Hell-Fire - Tried running the BH-5 at 2-3-2-x ?
Printable View
Hell-Fire - Tried running the BH-5 at 2-3-2-x ?
I have very old BH5 RAM (single side Twinmos) 2x256Mb sticks and they do 255 2-2-2-5. I also have some newer BH5 sticks (corsair xms3500c2 2x256MB) and they do 260 2-2-2-5. So on my board, there isn't much difference between new or old BH5 RAM sticks.
Which board do you have?
If are you asking about the 2x512 setup, yeh....tried about every timing arrangement I can think of.Quote:
Originally posted by STEvil
Hell-Fire - Tried running the BH-5 at 2-3-2-x ?
My Newer BH5 (2x256), will run fine in both boards, as will the CH5 (2x512), but the older BH5 (KHX3500 2x512) will not run for pooh in the A7N8X.
hehe.... yeah faking 2x 256 out of them would suk, but as a test just to see.....hmmm
I have 1 256M BH5 Corsair coming from another friend, so I can confirm the 4 boards here....3 NF7-S 1 A7N8Xd all work correctly with singlesided BH5.
I would always expect a certain drop off in max FSB when going from 16 chips to 32 chips.... but not 50-60Mhz :rolleyes:
I agree, both the added options in the DFI bios and the CPC control must help put doublesided BH5 in a HAPPY PLACE.
I read somewhere, NVidia charges $10K for the reference manuals, with tech support and a reference board, PCBoard films, etc. We'll have to see if I can manage to cut through the BS. If not, maybe we should all take up a collection from all NF7-S and A7N8X users.... everyone ponies up $5-10, we get a lawyer to oversee a trust fund for this purpose and pay NVidia the $10K ...DOH!
Anyone know Kyle or Anand? Maybe making a public stink could loosen the grip on the book. Hell I'd happily pay $100 for the reference manual, thats what the shop manual for my car cost.
The Linux community would certainly be alot happier campers.... and NF2 would be easier to implement Linux on soon after release of the book.
Just rambling....
I am game for the public stink-a-roo. Where is my picket sign and corduroy pants??Quote:
Originally posted by uwackme
hehe.... yeah faking 2x 256 out of them would suk, but as a test just to see.....hmmm
I have 1 256M BH5 Corsair coming from another friend, so I can confirm the 4 boards here....3 NF7-S 1 A7N8Xd all work correctly with singlesided BH5.
I would always expect a certain drop off in max FSB when going from 16 chips to 32 chips.... but not 50-60Mhz :rolleyes:
I agree, both the added options in the DFI bios and the CPC control must help put doublesided BH5 in a HAPPY PLACE.
I read somewhere, NVidia charges $10K for the reference manuals, with tech support and a reference board, PCBoard films, etc. We'll have to see if I can manage to cut through the BS. If not, maybe we should all take up a collection from all NF7-S and A7N8X users.... everyone ponies up $5-10, we get a lawyer to oversee a trust fund for this purpose and pay NVidia the $10K ...DOH!
Anyone know Kyle or Anand? Maybe making a public stink could loosen the grip on the book. Hell I'd happily pay $100 for the reference manual, thats what the shop manual for my car cost.
The Linux community would certainly be alot happier campers.... and NF2 would be easier to implement Linux on soon after release of the book.
Just rambling....
I know that they manufacture a ridiculous amount of boards (Abit, ASUS), and that the chipset makers also make tons of their products.....but $10K to purchase something so that we can make their products better for everyone...well, they can just bite my arse sir.
I say they give us the materials, we fix their mess, then they pay us a consulting fee.
I know that they (mainstream business) do not care as long as the product does the rated speeds...but they will certainly push it toward the enthusiasts who is not by any means going to run it at specs to begin with...arghh, now I am ranting...way to go uwackme. :p:
50-60 mhz is ludacris....thats not a drop off, thats the never ending pit of frustation. Nevermind overclocking, they arent even running at spec speeds...period. We all know that overclocking is a gamble, and some boards perform better than others, some ram performs better than others, video cards, etc etc etc. Sheesh, even 2 of the same motherboards (ram, video cards, etc etc) will react totally different when pushed past spec.....yet again, we are not talking about the case of overclocking, we are talking stock damn speed here. But when something refuses to run at rated speeds....Knock knock. Who's there?? Pssted Off. Pssted off who?? The end user, thats who!!
NF7-S off courseQuote:
Originally posted by Byron
Which board do you have?
what about places like this one, uwackme? - http://www.doom.com
i typed in NFORCE2 and got some sample results with MCPT, SPP... is that what we're looking for as far as chipset datasheets?
just wondering,
Yes, we need data sheets with register lists, bit-codings, etc for te chips.
SPP and MCP-T are our babies.
I signed up, will see what I can find.
Man we need this BIOS fix.
Im purring along at 220Mhz 11,3,3,2 with 2x 512M BH5 in my new DFI LanpartyUB....CPC off. Working with Farcry. Memtest is happy 225Mhz 6,2,2,2 for hours on end. 230Mhz gave errors after a couple hours, I think I have alot of COOLING work to do on all the mosfets, etc. Plus the Superflower550 I got has a low 12V rail, I need to goose it up.
I'd rather be doing this in my NF7-S. Though I like the actual DIODE temp reading. It was funny, I was able to port the RaptorRaid setup to the DFI.... used the Sil3112 PCI card as a bridge, booted in DFI, loaded drivers for the 3114, then move the connectors over and Presto! Without a peep of trouble that Raidarray has gone from one NF7 to another NF7 to an A7N8X to a DFI LPuB, WinXPpro happy as a lark the whole time..... so funny.
Wonder the the heck all the other settings are in this bios..."Superbypass" etc, etc.
Good to know all 3 sets of BH5 kick ass.
What leg on the vtt mosfet gives the correct vtt voltage? I cannot pass 260FSB no matter what I try so I want to know if the vtt tracks the vdimm or not. I did the vtt=vref mod.
Thanks, I found also this pic at Malves website:
http://www.malvescorner.com/images/nf7vddvbt.jpg
Yeah, almost the same way as on my 8rda+ with the upstanding pot's .. :p:
But I'd suggest the Vdimm=+3.3v mod instead of the mbfaq one and then regulate the vtt by hand (or do the vtt=vref mod)? Not?
Cant the Vbt also be checked at the ICs under the DIMM Slots?
I think it's pin 7 + 8 of the Vtt little mosfet.
But best place is simply at the bazillion resistor packs alongside the DIMMs.... notice they all are soldered to one big chunk of etch on the top of the board..... that is Vtt.
Thanks for the heads up on that.
I have been checking via the fets instead.
I will see how close the 2 different points are.
Right, it is the resistor packs that matter for Vtt.... that is the end of the line, where the voltage is used, so it needs to be accurate there. Best to set it monitoring the actual terminating resistors.
on my board (with 3.3V=vdimm and vref=vtt mod) the vtt tracks nicely the vdimm.
Well sort of ... it tracks the vdimm value I get from the 8rdavcore program, but not the real multimeter measured volts.
Eg, I measured a vtt of 1.77V when the vdimm was 3.55V in 8rdavcore (which corresponds to 3.65V measured with the multimeter).
So it tracks, but it undervolts a bit. Still I can get 265FSB with the vtt-vref mod before my mem craps out. With better mem, I should be able to do 270.
Jus got my board back from Malves, who did a great job in modding it (Vtt cut-trace,Vdimm mbfaqs, Vdd mbfaqs)!
Cheers Malves :D
Been doing some testing with 3x512MB BH5 sticks over the past two days...
All testing was done in Memtest only so far, until I decide which 2 sticks I want to keep... then I'l do windows based stability testing.
My settings for the Memtest runs were:
(All voltages are measured, but the BIOS reports the same values)
3v Vdimm, 1.5v Vtt, 1.6v Vtt, 1.65v Vcore, L12 mod, CPU Int enabled, APIC enabled, 6/6, 8x multi.
1st stick = Cosair PC3500 rev1.1
2nd stick = Mushkin PC3500 lvl2
3rd stick = Mushkin PC3500 lvl2
Dimm1 (nearest CPU) Single Channel, Single Stick Results:
1st stick : 231 FSB
2nd stick : 225 FSB
3rd stick : 223 FSB
Dimm3 (nearest Edge of board) Single Channel, Single Stick Results:
1st stick : 240 FSB
2nd stick : 227 FSB
Dimm1 & Dimm3, Dual Channel
1st stick Dimm1, 2nd stick Dimm3 225 FSB
1st stick Dimm3, 2nd stick Dimm1 230 FSB
note: Mushkin sticks tend to get a sheeeet load hotter so may gain a few MHz from having the spreaders removed.
note: Dimm1 and Dimm2 performed identical in terms of Memtest stability.
note: Module in Dimm3 tends to heatup a lot more than Dimm1
[EDIT] ongoing updates
________
VAPORIZER
Pff, Have 1 stick twinmos bh-5 , 512m, won't boot above 225fsb, :( I don't have any mods, but i'll try the L12 mod soon!
try putting it in Dimm3 (furtherest from CPU)Quote:
Originally posted by Rub87
Pff, Have 1 stick twinmos bh-5 , 512m, won't boot above 225fsb, :( I don't have any mods, but i'll try the L12 mod soon!
________
volcano vaporizers
At least your sticks are well burnt in, those are good results in a NF7 with 512M sticks Rabbi.
My Mushkin is new... and may account for it's lower performance compared with the Cosair.Quote:
Originally posted by uwackme
At least your sticks are well burnt in, those are good results in a NF7 with 512M sticks Rabbi.
From experience, does removing the Heatspreaders ever help gain a few FSB?
I ask this because without a fan on my modules I cant do 2-2-2-11 230 Dual Channel for prolonged MemTest runs... once sticks get nice and toasty they start to fail.
But, with a fan on low speed they run at that speed all day in Memtest.... Prime is another story completely!
________
Honda DYLAN125
Well, its a commitment if you take the spreaders off.... warranty is gone. Fan sounds cheaper in the "incase" case. I have good airflow in the case past the rams, so I am set, but I did take the spredders off my CH5 and BH5, Im not sending them back, if one fails, Ill use the remaining good chips for repairs to other sticks....I have access to SMD rework equipment.
Remember, I found those stickers under the Kingston spredder.... it DEFINITELY affected one chip, made it fail early....I figure by design, but its all gone now :D