mw86 Posted September 18, 2011 Share Posted September 18, 2011 Can i do a voltmod on a Pentium D 3.6ghz? I have a decent set of ram in it and I could mod the case for global points... if I understand what that is... and I would need lots of help I could provide pics of whatever pins etc. Also SVL7 you are awesome I didn't have chance to use a mod like this on my 920xm I had... Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 Thanks. Going to be doing it tomorrow afternoon. Also downloaded team viewer. May let johnksss tinker with it a bit. If this works i should be able to pick up some hwbot boints Your free to tinker with it too as soon as i get the vmod done. It'll probably be after 1pm my time on the east coast. Good luck! I don't think I'll be online at this time, but thanks a lot for the invitation!! You and Johnksss will make good team to torture the CPU Can i do a voltmod on a Pentium D 3.6ghz? I have a decent set of ram in it and I could mod the case for global points... if I understand what that is... and I would need lots of help I could provide pics of whatever pins etc. Also SVL7 you are awesome I didn't have chance to use a mod like this on my 920xm I had... I'm not sure, I guess it should work as well, though I'd need to do some research first. Those Pentium work quite a bit differently than the current CPUs. What is the exact name of the CPU? Can you provide a CPU-Z screenshot or similar? Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 If the same slots in the original pic don't work, what would I need to try in order to bump voltage for the 540M? Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 It totally depends on the maximum voltage the CPU uses at 100% load. Here's a part of the VID table: I measured 1.152V at 100% load and 26x on all cores on my 920xm, a look at this table tells me the actual voltage is 1.1125V, which means the VIDs are set to 0 0 1 1 1 1 1 (VID6, VID5, ... , VID0) As you can see in the table, VID5 equals 0.4V, VID4 = 0.2V, VID3 = 0.1V etc. Each VID set to "1" subtracts its value from 1.5V, which in the end delivers the core voltage. So the 1.1125V with 0011111 is actually 1.5V - VID4 - VID3 - VID2 - VID1 - VID0 = 1.5V - 0.2V - 0.1V - 0.05V - 0.025V - 0.0125V = 1.1125V Deactivating one of the active VIDs results in a voltage increase since the voltage of the deactivated (= set to 0) VID doesn't get subtracted from the 1.5V anymore. This means the 1.1125V is almost perfect for a VID mod since two relatively significant VIDs can be set to 0 (VID4 and VID3) If you're 540m happens to be at 1.0V, you've got a problem, because the only VID available would be VID5 which is 0.4V, and would lead to a total voltage of 1.5V which is pretty crazy and the maximum the VR can provide. If the 540m is at 1.125V or above you can do the same thing as I've done, VID4 to VSS will give additionally 0.2V and VID3 to VSS will give +0.1V. If you're between 1.0V and 0.9125V you can use VID3 to get +0.1V, but VID4 wouldn't be of any use since it is already at "0" for these voltages... Worst case scenario is max stock voltage from 1.0125V up to (and including) 1.1V, because VID4 and VID3 are set to "0" for these voltages, which makes a useful pinmod impossible. I recommend you measure the voltage first without any voltage boost set from the BIOS, then you see where you are... if it happens to be at a value which isn't suited for the mod, you might be able to get to a better value with the BIOS increase. I'm not 100% sure how the BIOS voltage increase works, but I guess it simply shifts the VID a bit, thus raising the voltage by a fix value. This means when you're at let's say 1.0875V without any additional voltage from the BIOS, you should be able to get to 1.1125V or above with the BIOS tweak and thus get to a region where you can do a pin mod. Hope my explanation are more or less comprehensible, I'm in a hurry and have to go now, will probably be back in 3-4h. Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 (edited) Ok just grabbed a multimeter. Getting dice now. Going to try and get some good work done. Hopefully the meter is accurate. Thanks for the info. .5v would be crazy, but maybe on dice it won't be so bad. That will be the 520 since its the stock chip. Edited September 18, 2011 by DR650SE 1 Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 (edited) Well just measured the 540M without any bios voltage adjustments, looks like 1.06v Edited September 18, 2011 by DR650SE Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 Good luck! I don't think I'll be online at this time, but thanks a lot for the invitation!! You and Johnksss will make good team to torture the CPU I'm not sure, I guess it should work as well, though I'd need to do some research first. Those Pentium work quite a bit differently than the current CPUs. What is the exact name of the CPU? Can you provide a CPU-Z screenshot or similar? Thanks I'll post em when I can get a screen shot. I have a pic of the chip itself somewhere. Thanks I'd try this if I can... might go in the 4ghz range Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 Ok quick update,looks like 1.06v with no bios voltage adjustment.With bios +150mV its about 1.177v with the bios +150mV 2 Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 Ok quick update,looks like 1.06v with no bios voltage adjustment.With bios +150mV its about 1.177v with the bios +150mVAwesome can you take a screen shot of the main window of hwinfo not summary and hit the check by 940xm and take a screen shot of the overclock voltage increase listing? I want to compare what kind of voltage i've been adding to my 2920xm Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 Ok quick update,looks like 1.06v with no bios voltage adjustment.With bios +150mV its about 1.177v with the bios +150mVCool, so it's time for a VID4 mod I'd say, this should give you additionally +0.2V if my assumptions about how the BIOS voltage boost works are correct. So you can easily follow the picture in the first post. If you need even more you can connect VID3 to VSS as well, which should be good for another +0.1V.Keep us posted! Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 Awesome can you take a screen shot of the main window of hwinfo not summary and hit the check by 940xm and take a screen shot of the overclock voltage increase listing? I want to compare what kind of voltage i've been adding to my 2920xmWill do later. At the moment I'm doing it with my 540M and my 720M is next. 1 Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 (edited) Ok VID 4 -> VSS = 1.06v without the vbios 150mv Let me up the bios voltage to 150mv With VSS4 and bios 150mV I have 1.22v. Need moar volts! Edited September 18, 2011 by DR650SE Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 can you add another 75? Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 (edited) Ok VID 4 -> VSS = 1.06v without the vbios 150mv Let me up the bios voltage to 150mv With VSS4 and bios 150mV I have 1.22v. Need moar volts! Ok, seems the voltage boost per BIOS works differently... hmm. Don't worry, we can do this. 1.06V you say, this means VIDs a 0100011 Now if you set VID5 to VSS and VID4 to VCC you'll get a boost of 0.3V. Hmm, VID4 is at a stupid location... it's almost impossible to connect it to VCC with a pinmod... don't think it's possible without accidentally shorting a pin. Other option: You could connect VID3 to VCC, this will bring down the voltage by 0.1V, and then connect VID5 with VSS, this will result in a total of+0.4V, pretty hardcore... my max voltage was 1.4125V... dunno whether this is healthy. VID0 is at the edge as well and could be connected to VCC too, so you can bring down the 1.4625 to 1.45V, which is the Intel "absolute maximum rating" for the i7, probably as well for the i5, I didn't check, so this would be hardcore, but should be ok, but make sure to deactivate the BIOS voltage boost. So summarized, to get from 1.0625V to 1.45V: VID3 to VCC, VID0 to VCC and VID5 to VSS. Edit: And remove the VID4 mod of course, just to be safe. @DR650SE - sorry !!! EDIT 2: Wait, this would get you to 1.35V, seems I'm a bit tired, sorry! So this should be safe. Let me recalculate... 1.0625V-0.1V-0.0125V+0.4V= 1.35V Yep Edited September 18, 2011 by svl7 Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 im not sure but at max the unlocked bios allows i think it would put me in 1.5-1.6v range thats too much if it is right Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 1.45v sounds great, but now i'm a little confused between vcc and vss. Do you mean VID3 and VID0 to holes "AJ/35" and "AJ/34"? Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 im not sure but at max the unlocked bios allows i think it would put me in 1.5-1.6v range thats too much if it is rightIntel maximum rating for the 2920xm is 1.55V for the core voltage (with respect to VSS), and 1.65V for the processor PLL voltage (with respect to VSS).Seems the 2920xm can handle higher voltages than the 920xm (or Intel just adjusted the limit, lol) Anyway, the voltage regulator of the first gen i7 can only go up to 1.5V anyway, the SB CPUs have 8 VIDs, so one more than the first gen i7, this allows a finer voltage tuning and the VID table goes up to 1.6V. 1 Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 ill have to post a few pics for you to help me see what I am setting Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 1.45v sounds great, but now i'm a little confused between vcc and vss. Do you mean VID3 and VID0 to holes "AJ/35" and "AJ/34"?Sorry, added some edits to my post, my calculation was wrong, you'll get to 1.35V, a very nice boost, but it should still be in a safe level. You could still add 0.075V per BIOS if you need even more.The best VSS is pin is obviously AL/34 since it is just next to the VID pins, you already used it for the VID4 mod.The best VCC pin will be AG/35 or AF/35 since they're at the edge, the VSS_Sense and VCC_Sense pins can't be used, they're only the feedback lines for the voltage regulator. Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 http://i1237.photobucket.com/albums/ff474/marcwwaters86/unlocked bios pics/bios features pics/25flex098voltsor14962.png25flex vid stock bios limitswhat is that in pll... vss etc etc? Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 (edited) So I booted with VID5 -> VSS and it's at 1.4 with no bios voltage. Going to see what I can do on this, but it doesn't look to offer more stability And the 540M is stuck at 19x for the multip rather than the 23 it should be at. Edited September 18, 2011 by DR650SE Quote Link to comment Share on other sites More sharing options...
mw86 Posted September 18, 2011 Share Posted September 18, 2011 Strange maybe you need more he he Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 So I booted with VID5 -> VSS and it's at 1.4 with no bios voltage. Going to see what I can do on this, but it doesn't look to offer more stability Well, the clock generator isn't on the CPU itself, and you can only overclock the i5 via the base clock, so it might that the clock generator get to it's limit, or that the CPU simply can't handle a faster speed (the memory controller is on the CPU if I remember correctly). I had an idea yesterday... maybe I'll try it someday. Lower the RAM speed and increasing the timings by using Taiphoon burner, maybe this will allow to push it further... don't know. Is there a QPI bus option in the unlocked r2 BIOS? Lowering the QPI multi might help as well. When overclocking via base clock there are a lot of factors which can affect the stability. The voltage increase should definitely allow you to do some crazy stuff with the 920xm, but remember that you'll have to use a different pinmod for it, best will be to measure the voltage first. Quote Link to comment Share on other sites More sharing options...
svl7 Posted September 18, 2011 Author Share Posted September 18, 2011 And the 540M is stuck at 19x for the multip rather than the 23 it should be at.Just seen your edit, strange... I've never experienced such an issue when overvolting the 920xm. Quote Link to comment Share on other sites More sharing options...
DR650SE Posted September 18, 2011 Share Posted September 18, 2011 Think I killed my 540M lol. It' just gives a bunch of BSOD boots. Quote Link to comment 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.