• 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.

Can't get my E2160/IP35-E stable at 400MHz FSB

V

vj8usa

Senior member
With the rig in my sig, I just can't seem to boot up at 400MHz. I dropped my multi to 6x to take clock speed out of the equation, but even then it bluescreens in mid boot. I don't think my RAM's the issue, since I ran it at 400MHz using a divider for the past few months (back when I was running a 333MHz bus). I gave NB, SB, and VTT slight bumps in voltage, but that didn't seem to help. Should I try increasing NB (MCH voltage on my IP35-E) a little more? I'm not sure what values are considered safe (last I tried was 1.33v, up 2 notches from the stock 1.25).
 
Some component isn't up to the task but you can't blame the mobo without proper troubleshooting.

List your components and we can go from there.
 
CPU, Mobo, RAM:
E2160
Abit IP35-E (BIOS 15)
2x1GB G.Skill, 2x2GB Crucial RAM (both rated for DDR2-800 at 1.8v, though the G.Skill's rated for 5-5-5-15 timings while the Crucial's 6-6-6-18)

And other components, just in case:
4850
600W OCZ StealthXStream
X-Fi
IDE controller

Something just occurred to me - I think I might have been running my RAM 1:1 at 333MHz earlier. Maybe my Crucial sticks aren't holding up so well at higher speeds (though I tried upping vDIMM to 1.9v before and it didn't seem to help). If I leave timings at auto and try to go to 400MHz, will it use the SPD settings of the G.Skill or the Crucial RAM?
 
What is the max bootable FSB? Did you try all available multies? (not just x6, but try x7 and x8 as well)
 
It got into Windows at 380, but I haven't tried anything between 380 and 400 yet. 378x9 was actually reasonably stable. I did try 400 with all multis, and none worked (8x seemed to be a bit better though; it'd get to the green Windows loading bar before crashing). I'll try slowly going past 380 on a lower multi and see how high I can get, though.

The last thing I did was try to set my RAM timings, but something got messed up - it kept giving me errors about a floppy disk ("Drive A error" with a really quickly repeating beep), and then when I got it to boot the BIOS had reset itself. I guess I won't try that again for now.

edit: Hm, looks like 385 is bootable, 390 is close but locks up before the desktop finishes loading.
 
Does the p35 have fsb "holes" like the old P965 did due to strap change? The P965 had a hole from 370-400 fsb, and lots of times even if you couldn't hit 395 or 400 fsb, you could hit 401fsb and take it all the way up to 420 stable.
 
Originally posted by: vj8usa
CPU, Mobo, RAM:
E2160
Abit IP35-E (BIOS 15)
2x1GB G.Skill, 2x2GB Crucial RAM (both rated for DDR2-800 at 1.8v, though the G.Skill's rated for 5-5-5-15 timings while the Crucial's 6-6-6-18)

And other components, just in case:
4850
600W OCZ StealthXStream
X-Fi
IDE controller

Something just occurred to me - I think I might have been running my RAM 1:1 at 333MHz earlier. Maybe my Crucial sticks aren't holding up so well at higher speeds (though I tried upping vDIMM to 1.9v before and it didn't seem to help). If I leave timings at auto and try to go to 400MHz, will it use the SPD settings of the G.Skill or the Crucial RAM?
Click to expand...

Try with 1 stick of ram, then go from there. multiple sticks of ram stresses the NB.
 
No FSB hole with IP35-E, all the way up to 530MHz with E8400 E0 stepping.

Bottleneck is the bum chip. No need to mess around with VTT, VNB, and VSB with E2180. Most will top out around 3.2GHz. Use default 10x for best performance. E2180 doesn't like high FSB.
 
so you are mixing two different kinds of ram ?

I agree with the poster above:
try removing all but 1 stick of mem and see what that does for you
 
It may sound weird but my E21x0 chips did higher FSB with higher multiplier. I had an E2160 that would do 400+ FSB with x9 and x8, but x6 and x7 wouldn't go anywhere. Since you have x9 multi available, I'd aim for the highest CPU frequency possible irregardless of FSB.
 
Originally posted by: lopri
It may sound weird but my E21x0 chips did higher FSB with higher multiplier. I had an E2160 that would do 400+ FSB with x9 and x8, but x6 and x7 wouldn't go anywhere. Since you have x9 multi available, I'd aim for the highest CPU frequency possible irregardless of FSB.
Click to expand...

That's interesting. It looks like my realistic max is somewhere in the 370-380 region, as anything higher isn't really stabilizing.
Oh, and Gillbot, thanks for the suggestion - I might try removing 2-3 sticks of RAM out of curiosity to see what happens, but I'm feeling lazy at the moment (and even if running 1-2 sticks does help me push the OC higher, I'd rather have the full 6GB than a few hundred more MHz).

I've been doing some more experimenting, and came across something really strange. When I increase vcore past 1.47/1.49 or so in BIOS, it actually decreases in CPU-Z. For instance, 1.49 droops to around 1.46 in CPU-Z, while 1.55 was drooping all the way to 1.43. 1.51 drooped to something like 1.42 if I remember right (at the FSB I was testing, it crashed after about 2 seconds). This messed up droop seems to be reflected in stability as well (1.51 crashed much faster than 1.49). That actually might be the reason nothing I do helps stability at higher FSB, since I can't seem to supply the chip with over 1.46v.
 
Originally posted by: vj8usa
Originally posted by: lopri
It may sound weird but my E21x0 chips did higher FSB with higher multiplier. I had an E2160 that would do 400+ FSB with x9 and x8, but x6 and x7 wouldn't go anywhere. Since you have x9 multi available, I'd aim for the highest CPU frequency possible irregardless of FSB.
Click to expand...

That's interesting. It looks like my realistic max is somewhere in the 370-380 region, as anything higher isn't really stabilizing.
Oh, and Gillbot, thanks for the suggestion - I might try removing 2-3 sticks of RAM out of curiosity to see what happens, but I'm feeling lazy at the moment (and even if running 1-2 sticks does help me push the OC higher, I'd rather have the full 6GB than a few hundred more MHz).

I've been doing some more experimenting, and came across something really strange. When I increase vcore past 1.47/1.49 or so in BIOS, it actually decreases in CPU-Z. For instance, 1.49 droops to around 1.46 in CPU-Z, while 1.55 was drooping all the way to 1.43. 1.51 drooped to something like 1.42 if I remember right (at the FSB I was testing, it crashed after about 2 seconds). This messed up droop seems to be reflected in stability as well (1.51 crashed much faster than 1.49). That actually might be the reason nothing I do helps stability at higher FSB, since I can't seem to supply the chip with over 1.46v.
Click to expand...

Haha, yeah I had that too. Update bios to v18, that solves it. My e2180 drooped to 1.48 under load with bios voltage at 1.5075. You can put the voltage on your e2160 to 1.5v no problem per the Intel specifications on the website. My vdroop worsened and I had to increase the voltage to 1.52v in bios, vdrooped to 1.50.

Set your voltage to 1.5v in Cpu-z loaded and see if you can't get higher.
3.6Ghz is very high, but you've got a good chance at 3.4Ghz. My e2180 simply hit a wall past 3.4. Count yourself lucky if you get to that, most won't go past 3.2.

Oh and you can let it get up to 73-4C with that voltage and still be fine.
 
Originally posted by: soccerballtux
Haha, yeah I had that too. Update bios to v18, that solves it. My e2180 drooped to 1.48 under load with bios voltage at 1.5075. You can put the voltage on your e2160 to 1.5v no problem per the Intel specifications on the website. My vdroop worsened and I had to increase the voltage to 1.52v in bios, vdrooped to 1.50.

Set your voltage to 1.5v in Cpu-z loaded and see if you can't get higher.
3.6Ghz is very high, but you've got a good chance at 3.4Ghz. My e2180 simply hit a wall past 3.4. Count yourself lucky if you get to that, most won't go past 3.2.

Oh and you can let it get up to 73-4C with that voltage and still be fine.
Click to expand...

Thank you for the advice. I went to BIOS 18 and that certainly did help a little, but I'm still encountering the same problem. Now when I go past 1.49 in BIOS, the droop makes everything lower again. I can't seem to get 1.5+ in CPUZ no matter what (I think 1.51 droops to around 1.48; any higher than 1.51 droops to much less). 1.48 in CPUZ definitely made a difference over the 1.46 I was getting with the old BIOS, as 3.4 is stable. I feel like I could get 3.5 to be stable if I could get a bit more voltage, but I can't figure out how to fix this weird droop issue.

edit: I did some searching, and saw a post saying that CPUZ can't accurately show voltage beyond 1.47/1.48v. Is this true?
Someone else mentioned they didn't see the vcore changes they were making in their BIOS in Windows until they cleared the CMOS. I guess I'll try that next to see if it helps.
 
Originally posted by: lopri
It may sound weird but my E21x0 chips did higher FSB with higher multiplier. I had an E2160 that would do 400+ FSB with x9 and x8, but x6 and x7 wouldn't go anywhere. Since you have x9 multi available, I'd aim for the highest CPU frequency possible irregardless of FSB.
Click to expand...

I had a similar experience with my e2160.
 
Quick update: clearing the CMOS fixed the voltage issue; nothing droops more than 0.04v or so. It looks like my chip can do at least 3.5GHz, but it's taking a lot more vcore than I'd like. At 3.5GHz with 1.55v in BIOS, it boots fully but fails Orthos after a minute or two. Temps were hitting 70C as well, which is a bit too close for comfort for me.

I think I can get 3.4 stable with 1.47 or 1.49v in BIOS, which seems to be a good compromise. Orthos load doesn't get far past 60C at that voltage.

On a side note, this is with all voltages (other than vcore) at stock. I have high hopes for OCing whatever I upgrade this CPU to in the coming years.

edit: Damn, spoke too soon. I have to get up to at least 1.51 to keep 3.4 stable.
 
Originally posted by: ExcaliburMM
3.2 was the max for my E2140. I was sad. :C
Click to expand...

A 100% OC? Not bad. Did you hit an FSB wall at 400, or did it just need too much juice to keep it stable?
 
For the price of those E21x0 cores, you could try for a 100% OC.

I used cheap Gigabyte mATX GA73VM-S2 motherboards for an E2140 and an E2180. I couldn't change the RAM voltage, although I could tweak the VCORE, VTT etc. I chose to fix the vCORE at the stock (reported under "auto") voltage, and they wouldn't push past 333 Mhz. For me, the limitation was the motherboard.

In a way, it's interesting, because those cores were made for a default FSB of 800 Mhz . . .
 
For the price of those E21x0 cores, you could try for a 100% OC.

I used cheap Gigabyte mATX GA73VM-S2 motherboards for an E2140 and an E2180. I couldn't change the RAM voltage, although I could tweak the VCORE, VTT etc. I chose to fix the vCORE at the stock (reported under "auto") voltage, and they wouldn't push past 333 Mhz. For me, the limitation was the motherboard.

In a way, it's interesting, because those cores were made for a default FSB of 800 Mhz . . .
 
Originally posted by: BonzaiDuck
For the price of those E21x0 cores, you could try for a 100% OC.
Click to expand...
I'd probably have to be pushing close to/greater than 1.6v to get a 100% OC (3.6GHz) stable, assuming I could run a 400MHz FSB at all. It might be a relatively cheap chip, but I'd rather not risk dramatically shortening its lifespan just to squeeze out a bit more speed (besides, I don't think my cooler would be able to handle the heat at much higher than 1.5v; it already hits 70C under Orthos load with 1.5v in CPUZ).


Originally posted by: BonzaiDuck
...they wouldn't push past 333 Mhz. For me, the limitation was the motherboard.
Click to expand...
Incidentally, 333x9=3GHz is probably the "sweet spot" for my chip, as I can hit it with under 1.3v. I have to go all the way up to 1.5v to keep 378x9=3.4GHz stable. I figure I'll still run it at 3.4 though, since 1.5v is within Intel's specifications.
 
Originally posted by: lopri
It may sound weird but my E21x0 chips did higher FSB with higher multiplier. I had an E2160 that would do 400+ FSB with x9 and x8, but x6 and x7 wouldn't go anywhere. Since you have x9 multi available, I'd aim for the highest CPU frequency possible irregardless of FSB.
Click to expand...

Yeah. My E2140 was fine at 8x400, but 7x400 made my system freeze and flake out.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top