Asrock 4CoreDual-SATA2

[imported_weblurker]

Originally posted by: brokencase
Weblurker,

I can up the bios fsb to the point where XP won't boot. However I find I can go higher if I set it with the fsb utility after the OS is up and running. I encounter no problems when I do this. I haven't seen any drive related issues and Prime 95 seems stable, even at 2.7ghz.

So, for the most part, I just leave it at 266 in the BIOS. If I need to go faster for a certain application I can use the fsb utility to do so.

Yes, I see. I was wondering if you were seeing the same problem I'm having with the E2180 and the SATA2, that the BIOS is messing up the cpu frequency. But it sounds like the SATA2 BIOS is handling the Q6600 correctly.

I don't have CPUid but I am running a Q6600 ACLR. What did you want to know from CPUid?

I had a theory that the new M0 stepping might have caused the SATA2 BIOS to set the wrong parameters for the cpu and memory, I was curious to see if your Q6600 was also an M0 part. But I read elsewhere that the quad core parts have their own entirely different stepping label.

Also, I read in another forum that someone successfully ran an E2180 on the older VSTA motherboard, the VSTA BIOS was correctly calculating the cpu frequency as 10*FSB. So it looks like the problem I'm having is limited to the E21xx cpus on the SATA2 BIOS.

 

[snor]

nt

 

[BAUBAU]

Originally posted by: weblurker

Originally posted by: BAUBAU
E2180 + 2x256MB DDR400 (kingmax pc3200u/ddr433) on 4core dual-SATA2.
BIOS 1.60A modded to have speedstep and thermal throttling enable/disable
Could this have helped with my clock speed beeing ok?

Hi Baubau.
I'm also running an E2180 but I find that the BIOS doesn't generate the correct cpu speeds. When I set the FSB host frequency to 266Mhz, the BIOS sets the cpu to 1995Mhz instead of 2660Mhz.
Did you seeing proper cpu speed calculations from the original Asrock v1.60 BIOS?
Edit: I tried v1.60a. It did the same thing as the original Asrock v1.60. It doesn't set the cpu to 10*fsb, instead it sets the cpu to 10*0.75*fsb, or, as I suspect, 10*(200/266)*fsb.
Here is an easy test. I set the host frequency setting option to "Manual" and manually set the fsb host frequency to 200. That is the default setting for the E2xxx series of cpus and should be identical to the BIOS default setting for the E2180.
However, when the BIOS next starts, it has set the cpu to 1509Mhz instead of 10*200Mhz=2000Mhz (same as NeonGerbil found).

I have no such issues. We have the same board, the same CPU.
So stop blaiming the BIOS. Blame the hardware. Things that might have happened:
- LGA socket inside mainboard is not clean or is damaged, CPU not recognised properly
You should test another cpu in (a pIV would do just fine)
- CPU is not clean or is damaged.
You should test it in another mainboard (ask a friend or go to a pc repair shop)
- CPU has strange revision not recognised by the current BIOS
You think this now, wich I found unlikely. My E2180 has:
PACK DATE:09/18/07 VERSION:E14253-001 MM:892608 and PC BX...SLA8Y (wich is M0)
- some electric components on the board are broken (power limiter... etc)
Unlikely, this would mean RMA as only option. Just try to remount the cpu on the mainboard, clean a litle and try again.
What are your other components? PSU + RAM + VGA + HDD A small chance exists that you get a low rating because of total power drained by the PC (or lack of), so the BIOS is downclocking your CPU.

Some BIOS settings i use: (overclocked until 290FSB, all fine)
^^ CPU, PCIE Async ; CPU Freq=266* ; PCIE Freq=100 ; PCI Freq=33,33 _
^^ Spread Spectrum=Disabled ; Boot failure=Enable ; _
^^ No-Execute Memory Protection=Enable ; Max CPUID=Disabled ; _
^^ Enhanced Halt State=Enable ; CPU Thermal Throttling=Enable ; Intel Speedstep=Enable ; _
^^ DRAM Freq=200* ; DRAM CAS=2* ; TRP=3* ; TRCD=3* ; TRAS=8* ; _
^^ TRFC=16* ; TRRD=3* ; TRTP=3* ; TWTR=2* ; TWR=3* ; DRAM Command Rate=1T* ; _
^^ Flexibility=Disabled ; DRAM Bank Interleave=4 Way; DRAM Bus Selection=Auto ; _
^^ All advanced on auto ; Dram voltage=Normal ; AGP voltage=Normal ; _
^^ Vlink=Fast; IDE Drive strenght=Ultra ; PCIE downstream=Disabled.

 

[imported_weblurker]

Originally posted by: BAUBAU

Originally posted by: weblurker

Originally posted by: BAUBAU
E2180 + 2x256MB DDR400 (kingmax pc3200u/ddr433) on 4core dual-SATA2.
BIOS 1.60A modded to have speedstep and thermal throttling enable/disable
Could this have helped with my clock speed beeing ok?

Hi Baubau.
I'm also running an E2180 but I find that the BIOS doesn't generate the correct cpu speeds.

I have no such issues. We have the same board, the same CPU.
So stop blaiming the BIOS. Blame the hardware. Things that might have happened:

I'm not the only person who has had this problem. If you go back in this thread a few pages and look at posts by NeonGerbil and Farfle, they had the same problem as I've experienced.

I didn't post all my notes about the issue (which now run about 10 or 15 pages) but I'm very sure that neither the cpu nor the socket are damaged.

I'll add some more details about the tests I've run.

When I use the F9 "set defaults" option, the system comes up properly at 2000Mhz. That must mean the cpu is recognized correctly.

Test: compare automatic cpu setting to manual setting.

1)Use F9 load defaults, BIOS comes up showing the cpu is set to 2000Mhz.

2)Set cpu host frequency to "manual", set host frequency to '200", BIOS comes up showing the cpu is set to 1508Mhz (for some reason, the actual number varies from 1500 to 1509).

Those two settings should result in identical settings, the cpu set to 2000Mhz. The automatic setting works but when the BIOS calculates the setting, it comes up with the wrong answer.

The maximum cpu host frequency the BIOS will allow is 340Mhz. When I use that setting, the cpu is set to 2550Mhz, or 10*(200/266)*340.

I've set the cpu to run at 2550Mhz and run OCCT for 12 hours and Orthos for 24 hours. Then I set the cpu to 2660Mhz using setfsb and again ran OCCT for 12 hours and Orthos for 24 hours. No errors came up. Because of that, I think it's extremely unlikely that the cpu or socket is damaged. The long error free OCCT and Orthos runs also should show that the hardware is stable, including the power supplied by the PSU.

............

The other problem I've run into is that the system won't come up out of a power up when the dram is set to DDR400.

Test: System boot after turning power on.

1) Set cpu host frequency to "auto". BIOS sets cpu to 2000Mhz. Turn power off, then turn power on and the system boots normally.

2) Set cpu host frequency manually to 200Mhz. Save and exit the BIOS. The BIOS does not come up after the warm boot. It beeps three times, indicating a lower memory fault. The system won't boot up until the reset button is hit and when it comes up, the BIOS says the cpu is set to around 1500Mhz. However, cycling the power will change the cpu frequency to 2000Mhz.

3) Set cpu host frequency manually to 266Mhz, BIOS sets cpu to 10*(200/266)*266 = 2000Mhz. It should be set to 2660Mhz but 2000Mhz is the correct default speed of the E2180, which is the same cpu setting as part 1) of this test. Cycle the power. The BIOS won't come up and beeps 3 times. Again, hit the reset button and the system boots up.

The BIOS automatic cpu parameter set feature works but setting the cpu host frequency manually results in cold boot problems, even when both the automatic and manual settings result in the same CPU frequency, 2000Mhz.
............

CpuID reports that the E2180 I have is the M0 rev, same as your 2180. In fact, the data on the package is virtually identical to what you have:

Version # E14253-001 Pack Date 09/13/07.

The hardware I'm running is nothing high powered, Asus fanless 6600 video card, one 250G Seagate HD and Samsung DDR400 ram rated at 3,3,3,8,2T. The E2180 and Asrock replaced a Sempron+DFI Infinity NF4X, everything else remained the same. The Sempron+DFI only used 90W measured at the AC plug. The PSU is an NSpire 350W, so power shouldn't be an issue.

The BIOS settings I'm using are the same as yours, except for the dram settings which are slightly slower.

Farfle solved this problem by implementing the BSEL cpu mod, which effectively made the cpu into a 266Mhz part. I suspect that worked because the cpu was forced to match what the BIOS may be erroneously assuming about the E21xx cpus, that it is a 266Mhz FSB cpu.

However, the fact that you haven't experienced the same problem as NeonGerbil, Farfle and me is extremely interesting.

All my tests seem to show that the BIOS is miscalculating the cpu frequency and probably the memory parameters as well. The F9 set defaults and automatic cpu host frequency work properly, none of the manual cpu host frequency settings work correctly.
................

BTW, is it easy for you to examine parts on your motherboard? If so, could you check the part number of the PLL? That part is right below the purple PCIe x4 socket and beside the PCI socket that is above the silver 2032 coin cell battery.

The reason I ask is that a friend has the older VSTA motherboard and that board used an RTM682-800 PLL and he didn't have any problems with his E2160 cpu settings, but Farfle had problems with his E2160+SATA2. The SATA2 motherboard I have uses an ICS953002DFLF PLL, a different part number from the VSTA.

I wonder if Asrock changed the PLL part way through the SATA production run? If so, it's possible the BIOS has to be updated to reflect the new clock generator.

 

[BAUBAU]

Asrock makes cheap mainboards, that means they use not-so-good parts, but that is not the issue.
The thing is they have a not-so-large stack of parts. When they run out of them, they order replacements, the cheapest price/performance they can find. It means that in the next batch of mainboards, some parts are replaced by compatible but not identical ones. Usually the same function IC can be found from lots of manufacturers with lots of different names, but esentially all of them perform the same. So it does not mean the next mainboard batch will be necessarily better/worse, because all of them goes into testing to assure at least stock functionality.

If your board has flaws (and you are shure of that), stop wasting time, you dont have the tools to figure it out. Some google after part numbers and comparing is also superficial because they dont describe how parts react on different applications. Is enought to know that other people have this mainboard with E2180 and it works just fine. If yours does not, RMA it now.

My mainboard:
4CoreDual-SATA2 with Rev G/A 1.02 (written between pci slots) and on the BIOS sticker P1.10-11B
LFE8674 0720L (for power near PSU connector)
RTM682-800 (for memory slots near them)
RTM866-890 (for chipset near PCIE slot)
[no easy task, had to open PC/dismount cards]

 

[imported_weblurker]

Originally posted by: BAUBAU
Asrock makes cheap mainboards, that means they use not-so-good parts, but that is not the issue.
The thing is they have a not-so-large stack of parts. When they run out of them, they order replacements, the cheapest price/performance they can find.

My mainboard:
4CoreDual-SATA2 with Rev G/A 1.02 (written between pci slots) and on the BIOS sticker P1.10-11B
LFE8674 0720L (for power near PSU connector)
RTM682-800 (for memory slots near them)
RTM866-890 (for chipset near PCIE slot)
[no easy task, had to open PC/dismount cards]

Many thanks BAUBAU for taking the time and effort to find out the part number of the chips and posting them. That was excellent and effective work. I think the difference between our two motherboards has answered the question, almost certainly completely.

Your motherboard uses the RTM682-800 PLL, exactly the same as what is on my friend's VSTA motherboard, which worked correctly with his E2160. As you mentioned, Asrock must have changed parts at some point and the new PLL must not be exactly the same as the RTM part.

I suspect my SATA2 motherboard is a later version than yours, the sticker on the BIOS read v1.50.12B but the board itself was labelled Rev G/A 1.02, same as yours. It looks like Asrock thought the PLL chips were identical but they probably aren't. The current Asrock BIOS works correctly for the E6xxx and E4xxx cpus, there must be a bug in how the BIOS handles the ICS PLL and the E21xx cpus.

I've been having an email conversation with someone at Asrock for the past 10 days about this problem. But I think with the information you've provided, a new version of the BIOS will surely be coming out shortly.

I'll post in this thread how things turn out.

If a BIOS change fixes the problem, then it would appear that the BSEL mod won't be necessary to make things work properly with the E21xx cpus and the SATA2 motherboard.

Thanks again BAUBAU.

 

[martianmisfit]

asking again, can i get some help please;

i'm in need of a power supply that will be compatible with my new pc.
can you guys/gals that are running the following specs list your power supplies?
mine keeps going out and "corrupting" my bios. i have to clear my cmos every time i turn on the pc.

Asrock 4CoreDual-SATA2
Q6600

thanks

 

[m00ndancer]

Zalman ZM600-HP 600W

 

[LLopps]

Anyone know if the ASROCK 4CoreDual SATA2 supports the new E4600, I see that it is a 65watt processor?

 

[kekewons]

Unaware of a reason it wouldn't.

I'm running a E4500 here without problems (haven't tried overclocking it yet though).

k

 

[TungFree]

I read the Asrock 4CoreDual-SATA2 thread and am unclear about the Q6600 with it.
I do not play games nor over clock
Presently I have a Athlon 2 year old CPU and an asus Motherboard that does not let me use my ATI AGP card. I have a ATI AIW 9800 pro 128meg one AGP
The Q6600 quad enchants me and I will not yet spring for the new $1000 one
that is barely out. The Q6600 price drop makes me want to act on it.

Though I build my own computers I am a Teacher and time forbids me to be fully informed on what is the best to do with my money on a new system.

I use Photoshop lots of CD and DVD burning and a dozen windows open in IRC chats.

I posted another tread asking advice about motherboard and this oine was recommended.

What ram should I get for it? and how much? What size of power supply? I have some old power supplies about 350 watts in the closet, I have no Ram sticks.

As I read the ones here with Q6600 they all had problems setting things up and the advice was to set the Bios to default as it comes.

Will I have problems if I do that, what ?
I need Ram recommendations
Will my system run slower set to default as it comes on the board ( slower than the specs say on the Q6600?)

 

[BAUBAU]

PSU recommandation:
- DO NOT USE OLD POWER SUPPLIES FROM THE CLOSET!
buy new, certified PSU with high efficiency 80%+
Most of Seasonic, Corsair etc. brands are good
Litle help on http://forums.extremeoverclock...?t=136602&page=1&pp=20

Ram recommandation:
- any brand-name DDR2 666 that defaults to 1.8-1.9V voltage requirements should work
Personally I've tested Corsair CM2X1024-5400C4 and works wonderfull. Highly awarded for real, and cheap too. 2x1GB is good for now.

Marginally slower by default, but you can always tweak it and even overclock it to gain more performance.

 

[colonel]

nice, I dont want to pay $100 dollars for a mobo

 

[brokencase]

TungFree...

I was in a similiar situation as you. I have an ATI AIW 9000 AGP. I upgraded from an old dual talutin 1.4ghz board. I also jumped on the Q6600. So, this was a big jump for me.

I also don't game (very much), I do some video transcoding for my portable media player, plus a lotta of CAD/CAM work.

If you go for this board & Q6600, buy a matched pair of DDR2-677 memory (2 gig) and if your power supply is old get a new one, at least 500W. I have no problem running the Q6600 at the rated 2.4ghz and higher. ASRock indicates the Q6600 may run at 5% below rated, but I think that happens if you use older DDR 400 memory.

It used to take 40 minutes or so to trancode a movie for my PMP. It now takes 11 minutes! Haven't tried coping a DVD yet.

I might add that the ATI Catalyst driver runs optimally on this board. 4x agp and all the performance options enabled in ATI's SMARTGART.

This board also has a very good built in audio chip.

It's a good bargain if you want to hold onto your agp. Go for it. If you have any trouble during the upgrade just post here and I will respond. The only downside is the PCI express implementation is substandard on this board. So if you upgrade video in the future you may not get the most out of it.

 

[Hazelnut]

I suspect my SATA2 motherboard is a later version than yours, the sticker on the BIOS read v1.50.12B but the board itself was labelled Rev G/A 1.02, same as yours. It looks like Asrock thought the PLL chips were identical but they probably aren't. The current Asrock BIOS works correctly for the E6xxx and E4xxx cpus, there must be a bug in how the BIOS handles the ICS PLL and the E21xx cpus.

Thanks for this weblurker! I was just trying to overclock my new 4CoreDual-SATA2 + E2140 that I built last night (to replace a XP2200+) and I got exactly the same behavior. Really confused me to get 340MHz Host freq in the bios give a 255MHz FSB.

So, can you confirm that ASRock have replicated this problem? And do you have any updates on when they might be releasing a new BIOS?

It's a real shame in a way that I needed to use this board to keep my AGP 6600gt and DDR ram, but I can't afford to upgrade it all at once. I knew that I'd probably be limited to getting the E2140 up to 2.4GHz with this board, but 2GHz is even more disappointing.

 

[imported_weblurker]

Originally posted by: Hazelnut

I suspect my SATA2 motherboard is a later version than yours, the sticker on the BIOS read v1.50.12B but the board itself was labelled Rev G/A 1.02, same as yours.

Thanks for this weblurker! I was just trying to overclock my new 4CoreDual-SATA2 + E2140 that I built last night (to replace a XP2200+) and I got exactly the same behavior. Really confused me to get 340MHz Host freq in the bios give a 255MHz FSB.

So, can you confirm that ASRock have replicated this problem?

I knew that I'd probably be limited to getting the E2140 up to 2.4GHz with this board, but 2GHz is even more disappointing.

Hi Hazelnut. I was considering writing another post summarizing the problems I've had with the SATA2 and E2180, but I wasn't sure if the problem would affect anyone else. I'll post a summary later today or tomorrow. I was hoping that Asrock would fix the problem quickly.

I'm glad you posted your result because I wasn't sure if the problem would occur with the E2140. I'll email Asrock that a user with an E2140 has experienced the same problem. IIRC, the E2140s are at the L2 rev. I didn't know if the bug only occurred with the new M0 stepping.

I sent Asrock a long report on the problems I had, plus links to the posts from this forum from other users and exact instructions on how to replicate the bug. I sent that email on Sunday (Monday in Taiwan) but I haven't heard anything from them. The other times I've sent Asrock reports about the bug, they sent me test versions of the BIOS the very next day. I'll send them your result, that the SATA2 board is also buggy with an E2140, and see if they reply.

I did a lot of testing with the SATA2 at 255Mhz (set with the BIOS at 340Mhz) and 266Mhz (set with SetFSB) and it worked perfectly. I ran Orthos for 24 and 36 hours and OCCT for over 12 hours at those FSBs and no errors came up.

You should be able to use SetFSB to bump the FSB to a higher value than 255Mhz. You'll have to set the PLL in SetFSB to the ICS chip.

I found that most apps work after using SetFSB, except programs that need precise timing. I have an HDTV Wonder and when I used SetFSB to set the FSB to anything other than what XP booted up with (higher or lower FSB), the audio and video quickly went out of sync.

I wouldn't be surprised if your E2140 is able to successfully run at close to 2.4Ghz with the FSB around 300Mhz. It seems most of the current Core Duo cpus end up overclocking into the 2.4-2.6Ghz range. I suspect there are two limits, the VIA chipset has a 300-310Mhz limit and how much voltage is being delivered to the cpu, which can't be changed by the user on the SATA2.

I have a 10x E2180 and I can get the FSB to 266Mhz with no errors but it won't get past 272Mhz, so the limit is about 2.66Ghz.

Someone else has a 9x Q6600 and he can get the FSB up to 290Mhz, or 2.61Ghz.

I remember reading someone tried a 11x E4500 and he couldn't get past 240Mhz, or 2.64Gz.

So your 8x E2140 will probably bump up against the 300-320Mhz VIA chipset limit, but that will probably get you to what you wanted, 2.4Ghz.

 

[brokencase]

I don't think you'll get very far with your report to asrock.. Heck, you guys want to overclock your E2140, E2180's. But everybody knows this board is not an "overclocker's delight". Even better, your trying to do it with with DDR400 memory and you can't figure out why the BIOS won't let you do it!

Does the board run those cpu's at the rated speed? Yes. That's it. Your there, dude. The manufacturer has met their obligations.

When I fired up my Q6600 it defaulted at the 5% underclock (as indicated on asrocks website), I thought "gee whizz. I guess that's it then." I then went into the bios and changed the fsb to 266 and it ran fine. Then I thought "That's wierd, I thought they were'nt going to let me run at the rated speed".

Here's my guess. For folks wanting to hold onto thier old DDR400 and want to run a newer Quadcore, it will work but at 5% underclock. For those that spring for faster DDR2 533/677 memory the bios will release the limitation and even allow for some amount of overclock.

Now, is that such bad deal?

I suspect similiar situation with the E2140, E2180. Your trying to crank the fsb beyond what your memory can handle, the Bios knows this from the spd, and the bios failsafe is kicking in.

Remeber that old saying... "Horsepower cost money. How fast do you want to go?"

 

[imported_weblurker]

Originally posted by: brokencase
I don't think you'll get very far with your report to asrock.. Heck, you guys want to overclock your E2140, E2180's. But everybody knows this board is not an "overclocker's delight". Even better, your trying to do it with with DDR400 memory and you can't figure out why the BIOS won't let you do it!

Does the board run those cpu's at the rated speed? Yes. That's it. Your there, dude. The manufacturer has met their obligations.

I suspect similiar situation with the E2140, E2180. Your trying to crank the fsb beyond what your memory can handle, the Bios knows this from the spd, and the bios failsafe is kicking in.

Remeber that old saying... "Horsepower cost money. How fast do you want to go?"

Actually, brokencase, as I mentioned in my previous post, when I used SetFSB to set the cpu host frequency to 266Mhz, the E2180 with DDR400 ran for 24 hours under Orthos and 12 hours under OCCT. It works overclocked to the same speed as your Q6600, which is about 2.6Ghz. CpuID shows that the DDR400 memory is running within spec. At 266Mhz FSB, CpuID said the memory was running at 199.5Mhz, just below the 200Mhz it was designed to run at.

As I'm sure you knew, the entire Intel line of Core Duo cpus are all the same design, all produced on the same production lines and all are designed to run at the same range of voltages and FSBs, Intel just varies the size of the cache for each cpu price segment.

I know that the E2180 can run for many hours under Orthos and OCCT at 266Mhz. The only problem is that the BIOS won't let me set the FSB to 266Mhz, the bug limits the FSB to 255Mhz.

I could live with that problem, 255Mhz isn't that much different from 266Mhz. The really annoying bug occurs when I set the FSB manually to any speed and the SATA2 won't boot up properly. It puts out three slow beeps in a loop until I hit the reset button. That gets old really quick.

The reason I think that's a BIOS bug is that if I set the FSB manually to 266Mhz, the BIOS sets the cpu to 2000Mhz (which is wrong BTW, it should be 2660Mhz), which happens to be the stock speed of the cpu. But under those conditions the SATA2 won't boot up until I hit the reset button.

Set the cpu automatically to 2000Mhz and the SATA2 boots up fine.
Set the cpu manually to 2000Mhz and the SATA2 won't boot up.

That would seem to strongly suggest that the BIOS calculation of the cpu/memory parameters is wrong. Doing it automatically or manually to the same cpu speed should result in the same behavior. But it doesn't.

BTW, the BIOS does try to do the right thing with the memory dividers.

At the stock 2000Mhz cpu speed and 200Mhz bus speed, the memory divider is set to 1:1 for DDR400, which is correct since DDR400 memory runs at 200Mhz.

At anything above a cpu host frequency of 200Mhz, the BIOS sets the fsb : dram memory divider to 4:3. At an FSB of 266Mhz, the 4:3 fsb : dram divider that the BIOS chose was correct, 4:3 ratio means 266:200, a 266Mhz FSB and 200Mhz memory, which was exactly right.

The issue isn't whether or not the system runs overclocked. It does. The issue is that the BIOS miscalculates the host cpu frequency anytime you manually set the FSB.

For example, when I manually set the cpu host frequency to 200Mhz (i.e. at the stock speed of the E2180) the BIOS sets the cpu frequency to 1500Mhz, not 2000Mhz. That's clearly wrong.

As to the issue that if the system runs at the stock speed and doesn't work correctly at any other speed, you have a point. Asrock could tell all of us with E21xx cpus to stuff it.

The problem is that people with SATA2 motherboards with RTM PLLs calculate the cpu speeds properly, the VSTA motherboards work properly, 266Mhz FSB cpus and the SATA2 work properly, it appears the E4xx cpus and the SATA2 work properly.

The expectation I and others had was that since the SATA2 BIOS was able to properly calculate the cpu frequencies for all other Core Duo cpus, it should be able to do so with the E21xx cpus. Was that a reasonable assumption? I think so.

Edit: I just read over the emails that the Asrock BIOS programmer sent me. They were very terse but I think I see the pattern, it seems he thinks the issue is whether or not the SATA2 is able to overclock. I don't think he realized until my third email that the issue has nothing to do with overclocking, the BIOS is simply miscalculating the cpu frequency and in an especially obvious way.

As I mentioned before, set a 10x multiplier, 200Mhz E2180 to 200Mhz FSB in the BIOS and the BIOS sets the cpu speed to 1500Mhz, not 10*200Mhz=2000Mhz.

They really should have caught this bug long before the BIOS was released to the public.

 

[kddoj]

Has anyone tried the new 8800GT in this board yet? I would be really interested to know if it works and with what bios revision...




Kris

 

[Radski]

Well gentlemen, after reading this thread off and on for the past 2 months I'm starting to wonder at the sanity of spending 100's of hours poking and twiddeling around just to get past 3.0 GHz with a $80 processor and a $60 motherboard. Is it worth the trouble?????

The bottom line is, this is a budget system. I'm not sorry I bought it 'cause it got me from an old Athlon XP 2700 into a dual core that does twice as good, stock.

But spending dozens of hours screwing around trying to make a silk purse out of a sow's ear is not worth the time and effort. Aside from maybe doing the BSEL mod to upgrade to 1066 FSB, I'm not interested in spending any more time on this. A quality motherboard with all the proper setting already in the BIOS is a much more reasonable and rewarding effort in overclocking.

Good luck to you boys: count me out of this project. I'm getting old enough to realize that buying quality stuff the first time is cheaper in the long run than trying to modify budget crap. The only reason I went this route was because I have that OCZ EL DDR-400 which was, in effect, an effort to get more performance out of my old XP.

 

[brokencase]

Originally posted by: weblurker
I could live with that problem, 255Mhz isn't that much different from 266Mhz. The really annoying bug occurs when I set the FSB manually to any speed and the SATA2 won't boot up properly. It puts out three slow beeps in a loop until I hit the reset button. That gets old really quick.

They really should have caught this bug long before the BIOS was released to the public.

Why can't you just boot normally (with it at the default stock setting) and then use the fsb utility to bump it up after it boots? Like we discussed previously.

Hopefully we'll see some new bios revs from Asrock. Last one was at the end of Sept. Looks like the only changes they are making is to tweak it to work with various PCIe video cards. That's probably a higher priority than your issue. I suspect that being the case of this boards limited implementation of PCIe, they are trying to get as many video cards as possible on thier qualified list.

 

[imported_weblurker]

Originally posted by: brokencase

Originally posted by: weblurker
I could live with that problem, 255Mhz isn't that much different from 266Mhz. The really annoying bug occurs when I set the FSB manually to any speed and the SATA2 won't boot up properly. It puts out three slow beeps in a loop until I hit the reset button. That gets old really quick.

They really should have caught this bug long before the BIOS was released to the public.

Why can't you just boot normally (with it at the default stock setting) and then use the fsb utility to bump it up after it boots? Like we discussed previously.

SetFSB works quite well and the E2180 and SATA2 ran Orthos and OCCT for 24 and 12 hours at 266Mhz.

But I found that when SetFSB changes the cpu speed, XP doesn't know about the change (unlike CpuZ, which checks dynamically). I use HDTV video apps that use the reported cpu speed to calculate screen refresh rates and they quickly lose video and audio sync when the fsb is changed after boot up.

The other problem is that if I boot up at the 200Mhz stock speed, the DDR400 memory is set at 1:1 fsb : dram at 200Mhz. So if I raise the fsb beyond 200Mhz to 266Mhz, memory access fails because it will be running at 266Mhz instead of 200Mhz.

This bug, as often happens, creates a situation where you can't get everything working at the same time.

I'm running the machine at 255Mhz and it doesn't boot up properly.

If I raise the fsb to 266Mhz, video apps fail.

If I run the ddr400 at ddr333 the Asrock boots up properly but I find that everything runs about 15% slower, losing most of the overclocking speed increase.

So, until Asrock fixes this problem, I'm doing a power up two step each time I turn on the machine. Power up the Asrock, wait until it starts beeping, then hit the reset button.

 

[imported_weblurker]

Originally posted by: Radski
Well gentlemen, after reading this thread off and on for the past 2 months I'm starting to wonder at the sanity of spending 100's of hours poking and twiddeling around just to get past 3.0 GHz with a $80 processor and a $60 motherboard. Is it worth the trouble?????

I absolutely agree with you, Radski. Had I known what I know now about the SATA2 with the ICS PLL, I would never have bought this motherboard. I would have waited until Asrock sorted out the BIOS problems, or just ditched the DDR ram I had and gone with DDR2 and another brand of motherboard.

Come to think of it, I would have just kept running the overclocked Sempron I was using until the new 45nm Intel cpus come out next year.

Unfortunately, having bought and installed the board, as a Greek general once said, "it's a bad plan but we can't change it". If I RMA the board, I'll probably just get another SATA2 with the same ICS PLL.

But I'm seriously thinking about pulling out the Asrock motherboard and returning it and getting something else.

 

[Hazelnut]

I'm RMA'ing my board. I contacted ASRock support about the BIOS bugs with the PLL and have not had any reply. Also, even without any overclocking I was getting BSOD's and freezes in games more and more as I used it more. So, it's out of my PC now and I'll be wrapping it for return tonight. Gonna have to bite the bullet and get new ram/gfx as well as cpu & motherboard.

What a waste of time!

Anyway, thanks to everyone who posted info, especially weblurker - it's been incredibly helpful.

 

[imported_weblurker]

Originally posted by: Hazelnut
I'm RMA'ing my board. I contacted ASRock support about the BIOS bugs with the PLL and have not had any reply. Also, even without any overclocking I was getting BSOD's and freezes in games more and more as I used it more.

Hi Hazelnut. It's very disappointing that you've contacted Asrock technical support and they haven't replied. Recently, that's happened to me as well.

I first sent them a bug report on Oct 16 and received a test version of the BIOS the next day, which didn't fix the bug. I emailed them about the failure and they sent another test BIOS the next day, which also didn't change anything.

What disturbed me was the message that came with the second test BIOS. The Asrock rep said that the SATA2 worked at the stock 2Ghz speed, that the BIOS might not display the correct cpu speed anyway and that they couldn't guarantee overclocking performance. Then he asked me to test the new BIOS version and as a postscript, added "BTW, we don't guarantee overclocking". So on the one hand they're saying that there really isn't a bug but on the other hand they send me a new BIOS to try.

It looked like they were laying the foundation for the infamous figure skating move, The Triple Weasel. Asrock tried to say that I'm not seeing a real problem, it works anyway and even if I am seeing a problem, it's in an area that they don't support.

Since then I've sent them links to posts from other users from this forum and other discussion areas, exact test procedures for replicating the problem and benchmarks showing that the bug isn't a display problem but a real issue. They haven't replied to a single one of those emails.

The SATA2 v1.60, v1.60a and v.160b BIOS bug boils down to this, for an E2180:

Set the cpu frequency from auto to manual.
Set the fsb to 200Mhz.
The BIOS then sets the cpu to 1500Mhz.

Basically the BIOS is saying this: 10*200Mhz=1500Mhz, which is wrong.

If anyone wants to return an SATA2 motherboard, that demonstration should be enough to convince a vendor to give you a refund.

The memory settings, which seem to be derived from the incorrect cpu frequency setting, are also wrong, which in my case results in the motherboard never booting up cleanly after turning the power on.

I'll send Asrock one last email, asking them if they're going to fix this bug or if they've decided to leave the BIOS as is.