Page 11 of 57 FirstFirst ... 89101112131421 ... LastLast
Results 251 to 275 of 1415

Thread: **Unified** DFI LanParty UT 790FX-M2R-M2RS / Phenom /Data Base Thread**

  1. #251
    Registered User
    Join Date
    Apr 2008
    Posts
    9
    Quote Originally Posted by justapost View Post
    Good news, the board is back.

    I tried tony's reverse battery method. I left the totaly unplugged board with the clear cmos jumper set and the batter mounted revers alone fopr a few hours and now it's working again. Glad I had luck this time.
    Interesting. My old Epox 939 board used to lock up occasionally. I could shock it back to life by removing the battery, setting the clear cmos jumper for about 10 minutes. Then I would set everything back, and boot up while pressing the insert key. The Epox also had a POST LCD, and would lock up at "FF" rather than "C1".

    I recently ran lm-sensors configure while installing Debian Linux on my DFI M2R + 9850BE. It locked everything up and the LED read "C1". I tried my "Epox" technique but it didn't help. I had to RMA both the board and cpu. Since then I've stayed away from lm-sensors and speedfan. Unfortunately those were my two favorite monitoring programs...

    If it happens again, I'll try your method. The difference is that you reversed the battery and let it sit for a long time.

  2. #252
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    Quote Originally Posted by vext View Post
    I recently ran lm-sensors configure while installing Debian Linux on my DFI M2R + 9850BE. It locked everything up and the LED read "C1". I tried my "Epox" technique but it didn't help. I had to RMA both the board and cpu. Since then I've stayed away from lm-sensors and speedfan. Unfortunately those were my two favorite monitoring programs...
    There is no need to run sensors-detect. The only module you need for sensor monitorin is "it87". Seems I have an robus bios chip here.
    The second time I ran into the C1 issue, removing one dimm helped. That dimm made problems for a few minutes and seems to work ok now.
    The old sensors-detect routine will trigger the chip at 0x2e to appliy an >3V vdimm. That should harm the dimm's and the chip's nb. I wonder why it also results in faulty bioses.
    Did you get both RMAed? If so did they say what was defect.
    Quote Originally Posted by vext View Post
    If it happens again, I'll try your method. The difference is that you reversed the battery and let it sit for a long time.
    That was tony's tip from a few weeks ago, whom I crossposted here.

  3. #253
    Xtreme Addict Rammsteiner's Avatar
    Join Date
    Oct 2005
    Location
    EvE-Online, Tranquility
    Posts
    1,978
    I was just gaming with stock settings. Although I had CnQ and Vcore/Vnb still at auto because of MSR OC attempts. But this shouldnt matter.

    But I was gaming and system froze, blackscreen... So I set CnQ off and set the default Vcore and Vnb manually.

    If it happens again Ill try to downclock the GPU (maybe the pencil is degrading on it). If not... Im going mad...

    Then Ill RMA the M2R and Phenom as well. Im even considering to get Crosshair II instead. This is driving me nuts.

    However, not making any conclusions yet. But Im getting a deja vu from my X2 + M2R...

    [EDIT]
    Didnt help crashed again. Downclocked the GPU this time.
    Last edited by Rammsteiner; 05-08-2008 at 07:04 AM.
    Synaptic Overflow

    CPU:
    -Intel Core i7 920 3841A522
    --CPU: 4200Mhz| Vcore: +120mV| Uncore: 3200Mhz| VTT: +100mV| Turbo: On| HT: Off
    ---CPU block: EK Supreme Acetal| Radiator: TCF X-Changer 480mm
    Motherboard:
    -Foxconn Bloodrage P06
    --Blck: 200Mhz| QPI: 3600Mhz
    Graphics:
    -Sapphire Radeon HD 4870X2
    --GPU: 750Mhz| GDDR: 900Mhz
    RAM:
    -3x 2GB Mushkin XP3-12800
    --Mhz: 800Mhz| Vdimm: 1.65V| Timings: 7-8-7-20-1T
    Storage:
    -3Ware 9650SE-2LP RAID controller
    --2x Western Digital 74GB Raptor RAID 0
    PSU:
    -Enermax Revolution 85+ 1250W
    OS:
    -Windows Vista Business x64


    ORDERED: Sapphire HD 5970 OC
    LOOKING FOR: 2x G.Skill Falcon II 128GB SSD, Windows 7

  4. #254
    D.F.I Pimp Daddy Brother Esau's Avatar
    Join Date
    Jan 2007
    Location
    Still Lost At The Dead Show Parking Lot
    Posts
    5,188
    Quote Originally Posted by seeker010 View Post
    how come new bios don't allow nb fid/did change when enabling cool'n'quiet?

    Which new Bios? 4/16/08 Official or 5/5/2008 Beta?
    SuperMicro X8SAX
    Xeon 5620
    12GB - Crucial ECC DDR3 1333
    Intel 520 180GB Cherryville
    Areca 1231ML ~ 2~ 250GB Seagate ES.2 ~ Raid 0 ~ 4~ Hitachi 5K3000 2TB ~ Raid 6 ~

  5. #255
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    I think 3-07 was the lates version where it was possible to change the nb vid/did if cnq was enabled.

  6. #256
    D.F.I Pimp Daddy Brother Esau's Avatar
    Join Date
    Jan 2007
    Location
    Still Lost At The Dead Show Parking Lot
    Posts
    5,188
    @Achim.....question? Through you're personal Experience with Phenom and 790FX do you find that HCI MemTest works well and is accurate for testing Ram Stability in Windows?
    SuperMicro X8SAX
    Xeon 5620
    12GB - Crucial ECC DDR3 1333
    Intel 520 180GB Cherryville
    Areca 1231ML ~ 2~ 250GB Seagate ES.2 ~ Raid 0 ~ 4~ Hitachi 5K3000 2TB ~ Raid 6 ~

  7. #257
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    Quote Originally Posted by Brother Esau View Post
    @Achim.....question? Through you're personal Experience with Phenom and 790FX do you find that HCI MemTest works well and is accurate for testing Ram Stability in Windows?
    Good question.
    I never used HCI memtest, I rely on memtest86+ http://www.memtest.org/.

  8. #258
    Xtreme Member AlabamaCajun's Avatar
    Join Date
    Dec 2007
    Posts
    157
    Ramm, sorry if I missed it but have you visited the screens such as Genie, RAM and PWM and hit F7. I had to do this to get stable. Do RAM first and reboot up to the Save DMI, I think Post 96. Then Restart there and reenter BIOS and F7 the other 2 then reboot. This is the only way I've found that completely clears the controller tables.

    JP, thanks for the Register info, that's GOLD

    I could not resist playing with the clocks again. Still on the 5-X-2008 Beta which for me seems stable.
    I went ahead and hit the volts back to 1.28 after a few days running stock at 1.275. Bumped the lock to 209 set CPU multi back to Auto, not 12.5. It booted and ran all night, folding from 8pm to 8am before I shut it down to let the rooms cool. Night temps are on the low 60s requiring little AC so not much cool air goes there, too muggy for cracking the window.

    2.61G is not too back but I may still be able to get 2.7G+ solid if I can find the best way to cool this thing.
    Last edited by AlabamaCajun; 05-08-2008 at 11:00 AM.
    Main Win64; Phen940BE, Sunbeam Cooled, DFI LP790FX-M2R, 2x2G Corsair, R3850XT
    Production XP; Athlon 6KBE, GByte POS mobo, 2x2G Gskill 800, R3750XT.

    Vista 64 Test; GP9500, ECS740M, 2x2G GSkill, IGP.
    Server; 4 X 8347 Opterons, 8x1G Kingston.
    XP Home Test; A5000BE X2, ECS RS485M, 2x1G Wintec, IGP.
    S&Gs Test; A5000BE X2, ECS A770M-A, A5800 X2, ECS KA3-MVP-580X, 2x1G ?

  9. #259
    Registered User
    Join Date
    Apr 2008
    Posts
    4
    Quote Originally Posted by justapost View Post
    I think 3-07 was the lates version where it was possible to change the nb vid/did if cnq was enabled.
    yes. you are right. do we know why DFI made this change? Also, what does cpu stepping p0 and p1 mean?

  10. #260
    Xtreme Addict Rammsteiner's Avatar
    Join Date
    Oct 2005
    Location
    EvE-Online, Tranquility
    Posts
    1,978
    Quote Originally Posted by AlabamaCajun View Post
    Ramm, sorry if I missed it but have you visited the screens such as Genie, RAM and PWM and hit F7. I had to do this to get stable. Do RAM first and reboot up to the Save DMI, I think Post 96. Then Restart there and reenter BIOS and F7 the other 2 then reboot. This is the only way I've found that completely clears the controller tables.
    Thx for that, didnt know that yet.

    How do you mean stable though? As in less/no BSOD's, no strange freezes? At stock or as well with OC'd settings?
    Synaptic Overflow

    CPU:
    -Intel Core i7 920 3841A522
    --CPU: 4200Mhz| Vcore: +120mV| Uncore: 3200Mhz| VTT: +100mV| Turbo: On| HT: Off
    ---CPU block: EK Supreme Acetal| Radiator: TCF X-Changer 480mm
    Motherboard:
    -Foxconn Bloodrage P06
    --Blck: 200Mhz| QPI: 3600Mhz
    Graphics:
    -Sapphire Radeon HD 4870X2
    --GPU: 750Mhz| GDDR: 900Mhz
    RAM:
    -3x 2GB Mushkin XP3-12800
    --Mhz: 800Mhz| Vdimm: 1.65V| Timings: 7-8-7-20-1T
    Storage:
    -3Ware 9650SE-2LP RAID controller
    --2x Western Digital 74GB Raptor RAID 0
    PSU:
    -Enermax Revolution 85+ 1250W
    OS:
    -Windows Vista Business x64


    ORDERED: Sapphire HD 5970 OC
    LOOKING FOR: 2x G.Skill Falcon II 128GB SSD, Windows 7

  11. #261
    Xtreme Member AlabamaCajun's Avatar
    Join Date
    Dec 2007
    Posts
    157
    Stable is when I can carry out normal tasks, surfing, folding, gaming and not see any glitches. Solid is what I call one that runs weeks without any problems though I will through a game error if it cleanly quits to desktop and reports an error that happens after doing something wild that may have just pushed the game into a buggy area of code. Some of the simulator games can be pushed over the edge. Even though I get suspect of a hardware problem and will be on guard.
    After all the flakiness I saw with the GBT 790FX-DS5 that sits in a box, this is stable.
    Main Win64; Phen940BE, Sunbeam Cooled, DFI LP790FX-M2R, 2x2G Corsair, R3850XT
    Production XP; Athlon 6KBE, GByte POS mobo, 2x2G Gskill 800, R3750XT.

    Vista 64 Test; GP9500, ECS740M, 2x2G GSkill, IGP.
    Server; 4 X 8347 Opterons, 8x1G Kingston.
    XP Home Test; A5000BE X2, ECS RS485M, 2x1G Wintec, IGP.
    S&Gs Test; A5000BE X2, ECS A770M-A, A5800 X2, ECS KA3-MVP-580X, 2x1G ?

  12. #262
    Xtreme Mentor
    Join Date
    May 2007
    Posts
    2,799
    If they removed the option of FID+CnQ, it may be because of a problem with it though, someone should ask them
    They rarely remove it for no reason.

  13. #263
    Xtreme Member aGeoM's Avatar
    Join Date
    Nov 2007
    Location
    Portugal
    Posts
    397
    Quote Originally Posted by KTE View Post
    If they removed the option of FID+CnQ, it may be because of a problem with it though, someone should ask them
    They rarely remove it for no reason.
    The reason is, before you could set it the way you wanted, but after you reboot/boot to windows, it reseted to default, now you can't have it enabled just because it wouldn't work. Makes sense to me. Don't know if my English makes for you.



  14. #264
    D.F.I Pimp Daddy Brother Esau's Avatar
    Join Date
    Jan 2007
    Location
    Still Lost At The Dead Show Parking Lot
    Posts
    5,188
    Update:.........5/5/2008 Beta Run at the same settings a my Bios Tempate Overclock on Front page with Ram @ 1066 OCZ Tony's Modified SPD @ 1.87Vdimm CPU @ 1.23750vcore set OCCT Crashed 4hrs 46 minutes so technically that equals 8 1/2 hrs of Prime95 I am quite positive its due to heat nmore then likely as it is 90F in my Room Currently

    Overall Bios seems OK but needs some refinement is all

    Once again there are USB Keyboard Issues Present in this Bios when in Pre Boot Prior to Post the USB Keyboard does not Respond and you have to Restart Manually to get the Delete Key to Function to get into Bios!

    Also some NIC Latency Issues "Not Much' but it is there.

    Also another Minor Annoyance is that after saving settings and Restarting the Splash Screen / Post Screen does not appear and just is Black until Raid Controller Bios Post appears!


    Can any of you gentlemen Running 5/5/2008 "Beta" with the Board set up on the Bench get a accurate DMM Reading on CPU , NB Voltage please and post offset variants?
    Last edited by Brother Esau; 05-09-2008 at 09:54 AM.
    SuperMicro X8SAX
    Xeon 5620
    12GB - Crucial ECC DDR3 1333
    Intel 520 180GB Cherryville
    Areca 1231ML ~ 2~ 250GB Seagate ES.2 ~ Raid 0 ~ 4~ Hitachi 5K3000 2TB ~ Raid 6 ~

  15. #265
    Xtreme Mentor
    Join Date
    May 2007
    Posts
    2,799
    Thanks aGeoM. It was a problem on many AMD boards, even when they disable it.

    When they leave CnQ and FID changing on together, the clockspeeds through the TSC are read wrongly many times [downclock to read 2500 at all times].

  16. #266
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    The odd thing is that different nb vid's and did's worked fine here with the 9600BE and the 4-07 bios. The problem aGeom described did not occure with all phenoms. !

  17. #267
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    A small update about the phenoms smbus thermal interface.

    On the DFI/Sapphire 790FX boards this interface is accessible like a separate chip at the address 0x4c. It can be read out undel linux with

    i2cdump 0 0x4c
    Code:
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
    00: 00 16 00 00 08 00 00 46 00 00 00 00 00 00 00 00    .?..?..F........
    10: 80 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ?...............
    20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
    30: 47 80 00 00 00 00 00 00 00 00 00 00 00 00 00 00    G?..............
    This is a typical output. I found the value at address 0x01 holds the core temp in celsius. 0x30 is always that temp + 49°C-50°C.
    The value at 0x0a seems to be the max temp the chip is rated for. I only have an 9600BE here and that cpu shows 0x46 means 70°C here.

    However there is another way to read out this interface the smbus way.
    There are two registers on the Miscanselous Control Register (F3).
    Register 0x1E8 is used as an address pointer and 0x1EC can be used to read and write data to that address.
    So if you startup Baredit and select "B.0 D.24 F.3" you can write the desired address to the register 0x1E8 and after you select "B.0 D.24 F.3" again you get the value in 0x1EC


    This screenshot shows the data of register 0x01 of the thermal sideband smbus interface means it's the current core temp (0x18 =>24°C).

    Note: This way of access also works on the GBT780G board with the 9600BE. This board does not offer an i2c/smbus access at 0x4c and on no other address but via the registers 0x1E8 and 0x1EC it works. On both boards registers 0x1E4 has bit 3 set to 0b, which means the thermal side band smbus interface is not deactivated.

    May I ask that someone with an 9850BE or an other phenom different from an 9600BE posts the values of the registers 0x06 and 0x46 here.

    1. Start BarEdit
    2. Select "B.0 D.24 F.3"
    3. Write 0x06 into register 0x1E8
    4. Select "B.0 D.24 F.3"
    5. Write down the value of register 0x1EC

    Do the same with 0x46 instead of 0x06 at step three.

    This data will help to identify the chip and skip some probing routines.
    Last edited by justapost; 05-09-2008 at 10:54 AM.

  18. #268
    Xtreme Member aGeoM's Avatar
    Join Date
    Nov 2007
    Location
    Portugal
    Posts
    397
    Quote Originally Posted by justapost View Post
    The odd thing is that different nb vid's and did's worked fine here with the 9600BE and the 4-07 bios. The problem aGeom described did not occure with all phenoms. !
    NB-VID works if I set +xxx% over default.
    NB-FID if I set it to >06~what ever APM shows always 2000MHz, then after reboot and enter BIOS NB-FID=06, so the only way to set ie. FID=09(2400) is with Baredit tool after a reboot to windows.

    This was tested with 9850BE + Bios 307.

    Edit

    Bios 505:

    BIOS CPU-VID=+103.xx%
    BIOS CPU-NB-VID=103.xx%



    Well...VIDs are working as BIOS 307, CPU-FID can be changed but resets to default (2500), they should ghost this option to.

    Quote Originally Posted by justapost
    May I ask that someone with an 9850BE or an other phenom different from an 9600BE posts the values of the registers 0x06 and 0x46 here.

    1. Start BarEdit
    2. Select "B.0 D.24 F.3"
    3. Write 0x06 into register 0x1E8
    4. Select "B.0 D.24 F.3"
    5. Write down the value of register 0x1EC

    Do the same with 0x46 instead of 0x06 at step three.

    This data will help to identify the chip and skip some probing routines.
    I wrote 0x06 and 0x46 @ 0x1E8 nothing changed @ 0x1EC, as your instructions.
    Last edited by aGeoM; 05-09-2008 at 09:09 PM.



  19. #269
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    Quote Originally Posted by aGeoM View Post
    NB-VID works if I set +xxx% over default.
    NB-FID if I set it to >06~what ever APM shows always 2000MHz, then after reboot and enter BIOS NB-FID=06, so the only way to set ie. FID=09(2400) is with Baredit tool after a reboot to windows.

    This was tested with 9850BE + Bios 307.
    Hmm as usual I mixed up NB-VID and NB-FID in my previous post. I have in mind that i used different nb-fid's with CnQ enabled turing my p-state testings and it worked fine.
    Quote Originally Posted by aGeoM View Post
    I wrote 0x06 and 0x46 @ 0x1E8 nothing changed @ 0x1EC, as your instructions.
    Thank you for testing.
    Even if you select that "B0 D24 F3" pci register in the dropdown again?
    If i write someting into 0x1E8 it first shows 0x00 as it's new value. After I loaded the pci register again by selecting it in the dropdown the value i wrote in shows up and the data from that register shows up in 0x1EC.
    Is bit 3 ox 0x1E4 0b on your system?

  20. #270
    Xtreme Member aGeoM's Avatar
    Join Date
    Nov 2007
    Location
    Portugal
    Posts
    397
    Quote Originally Posted by justapost View Post
    Hmm as usual I mixed up NB-VID and NB-FID in my previous post. I have in mind that i used different nb-fid's with CnQ enabled turing my p-state testings and it worked fine.
    I whish it worked for me, this 9850BE, is stable if I set the MSR registers to 2700@default VID, but from BIOS I need to feed some more volts, NB it's ok @2400, so KTE help me to write the ".ini" to use with Sam2008 tool to set CPU speed at 2700, but for NB@2400 I have to manualy change it in Baredit and reboot, thats a PITA.

    Thank you for testing.
    Even if you select that "B0 D24 F3" pci register in the dropdown again?
    If i write someting into 0x1E8 it first shows 0x00 as it's new value. After I loaded the pci register again by selecting it in the dropdown the value i wrote in shows up and the data from that register shows up in 0x1EC.
    Is bit 3 ox 0x1E4 0b on your system?
    Well after 0x1E8 write first shows 0x00, but after I load it again it reads the value entered at the same adress 0x1E8. ie 0x1E8=0x06

    Here's a SS defaul system/0x06@0x1E8:
    Attached Images Attached Images
    Last edited by aGeoM; 05-11-2008 at 09:10 PM. Reason: Missing a quote, lol



  21. #271
    Registered User
    Join Date
    Apr 2008
    Posts
    9
    Quote Originally Posted by justapost View Post
    There is no need to run sensors-detect. The only module you need for sensor monitorin is "it87". Seems I have an robus bios chip here.
    The second time I ran into the C1 issue, removing one dimm helped. That dimm made problems for a few minutes and seems to work ok now.
    The old sensors-detect routine will trigger the chip at 0x2e to appliy an >3V vdimm. That should harm the dimm's and the chip's nb. I wonder why it also results in faulty bioses.
    Did you get both RMAed? If so did they say what was defect.

    That was tony's tip from a few weeks ago, whom I crossposted here.

    Whooee, 3.2v, I can see why things came to a halt. The dimms are ok but I could never get the cpu to work after that. I tried it in two other boards, including a new DFI M2RS. I didn't try a second cpu in the original M2R board, so it might have been ok.

    Yes, I RMAed both cpu and board. Newegg didn't get back with me about what the defect might be.

    So, sensors-configure is not necessary? I would really like to get lm-sensors working in Linux.

  22. #272
    Xtreme Addict
    Join Date
    Sep 2007
    Location
    Munich, DE
    Posts
    1,395
    Quote Originally Posted by vext View Post
    Whooee, 3.2v, I can see why things came to a halt. The dimms are ok but I could never get the cpu to work after that. I tried it in two other boards, including a new DFI M2RS. I didn't try a second cpu in the original M2R board, so it might have been ok.

    Yes, I RMAed both cpu and board. Newegg didn't get back with me about what the defect might be.

    So, sensors-configure is not necessary? I would really like to get lm-sensors working in Linux.
    Hi vext,

    I was wrong with the 3,2V the probing used random values so the applied value could have been anything up to 3.2V.
    Jean from lm-sensors made a few patches in the last days so the next version 3.0.2 will run just fine on the Sapphire/DFI boards.
    Everything lm-sensors dtects is the smbus controller (i2c-piix4) and the sensor-chip (it87). So if you add those modules manual you don't need to run sensors-detect.

  23. #273
    Xtremely Kool Final8ty's Avatar
    Join Date
    Jul 2006
    Location
    UK
    Posts
    1,876
    My Phenom 9850 Black Edition should be here tomorrow
    Some people have said in another forum that i will need to re-install Vista64
    going from Dual to Quad ?

  24. #274
    D.F.I Pimp Daddy Brother Esau's Avatar
    Join Date
    Jan 2007
    Location
    Still Lost At The Dead Show Parking Lot
    Posts
    5,188
    That would be a wise idea
    SuperMicro X8SAX
    Xeon 5620
    12GB - Crucial ECC DDR3 1333
    Intel 520 180GB Cherryville
    Areca 1231ML ~ 2~ 250GB Seagate ES.2 ~ Raid 0 ~ 4~ Hitachi 5K3000 2TB ~ Raid 6 ~

  25. #275
    Registered User
    Join Date
    Apr 2008
    Posts
    9
    Quote Originally Posted by justapost View Post
    Hi vext,

    I was wrong with the 3,2V the probing used random values so the applied value could have been anything up to 3.2V.
    Jean from lm-sensors made a few patches in the last days so the next version 3.0.2 will run just fine on the Sapphire/DFI boards.
    Everything lm-sensors dtects is the smbus controller (i2c-piix4) and the sensor-chip (it87). So if you add those modules manual you don't need to run sensors-detect.
    Thanks for the info. I just checked the lm-sensors site and they have a bulletin which explains the problem and their solution.

    I'll try version 3.0.2 as soon as it's posted.

    vext

Page 11 of 57 FirstFirst ... 89101112131421 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •