• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Ram Divider Holding me back!

I

imported_adama

Junior Member
My new arrival (just born last week!) is an Athlon 3200 Venice:

LBBLE 0522 CPEW

Got it stable at 2400MHz at the mo using a 10x multi and 4x HT with my Ram on a 180 9:10 divider.

I would love to get my 3200 chip somewhere near 2.7GHz (like so many others have done) - but the Ram or something else just DOES NOT like going to a lower RAM divider (ie 166 5:6 or less!) once you start to clock the HTT up....it just won't post!

Even if I left everything as it is now and dropped the divider to 5:6, the damned thing just can't handle it.

I've tried 4 different BIOS flashes for the DFI / Multiple different Ram timings and voltages. But the common underlying problem is always using the 5:6 166 Ram divider or anything lower.

If I left the board at stock speeds for my 3200 chip (10x220MHz etc) - then I can get the 166 divider to work 😕 But as soon as I push the HTT up past about 235 x 10, the 166 divider won't post.

So my poor Geil value 2GB set is running flat stick at about 216MHz (3200 value chips don't like going over about 220MHz). If I could lower it too the next divider down - I have no doubt my venice could easily get to 2.5 - 2.6GHz :thumbsup:
 
Many mobos will automatically tighten the timings when lowering the RAM divider. Have you checked that the timings are staying loose?

That's my guess anyway.
 
Something you can try,.... I had cpu vid control at 1.45v, cpu special at 113% and cpu vid start up at 1.45v and could not get over 2.4ghz, then I went to 1.3v cpu vid, 126% cpu special, and 1.5v start up,.... hit 2.6ghz. May be worth a shot,...dunno why it made a difference on mine but ,...
Also up the ldt to 1.3v and the chipset to 1.7v.
Have you tried upping the voltage to the ram?
Which bios are you currently using,... what about memory settings other than the divider?
 
I highly suggest using the 7/4/05 BIOS that can be found on df-street.com

When I first put together the system in my sig, any RAM divider lower than 9/10 caused a BSOD or a blank screen right after the windows splash screen. I was ready to toss this comp out the window, flashed to 7/4/05 the day it was released and wahla, worked like a charm.
 
Great suggestions there - thanks all 🙂

I had the default bios (?) - then tried 510-2 DFI BIOS - then tried the 623-1 Bios (of the flash BOOT CD from DFI Street). They didn't like anything above about 230 HTT and a 166 divider.

I will chk out the 7/04/05 bios and give it a try - thanks Matt2.

11427 - nice little voltage pointers there. I'm running 2.9v on the RAM. Have tried all mem voltage steppings up to 3.2v (to see if it would make any diff) - but still same result. Current timings are: 2.5-3-3-7 (1t). I've set all these manually and tried loosening them to varying degrees when I'm trying to get the 166 divider to work (as well as trying the AUTO settings).

I will try the new bios and let you know - many thanks again.

<can't find the damn link to that 7/04/05 bios - can you pls send me>


 
Originally posted by: 11427
Something you can try,.... I had cpu vid control at 1.45v, cpu special at 113% and cpu vid start up at 1.45v and could not get over 2.4ghz, then I went to 1.3v cpu vid, 126% cpu special, and 1.5v start up,.... hit 2.6ghz.
Click to expand...

While I was BIOS hunting - I came across this official warning (just in case you hadn't seen it):

Due to the a mistake made in 7/04 beta bios , do not use

'Above VID * 123%'
'Above VID * 126%'
'Above VID * 133%'
'Above VID * 136%'

option under "CPU VID Special Control"

The bios may have chance to set wrong vcore voltage under certain circumstances ...

If the cpu can not handle the wrong core voltage , it may be damaged ..

Also when you say use 7/04/05 bios - is this one of the 704-1 (and -2 /-3) bios's? Because I'm struggling to find a bios called 7/04/05 (thanks for the link coomar).

 
yeah, 704 refers to 7/04, the board only has bioses from this year

i use -2 which is optimized for utt ram

try -1 or -3

 
Originally posted by: adama
While I was BIOS hunting - I came across this official warning (just in case you hadn't seen it):
Click to expand...

I hadn't seen that, thanks for the info,.... I'm using the 6/23 bios
 
I had a similar problem on my DFI Ultra-D with the memory in sig.

It will not run any dividers except 1:1 and 3:4, even at stock speeds any other setting will fail, regardless of timings or voltage.

I popped in some absolute trash kingston value ram from my 400SC server that i paid $45 for 512MB for back in the day. Works on all dividers.
 
Just some results after spending about 5 hours yesterday tweaking!!

Tried flashing the DFI to the 7/04/05 (or 704-1) BIOS, still no joy getting the 166 Divider steady on anything higher than my usual <230 htt.

I combined some of the CPU VID Special '%' voltage tweaks that 11427 advised being VERY careful not to use anything above 110% on the 704 bios...nasty :Q

Didn't have much joy there - so I tried going back to the 310-1 BIOS (10/3/05 I'm guessing). Hey presto! I got my 166 (and even the 150) dividers to work 😀

I thought - lets see what this 3200 venice really CAN do now I've got my Ram stable on a safe divider.

I went from 240 to 250 (x10 x4 HT) and the cursed thing stopped posting. I was really on a mission by that stage to get this sucker stable at 250, so spent the next 4 hours playing with all sorts of voltage tweaks and clocks. Finding that it would lock at 250 - fine at 257 / lock at 255 / fine with 253 - very inconsistent. Just when I thought everything was good (except Prime's torture test!) - I would re-boot and it wouldn't post. grrrr...... :frown:

A few times I got it into Windows at ~2.7GHz (using 3 x HT). Was fine on memtest / failed Prime / locked up in 3dmark03.

Then ~2.6GHz (fine on memtest and 3dmark03 / but prime fails within 5 seconds).

So at midnight, I thought I had 2.5GHz stable (x10 x4ht - 2.9V Ram / ~1.6v CPU / also upped the HT volts and one other <?> up by the usual .1v) - even prime was running ok after a few mins (couldn't risk leaving it running o'night). I slept...resting assured my toil was worth loosing half a weekend over.....

Woke up this morning she wouldn't post!! So something (quite possibly my 3200 Venice DOES NOT like anything >2400MHz).

Alas - I have not given up yet!
 
my venice doesn't like high voltage, have you tried 2.5-2.6 at 1.475-1.55 voltage range, some of the newer cores are very sensitive to voltage
 
Yes - I tried most combinations of voltages to get the 166 and 150 dividers to work with the 310 bios (incl 1.45 up to 1.55v CPU core).

I've tried another 4 different bios's tonight to see if one would magically let me do over 230MHz on the 166 or 150 dividers - but no post I'm afriad.

Just a quickie - a guy in another post on this forum advised what the 3 different BIOS types were:

it's -1 TCCD -2 UTT -3 General purpose

Anyone using 3200 Geil Value know which category I'm in?

I tried all three of each BIOS I tested - but it didn't solve my problem 🙁

At the moment I'm using a 623-3 BIOS and back to my old 180 (9:10) divider, 250 x 10 CPU, but stressing my poor geil value to 227MHz!! I can get memtest and 3dMark03 stable - but prime gives it the flick after a few seconds (ps nice score on 3dmark03 - highest yet at: 11379 😎
 
if your value is new and blue than it should be winbond utt (it should be the bh-5 dyes)

if you look at the ram module on one side of the pcb, there should be a few letters/numbers, you want the combination bp in their (brainpower)

if it loves voltage its probably the utt and you should use -2

otherwise use -3

people argue between whether tccd is -1 or -3 and general is -1 or -3, all that matters is that utt is always -2
 
You're exactly right coomar - MY Geil's only 4 weeks old and blue. I was thinking the modules would have it written on the side, but wasn't game to take the heat casing off the outside to check!

The 3200 Geil pair runs stable in prime 95 up to about 220MHz on tight 2-3-3-6 1T (2.9v) - happy with that. When pushed to 227 (by running a 2500MHz clock on the 9/10 divider) - than I can't expect it to be stable in the prime torture test....this IS VALUE ram after all :roll:

I believe it's the processor that's preventing me from clocking on the lower dividers. I'll just be content with 'hovering' between 2400-2500 depending on how greedy I'm feeling.

 
You must log in or register to reply here.

TRENDING THREADS

Back
Top