Edward Ng Posted October 22, 2004 Posted October 22, 2004 I just wanted to mention some things I noticed, and to ask a question of the other brilliant minds in here... First off, I noticed that in order to solidify voltage changes, you need to actually save, reboot, reenter BIOS, verify setting (but notice the current actual voltage!), save, reboot and now your voltage is set. I'll go into more detail... The voltage setting on the LP nF3 250Gb is actually a two-stage setting, the base voltage (up to 1.55) and then a percentage modifier (up to 136%). Let's say my current voltage is 1.45 and 110% modifier. That makes my current voltage 1.595. Now, let's say I set it to 1.55 and 136%, save it, reboot and reenter. I will see that now, my saved setting are still 1.55 and 136%; however, the actual voltage right now is only 1.55, indicating it has set the primary vreg, but hasn't instated the modifier yet! Now, I save it again, reboot once more and then go back in again. Now my BIOS will indicate a full voltage of 2.108. This is how it's been working for me since updating to Mr. Wu's most recent BIOS; I've not yet tried Jess' extra VDIMM and MemTest BIOS, and I didn't check this operation with the factory BIOS. Come to think of it, it's likely that you actually only need to reboot twice to finish setting the new voltage without actually going in, but I like to go in and verify with my own eyes that it's really working. Secondly, there seems to be an interesting situation with the Secondary IDE Slave option on Mr. Wu's most recent BIOS. Whenever I make a change in my BIOS, or at least whenever I make a change within the Genie page, my Secondary Slave detection will change from the setting I prefer, "None," back to, "Auto." Thus, when I reboot next time after saving and restarting, it detects my HDD at Primary Master, my DVD-ROM at Secondary Master, and then makes an attempt to find a drive on my Secondary Slave, which, results as it should with, "None." I basically have to keep changing my Secondary Slave back to, "None," every single time I go in there and make a change. It's not critical and does not affect boot time, performance or stability, but combined with the double boot to set new voltages, it is somewhat annoying to keep having to correct this setting as well. Finally, the item that I need help with and that's bugging me. Currently, my system runs rock solid (Prime95 In-Place Large FFTs torture, MemTest86, all 3DMarks, PCMark04, Far Cry, Doom 3, UT2K4--all pass) at 8.5X286 10-3-3 CAS2.5 1T with a single stick of PQI PC3200 Turbo 512MB. If I stick in a second stick, I can only maintain 10-3-3 2.5 1T to HTT250, but that's off topic as I'm using just one stick for now to maintain my super-FSB. Anyhow, if I set my HTT to 287 or 288 (haven't tried higher), my rig just plain gets stuck after detecting drives, at a screen reading, "Detecting Array," and then either it reboots, or it freezes and I have no choice but to shut off with the ATX switch out back behind my PSU, because neither the power button nor the reset button works (unless I catch it early when the, "Detecting Array," first shows up). I know it has nothing to do with the HTT because I've booted into MemTest and successfully tested 10-3-3 2.5 1T at up to HTT295 on a single stick at lower multipliers. On the same token, I know that with a Multiplier of 10.0X, I can get into Windows at 250HTT, which is 2500MHz. Calculate 8.5X * HTT286 and that's only ~2433MHz; 8.5X * HTT288 results in only ~2448MHz. So, I don't think it's because of the CPU, either. I read in another post by someone else to try setting Max Async to 8-10, Write CAS (twcl) to 1 or Auto and Drive Strength to 1 or 2 for my CPC On, but none of these solutions worked. I have tried setting the LDT multiplier to 2.5X as well as 3.0X and it doesn't help; I understand Mr. Wu mentioned the 2.0X and 1.5X settings as being nonfunctional, and 2.0X didn't help me, anyway. So, I'm basically out of ideas, unless by some remarkable phenomenon, the combination of HTT286 * 8.5X is the limit, even though the HTT can be taken higher and the chip is not maxed. the primary thing stopping me is this, "Detecting Array," thing, and I just don't know what to try anymore. Mr. Wu mentioned that this is likely related to a problem programming the HTT, but why doesn't this happen when I use at a higher clock rate if I use a lower HTT, and why doesn't this happen at a higher HTT if I use a lower multiplier? PLEEEEEASE won't someone help me?!!? -Ed Quote Share this post Link to post Share on other sites More sharing options...
reject Posted October 22, 2004 Posted October 22, 2004 detecting array means u need more voltage try upping the vcore as the memory controller needs more vs even at a lower cpu speed at high HTs. Quote Share this post Link to post Share on other sites More sharing options...
Edward Ng Posted October 22, 2004 Posted October 22, 2004 How is it that I see other people getting such great results at only ~1.7volts and I need to dump over 2volts in? Or am I pushing harder than most people at this point? I am using a water cooled setup. Quote Share this post Link to post Share on other sites More sharing options...
RGone Posted October 22, 2004 Posted October 22, 2004 :confused: man you have no signature as per forums requests to do so. We don't have a clue how hard you are pushing it. Under normal circumstances if you run up against the hang at detecting array many of us will let it sit a minute and it will do one of two things usually. 1 Stay hung. 2. Reboot and do so 3 times and come up with a message that computer is in safe mode and please to enter bios and adjust speeds etc. At the point of it booting 3 times and the warning is not heeded and a reboot is initiated it will cycle 3 more times and re-present the warning to the user. When it does this to me personally> it has been because I was pushing harder than the pieces would go and I backed off a little on whatever setting "I had just made in bios" and bingo booting and going again. Something during booting is not being liked and the bios does not fully initiate and bingo it hangs at the detecting raid array as the computer is hung and cannot 'erase' that screen as it normally would and continue on. Further and who in hale knows why, is that turn of events where it will run 14x200 but won't run 10x240 and all of us who have played with overclocking very long have seen this. Why? Again I don't have a scientifice answer as the test instruments are beyond my means. Sincerely, RGone... :cool: Quote Share this post Link to post Share on other sites More sharing options...
Edward Ng Posted October 22, 2004 Posted October 22, 2004 Okay, I have modified my signature to indicate the best settings I've achieved so far... I tried the voltage suggestion, and it didn't make any difference, other than increasing my temps by a huge amount. It didn't get any more or less stable at the higher voltage and temps, and it didn't allow me to pass the HTT programming stage. The setting I tried was the full 1.55 and 136%. Actually, that's only .1Vots more than I have to run to get my current settings, anyway. It does seem strange to me, however, that while most people run 1.7-1.8Volts and get way higher clocks, I have to run 2.0 to get 2433MHz. :confused: Btw, thanks for your help so far, guys. -Ed Quote Share this post Link to post Share on other sites More sharing options...
Samurai Jack Posted October 22, 2004 Posted October 22, 2004 MR Edward NG First I would like to tell you about the 1/2 multiplier issue. Running the settings you are running does not have your mem at 1:1 because when you use 8.5x the mem runs at the 9x settings. This will cause you problems trying to set up because you think you mem is running this & it will be running that. You may be able to run higher than you are because if you went to a 9x from where you are then all of a sudden your mem is 1:1 & you have not tuned it for this speed because you have tuned with 8.5x and mem running slower. What you mem is running at these settings is 270. so if you set 9x270 you will have the exact same thing & know more about where you are. Now you can start tuning again. Hang up(detecting drive array) probably is a combo thing. One I would use 3x LDT. The rest is going to be Ram settings that is probably not leting it detect the array. SJ Quote Share this post Link to post Share on other sites More sharing options...
Edward Ng Posted October 22, 2004 Posted October 22, 2004 DAMN I knew I forgot to mention my Ratio setting... I have the memory forced to 1:1 in the BIOS; in Windows, CPU-Z confirms that my memory is indeed doing 286 10-3-3 2.5 1T, as does the POST screen (DDR400). Unless you're telling me CPU-Z is wrong (which I'd believe you, actually; nothing's impossible)... As I performed all memory testing at a multiplier of 4.0X, I know, firmly that, my memory can reach HTT295 10-3-3 2.5 1T (at 4.0X). I'll try 8.0 later with a different Ratio; that, I haven't yet tried, since I don't know how well Async works on nForce3... I am now using LDT multiplier 3.0X as well, and it isn't improving the situation, either. But thanks for those suggestions... -Ed Quote Share this post Link to post Share on other sites More sharing options...
TobyWong Posted October 22, 2004 Posted October 22, 2004 How is it that I see other people getting such great results at only ~1.7volts and I need to dump over 2volts in? Some A64s, Newcastles in particular get better OC's with a conservative voltage increase. Adding more voltage will not always give you better performance... in some cases it can actually slow you down. Quote Share this post Link to post Share on other sites More sharing options...
Edward Ng Posted October 22, 2004 Posted October 22, 2004 I have to run 1.55 & 126% to boot at HTT286 10-3-3 2.5 1T; I tried dropping it to 1.55 & 123%, as well as 1.55 & 113%, and in both cases, it gave me, "Detecting Array." Going up to 1.55 & 133% as well as 1.55 & 136%, it still boots fine, but it doesn't allow me higher than my current HTT286 10-3-3 2.5 1T. I think it's becoming clearer why I'm stuck. :confused: -Ed Quote Share this post Link to post Share on other sites More sharing options...
Rick-23 Posted October 22, 2004 Posted October 22, 2004 Is it save to use that much cpu volts, ive got it to boot at 2600 tho that was with 1.92 - 95 v and it didnt seem stable.. AMD Athlon 64 3000+ watercooled DFI NF3 1 GB ocz 3700 EB enermax 465 i currently run at 9.5x270 = 2560 3-2-2-0 1T i cant seem to get any higher than that , it will do 9x260 on the same timings tho 9.5x270 gives me more bandwidth Quote Share this post Link to post Share on other sites More sharing options...
ElixXx Posted October 22, 2004 Posted October 22, 2004 the volt changing thing , are the same for me, eg must reboot 2 times for the volt to apply. Quote Share this post Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.