timings. And even then I cant get them stabilized past 4000% on ram test. They still throw single errors every couple 1000%. Think im gonna do a hci test overnight on stock like you recommended. Ive been reseting the cmos after each failed mem training and its helped quite a bit. Ive only had to reflash the bios once in the past week doing it that way.
I think either I’ve got a bad stick, bad bios , or my memory holes stop right after 3400. Either way this has been a grind thats for sure. Ill look into your suggestions ket. The help is much appreciated.
Do you have any experience with cldp0_vddp voltage? Failing the previous suggestions thats the only thing left to try for better frequency. Otherwise I’m stumped on tweaking these sticks lol.
Try one at a time fella, run each one for an hour on memtest
Are you sure your CPU is stable, is it overclocked? HCI Memtest includes CPU Stress as well. If you are not 100% certain, take your CPU to stock and leave ram where you have now, then you’ll only be testing your ram OC/settings
@ket testing at stock speeds got me to a 1000% on HCI and swapping banks on the DIMMS did nothing either.Going to reflash the bios like you suggested and do a clean windows install, just in case i’ve corrupted the OS files. Checking the gold contacts there spotless except for the latch mark on them from inserting in the channel.
@Lost_N_BIOS The cpu overclock is rock solid I can actually run 4.2ghz at 1.44v but thats to high for daily use. Stability is my goal here so rest assured, everything else is sorted out on this rig. And just in case it had degraded or something I have ran these test at stock cpu speeds.
At this point I’ve already sunk 2 weeks into overclocking these things, think Im going to fresh install my stuff and just enjoy the 3400 14,15,14,30 i have stable. Wait for gigabyte to get off there ass and give us a new AGESA revision. pinnacle pi 1006 is already out on msi boards, and supposedly gives pretty decent improvements over the last used one.
Unfortunately just because someone else has 3466 ram at cas 14,13,10,11 doesnt mean you will, its all luck of the draw, 1.44v isnt too high thats about right. at stock the cpu can easily spike upto 1.5v
for example my 2700x runs at 4250mhz with the ram at 3520Mhz cas 14 and 53.3ohms, will even run 3600 at cas14
remember CBS pstate0 will automatically drop the vcore down to 0.4 to 0.8v
@redtopracer if you are using my modded firmware press the "Alt" key on a selected item in the firmware and you will get rewritten help information, actual help info that is, not the vague default crap GB give as "help".
@ket I know the function and purpose of the cldp0_vddp voltage and what it does. I was curious if you knew of a better way than just trawling threw the entire spectrum?
At any rate i actually need help. Whenever i try to flash using your dos method I get error 46 problem getting the flash information.
what am i missing here to flash this?
@redtopracer if I recall you’ll get that error with stock firmware because by default overwriting the boot block is locked, that issues is addressed with the modded firmware. Have you tried flashing leaving the switch out to flash the boot block? If you still get the error then you might need to update the EFI shell HERE is the link for that. As for good CLDO_VDDP voltage, 725 should work well for B-Die. You could also see what your VDDCR_SOC voltage is and try raising / lowering it. I wouldn’t go above 1.15v and you shouldn’t need more than 1.1v for the 3466 - 3600 range. You might also be able to drop vcore with an increased VDDCR_SOC.