Is anyone getting 1700+'s that DON'T O/C well? UPDATE: SOMETHING ALL OC'ERs SHOULD READ!

[Ilmater]

Alright, I read this forum and HardOCP's overclocking forum fairly regularly. I see everyone getting decent to incredible results with their 1700+'s. Well I got a DUT3C 0302 back in late march, and it won't do over 1.8GHz stable. Then I had a friend who wanted to buy a 1700+ and run it stock, so I sold him that one and got an 0310 DLT3C XMPW. Now I can't do over 1.5GHz stable. I have a Swiftech MCX462+ with a Panaflo H1A 92mm fan on top of it cooling the CPU. I have an Epox 8RDA+ with good NB cooling and passive SB cooling running at 200MHz FSB. However, a 1.5GHz T-Bred B isn't powerful enough for the games I want to play. My CPU runs at 38C under load and 35C idle and my system temps are 30C load and 28C idle. I don't think cooling is a problem.

The strangest part is, with this new proc, it boots as high as 2.4GHZ, but it's totally unstable. I can run 1.5GHz (7.5 x 200) all day long at 1.55V, but I can't run one ounce higher. Not only that, but running 1.6GHz, while unstable, is more stable at 1.55V than it is at 1.6, 1.7, or even 1.8V. It seems to get more unstable at higher voltages. Am I doing something wrong? (BTW, I know the RAM is good because I ran memtest all day at 200MHz and it came out just fine - it's Corsair's XMS after all)

 

[soja]

Wow, 7.5x200 not enough for a game, amazing game Couldn't help you much, waiting on a dlt3c 1700 b myself. Although I have read in various forums and sites about similar chips that couldn't run stable with vcore over default. Came about while I was searching google about jiuhb chips and whatnot. Happy searching

 

[Ilmater]

Originally posted by: soja
Wow, 7.5x200 not enough for a game, amazing game

It's only 23MHz over a stock 1700+. BF 1942 is needy. Even my gig of RAM doesn't make it happy enough.

 

[anomaly]

Drop the FSB down to like 180 and raise the multiplier and see if you can get a higher clock speed out of it. If not, dunno try the cpu on another mobo to see if it runs higher.

 

[soja]

ahh BF1942. That game's killer. Could your ti200 be a bottleneck?

 

[Ilmater]

I'm sure my Ti200 is somewhat of a problem, but the game is just choppy(like every 5 seconds it skips a few frames), and since I'm at 1152x864 at the lowest quality settings, I thought it was the proc instead. I could be wrong. Regardless, I think that this proc should get higher than this.

UPDATE:
Last night, I tried moving things up again. I just decided that maybe the proc needed more time to adjust. Well, I moved up and tested with 15 minutes of Prime95 at each multiplier. 8x - It's good! I'm pretty pumped at this point because it wasn't stable here before. I figure I'm probably at the end of my rope, but I decided to try again. 8.5x - EUREKA! I'm good! Hey, why stop there, right? 9x - Solid as a rock! I even ran through 1 run of 3DMark again with Prime95 in the background, and it was fine. So, I tried one more.

9.5x - this is where it gets weird. After I set this, the computer restarts and my motherboard says, "Instability caused due to faulty (something, I don't remember) or processor overclocking. Please reduce CPU speed (or something, blah, blah). Press F1 to continue or (something else)." That's strange. So I hit F1 and it went past the post screen and to my SCSI card initialization and I hit Ctrl+Alt+Del to restart it. I upped the voltage by .25V (I'm at 1.6V at this point) and restarted. No post. Mobo started beeping. I had to restart holding down Insert to get it to POST so that I could lower the voltage and lower the multi back down one.

What's the deal?! I ran 3DMark w/ Prime95 all night and I'm rock solid at 1.8GHz now at 1.575Vcore. Why did it suddenly decide to go up and will I be able to go up more? Does anyone know anything about the Epox beeping?

 

[Doh!]

What psu are you using? I'm asking this since upping the voltage resulted in instability.

 

[Ilmater]

Originally posted by: Doh!
What psu are you using? I'm asking this since upping the voltage resulted in instability.

Enermax 435W

How can I check to see if it's working correctly? Can I look at the voltages in the BIOS? Can I use a voltmeter? I don't even know what kind of voltages to expect from a +5V line or a +12V line! What's an acceptable variation?

 

[xSauronx]

Originally posted by: Ilmater
I'm sure my Ti200 is somewhat of a problem, but the game is just choppy(like every 5 seconds it skips a few frames), and since I'm at 1152x864 at the lowest quality settings, I thought it was the proc instead. I could be wrong. Regardless, I think that this proc should get higher than this.

UPDATE:
Last night, I tried moving things up again. I just decided that maybe the proc needed more time to adjust. Well, I moved up and tested with 15 minutes of Prime95 at each multiplier. 8x - It's good! I'm pretty pumped at this point because it wasn't stable here before. I figure I'm probably at the end of my rope, but I decided to try again. 8.5x - EUREKA! I'm good! Hey, why stop there, right? 9x - Solid as a rock! I even ran through 1 run of 3DMark again with Prime95 in the background, and it was fine. So, I tried one more.

9.5x - this is where it gets weird. After I set this, the computer restarts and my motherboard says, "Instability caused due to faulty (something, I don't remember) or processor overclocking. Please reduce CPU speed (or something, blah, blah). Press F1 to continue or (something else)." That's strange. So I hit F1 and it went past the post screen and to my SCSI card initialization and I hit Ctrl+Alt+Del to restart it. I upped the voltage by .25V (I'm at 1.6V at this point) and restarted. No post. Mobo started beeping. I had to restart holding down Insert to get it to POST so that I could lower the voltage and lower the multi back down one.

What's the deal?! I ran 3DMark w/ Prime95 all night and I'm rock solid at 1.8GHz now at 1.575Vcore. Why did it suddenly decide to go up and will I be able to go up more? Does anyone know anything about the Epox beeping?

when my epox beeped it meant "hey, you f'ed up, change it back" and was related to a bad oc i think, the cpu in my case

otherwise i go by the led codes....no led on the 8rda+?

anyway, one thing i dont think ill every try is upping the voltage, i think ill kill something when i do

 

[amcdonald]

I also suspect a faulty psu since the crashing appears to be somewhat at random.
If you have another one to test out that would be great.
The 8rda+ comes with USDM which will report your voltages from windows.
Did you try lowering the FSB? Drop to 166 and see if you are able to OC the Chip.

 

[Ilmater]

Originally posted by: amcdonald
I also suspect a faulty psu since the crashing appears to be somewhat at random.
If you have another one to test out that would be great.
The 8rda+ comes with USDM which will report your voltages from windows.
Did you try lowering the FSB? Drop to 166 and see if you are able to OC the Chip.

I checked my voltages in my BIOS this morning:

+5:_____4.81 - 4.83
+12:____11.85 - 11.91
-12:____-11.86 - -11.95
-5:_____-5.20 - -5.25
VBAT(V):_3.07 - 3.08
5VSB(V):_5.51 - 5.53

Isn't that OK?

 

[amcdonald]

Set USDM to log (if it does) and change the voltage refresh to 1 sec.
Then run a test that will probably crash the computer.
Check for a voltage spike before the crash.

 

[Ilmater]

Originally posted by: amcdonald
Set USDM to log (if it does) and change the voltage refresh to 1 sec.
Then run a test that will probably crash the computer.
Check for a voltage spike before the crash.

I can't even post at anything over 1.7GHz, but I'm stable as a rock at 1.7, so I can't run anything that will crash the computer. I ran 3DMark and Prime95 yesterday all day and not one crash.

 

[Jeff7181]

Those voltages look ok...

You haven't mentioned what type of RAM you're using have you? RAM can cause lots of instabilities, including no POST.

 

[Ilmater]

I'm using Corsair XMS. It's in my system rig profile (see my sig).

Well, I finally got one more proc to try: the 0319 from SVC. Actually, I've had it for about a week or so, but I just put it in last night. It stopped at the exact same place. Would not go higher than 8x at a 200MHz FSB. So, as I was desperate, I lowered the FSB and tried again.

To make a long story short, I'm running at 190MHz FSB with a 12.5x multi (2375MHz if you don't want to do the math).

That means that something else is to blame. I know it's not my memory as it can go to 200MHz without breaking a sweat, and I know it's not my PSU. One option is that it's the CPU. This is doubtful as it can run at 2375MHz and run at a 200MHz FSB, just not at the same time. Plus, my last CPU stopped at the same place as this one.

That leaves the probable culprit to be the motherboard. Is there a possibility that a different BIOS might help?

 

[Gage8]

what's your bios version? go to NforcersHQ.com and download the uber 1004 bios

 

[AkumaX]

tbred A?

 

[4x4expy]

Have you tried running a single stick of your memory? (I'm assuming you don't have a single 1gb stick)

 

[Ilmater]

Originally posted by: 4x4expy
Have you tried running a single stick of your memory? (I'm assuming you don't have a single 1gb stick)

That is correct, and no I haven't, but why?! It runs just fine at 200MHz 2-2-2-5 when the multiplier is low. You'd have a hard time convincing me that when the multiplier goes up it somehow affects the memory.

Originally posted by: Gage8
what's your bios version? go to NforcersHQ.com and download the uber 1004 bios

Has anyone used this here? Any testimonials? BTW, I'm using 3507 now.

 

[4x4expy]

If it is stable at 200fsb with a low multiplier, then that probably is not the answer. Most agree though that max testing of the CPU should be done with the simplest memory configuration with relaxed timings to suppress any potential problems from the ram. That way you can get a better idea if it is your CPU hitting a wall or something elsewhere. Because unfortunately tests like P95 gives you know clue as to the reason for failing. Once the max capability of the CPU is identified then start to tighten then tweak the memory for the best possible.

 

[Ilmater]

What I did was get the highest FSB that I could, and then try to ge the highest multi I could. However, it didn't work like that.

What I don't understand is why the processor won't overclock past a certain point when the memory and motherboard can CLEARLY handle a certain FSB.

 

[joshg]

Well...

I'm not sure if this has relevance to this or not, but...

Once I read where a guy was having weird problems where things started going screwey at even FSB settings (I.E. 133, 166, 200)... something about it thought he was running a different proc? I remember he said that he tried it at 202 (since that doesn't fall on a proc's default speed) and things worked fine....

Just an idea, but like I said I'm not sure if this has any relevance here or not

 

[Ilmater]

Originally posted by: joshg
Well...

I'm not sure if this has relevance to this or not, but...

Once I read where a guy was having weird problems where things started going screwey at even FSB settings (I.E. 133, 166, 200)... something about it thought he was running a different proc? I remember he said that he tried it at 202 (since that doesn't fall on a proc's default speed) and things worked fine....

Just an idea, but like I said I'm not sure if this has any relevance here or not

Actually, that's interesting. I might try that.

 

[Ilmater]

IT WORKED! It's 100% FINE at 199MHz FSB. At 200MHz FSB it takes forever to post and it posts the processor as 1100MHz, but at 199MHz, I'm currently stable at 2.388GHZ. Incredible. I never thought that would happen.

Well, that lends definite creedence to the idea that this is a real phenomenon, but the question is, why? Does anyone have any idea?

EDIT: BTW, I can boot into windows at 201MHz FSB too, but the voltage required isn't worth it.

 

[BmXStuD]

I just got my new 1700+ to 2ghz speeds from 1.46ghz.
\specs below (sig'd)