AOA Forums AOA Forums AOA Forums Folding For Team 45 AOA Files Home Front Page Become an AOA Subscriber! UserCP Calendar Memberlist FAQ Search Forum Home


Go Back   AOA Forums > Hardware > EPoX MotherBoards

EPoX MotherBoards EPoX Intel and AMD Motherboards.


Reply
 
LinkBack Thread Tools Rate Thread
  #1 (permalink)  
Old 11th February, 2004, 01:14 AM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

8rda+ v2.1, CPU voltage issue

I have a barton 2500+ (few months old, not multiplier locked), which currently I have at 2.37ghz (215fsb). I used magic bios, and have the latest bios release.

http://cafe.cic.hull.ac.uk/~icb01mp/MyTemps.JPG

This is my current voltage reading, with the vdd set at 2v, the memory at 2.9v (PC3200 twinmos, 2.5,3,3,8) and the CPU at 1.85v (windows reads 1.84v).

Now, if I set the voltage to 1.875V, the voltage reading in windows still reads 1.84v, and the system is a little unstable. If I set it to 1.9v or beond, the system simply doesn't boot. (As in, it powers on, all fans a-spinning, but no beep or screen output, no post)

Whats going on here, why can't get get more than 1.84v?

Seconday question:
Given that my memory is Twinmos 3200 cas 2.5, what is the maximum fsb that I should aim for, what is realistic for this RAM (which is currently at 215mhz, 2.5,3,3,8 @ 2.9v in bios, reading 2.83v in windows)?


System info:

2500+ @ 2.37ghz (215mhz fsb, sync)
PC3200 Twinmos RAM (764mb)
8rda+ v2.1 (Water cooled chipset, active cooled voltage regulators (copper heatsinks and low rpm 80mm red LED fan), Passive cooled south bridge)
Enermax 460W PSU
Koolance water cooling (200W lapped silver CPU water block, dust guard removed, all fans facing outward)
Radeon 9800SE AIW --> Radeon 9800 AIW pro (hardmod)
2x 7200rpm HDD
__________________

Last edited by MIP; 11th February, 2004 at 02:26 AM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #2 (permalink)  
Old 11th February, 2004, 03:25 AM
Member
 
Join Date: January 2003
Location: Peru, Illinois
Posts: 549
Send a message via ICQ to golfer49

Many of the new boards have overvolt protection, its not going to let you go any higher, it will just crash..............on the memory, try timings of 11-3-3 at cas 2 or 2.5, it should go in the 220 FSB range.
__________________
Golf is a good walk spoiled! ( Mark Twain )

Abit NF7-S rev 2.0 Barton 2500 AQZFA0340SPMW 11 x 214
w/SLK900A & 92 mm fan (38 deg C)
2 x WD400JD---Gainward Ti4200 8 X AGP
Twinmos PC3200 w/ Winbond chips 2 X 256
600 watt dual fan psu

Abit NF7-S rev 2.0 XP1700DLT3C 10 x 225
SK7 with 80mm smartfan ( 37 deg C )
Gainward FX5200 128 Mb DDR
2 x 40 Gb Maxtor ATA 133 drives
Buffalo PC3200 2 x 256 CH5 chips
650 watt dual fan power supply

WinXP Pro, SP1, nVidia 3.13 on both
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #3 (permalink)  
Old 11th February, 2004, 04:09 PM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

How do I disable the overvoltage protection?

I understand its cutting a pin, or removing a resistor, on the chip above the voltage regulators, but the picture in the guide I found shows a chip with 18 pins, and mine has 16 pins, and the layout of resistors around the edge is slightly different (though similar).

I only have a crappy webcam:

http://cafe.cic.hull.ac.uk/~icb01mp/8rda/Close.jpg
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #4 (permalink)  
Old 11th February, 2004, 04:51 PM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

It's not overvoltage protection, it's *overcurrent* protection.

The voltage regulator uses a resistor connected to pin 16 in order to sense the current passing through the MOSFETs. People have reported that replacing this resistor with one of a slightly higher resistance helps. Others have completely removed the overcurrent protection by either cutting traces or by removal of the current sense resistor.

Obviously removing the current sense totally means that the VCORE regulator won't shutdown, even if you short it. Instead, it'll just go pop, and take out the MOSFETs.

Edit: You will probably want to add some cooling to the MOSFETs (or replace them with MOSFETs with a lower resistance. The current sense helps prevent them from overheating.
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #5 (permalink)  
Old 11th February, 2004, 07:53 PM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

Quote:
Originally Posted by Áedán
It's not overvoltage protection, it's *overcurrent* protection.

The voltage regulator uses a resistor connected to pin 16 in order to sense the current passing through the MOSFETs. People have reported that replacing this resistor with one of a slightly higher resistance helps. Others have completely removed the overcurrent protection by either cutting traces or by removal of the current sense resistor.

Obviously removing the current sense totally means that the VCORE regulator won't shutdown, even if you short it. Instead, it'll just go pop, and take out the MOSFETs.

Edit: You will probably want to add some cooling to the MOSFETs (or replace them with MOSFETs with a lower resistance. The current sense helps prevent them from overheating.
Currently I've got copper RAMsinks on the MOSFET's, and an 80mm case fan directly above them blowing down on them, and they are cool enough to comfortably keep your finger on the heatsink. I'll put a copper sink on the vreg chip aswell when I've figured out what to do with it.

http://cafe.cic.hull.ac.uk/~icb01mp/8rda/Case.jpg

I would like to take the risk, and simply disable the overcurrent protection completly, so I can up my CPU voltage, however, I'm still unsure as to what to actually do. Like I say, the guide I've found appears to be looking at the early 8rda+ revisions, and I have version 2.1, with a different layout and vreg chip.

http://cafe.cic.hull.ac.uk/~icb01mp/8rda/vreg.JPG

I've drawn this to represent the 32pin chip that is on the 8rda+ v2.1.

Does anyone know which is the correct pin or resistor to remove?

I'm fully aware that I invalidate the warrenty and I own the problem if this goes wrong, obviously.

I'm no electrician, does this give anyone any clues?

Link
__________________

Last edited by MIP; 11th February, 2004 at 08:39 PM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #6 (permalink)  
Old 12th February, 2004, 01:30 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Just some notes/thoughts here...

Looking at the chip, it's a fairly comprehensive current/voltage limit system. It can handle transient current overloads across any of the three phases as well as average current overloads overall.

It looks like the voltage regulator has three seperate current sense inputs - that's pins 4,5 and 6.

Pin 7 forms the reference for the current sense opamps internally, but only for the transient response. I'm not sure that altering this would be of much benefit at this point in time.

Pin 15 forms part of the ILIM. This is much more interesting, as it forms part of the overall average current limiting system.

At a guess, there'll be a resistor from pin 16 to pin 15, and then another resistor from pin 15 to ground. This sets the reference voltage for the overall current limit. If this is true, then it should be possible to add a extra resistor between pins 15 and 16 to bring up the current limit.
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #7 (permalink)  
Old 12th February, 2004, 01:32 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Not sure that disabling overvoltage is really possible on this chip - at an initial look, the overvoltage protection is closely tied in with the voltage regulation. IE, disabling overvoltage protection might well disable voltage regulation too! See above for the overcurrent protection though!
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #8 (permalink)  
Old 12th February, 2004, 01:36 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

One last thing - ignore my initial comment about the overcurrent protection, as I'd been under the impression that the voltage regulator on the V2.1 was the same as V1.1.

As you've posted the link to the datasheet, I've since been working from that for the details of protection mechanisms.
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #9 (permalink)  
Old 12th February, 2004, 03:16 AM
Gizmo's Avatar
Chief BBS Administrator
BassTeroids Champion, Global Player Champion, Aim & Fire Champion, Puzzle Maniax Champion, Othello Champion, Canyon Glider Champion, Unicycle Challenge Champion, YetiSports 9: Final Spit Champion, Zed Champion
 
Join Date: May 2003
Location: Webb City, Mo
Posts: 16,178
Send a message via ICQ to Gizmo Send a message via AIM to Gizmo Send a message via MSN to Gizmo Send a message via Yahoo to Gizmo Send a message via Skype™ to Gizmo

If your goal is to disable overcurrent protection, then shorting pins 15 and 16 together will do it. I do NOT recommend this. If your problem is being caused by your overcurrent protection kicking in, then you are gonna toast some transistors.

You CANNOT disable the over-voltage protection, because it is inherent in the voltage regulation of the circuit. However, the sample circuit shows a resistor between pins 2 and 3 of the IC. Assuming your board uses a similar layout to the sample circuit (which it probably does), you can increase the value of this resistor slightly and increase your output voltage. I can't tell you how much to increase this resistor by, because the actual output voltage is controlled by a number of parameters. Since I don't know what the value of this resistor is in your particular mobo, I can't even give you a SWAG. The problem you run into with all of this is that, according to the manufacturer, this regulator requires dynamic compensation if the duty cycle of the switching FETs exceeds 50%. I don't know how this circuit is set up, so I can't say if this is going to be an issue for you or not. Increasing the voltage increases the current, and hence, the duty cycle of the FETs.

Bottom line, you are treading into dangerous territory. I don't have enough info to be able to advise you intelligently. Perhaps Aedan knows a bit more. If you want to pursue this, be CAREFULL! Or you will be replacing a mobo, like me.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #10 (permalink)  
Old 12th February, 2004, 03:32 AM
Gizmo's Avatar
Chief BBS Administrator
BassTeroids Champion, Global Player Champion, Aim & Fire Champion, Puzzle Maniax Champion, Othello Champion, Canyon Glider Champion, Unicycle Challenge Champion, YetiSports 9: Final Spit Champion, Zed Champion
 
Join Date: May 2003
Location: Webb City, Mo
Posts: 16,178
Send a message via ICQ to Gizmo Send a message via AIM to Gizmo Send a message via MSN to Gizmo Send a message via Yahoo to Gizmo Send a message via Skype™ to Gizmo

In looking at your post again, I'm really thinking that you are running into a problem with your current limit kicking in. Its possible that the circuit may be set up too conservatively to be able to reliably overvolt (after all, 1.85v is the max that AMD ever said a mobo had to support, and you are getting that, barely). If that is truly the case, then you can take Aedan's advice and add a resistor from pin 15 to pin 16, which will shift the current limit. Again, I don't know what value to add, because I don't know what value is there now. If the board is set up like the example, then adding a 33K resistor would increase the current limit by about 10%.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #11 (permalink)  
Old 12th February, 2004, 10:40 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Having considered things some more, the fact that you're getting some unstability when you set 1.875V *might* indicate that the transient overcurrent protection is beginning to kick in. That's not definate, as I haven't examined the board or any of the waveforms present at the regulator.

It's always much harder to do this by proxy!
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #12 (permalink)  
Old 12th February, 2004, 02:22 PM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

If I set somthing like the fsb to high, I normally get instability, error messages and crashing. However, if I set this voltage to 1.875v, it usually won't boot, but if it does, the voltage in windows still reads 1.84v, and the system will restart fairly quickly. At 1.9v the bios needs reseting.

At no point does my pc produce any errors in prime95, its simply a case of pushing any further results in this restarting.

Its quite fustrating as my whole reason for goin for this board was based on reviews telling me that overclocking wise, and voltage wise, I could set things alot higher than I actually can. I wouldn't have bought it if i'd known what I know now.

Pin 15 is linked to a resistor, which is in a similar place to the resistor you can remove in version 1.x of the board.

Picture of an 8rda+ 1.x (not same as mine) :
http://www.motherboardfaqs.com/Epox_...utoff/ov12.jpg

Pin 16 is linked to a seperate resistor on the board. At some point, I'll get hold of a good digital camera, and take a proper pic, maybe that'll help.


I'm not sure if any of the observations I make are signifigant or not, but I'll try my best to give you details, I really appreciate your help thanks
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #13 (permalink)  
Old 12th February, 2004, 04:10 PM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Forget about the v1.1 board, as it's circuitry is completely different, even if components are in similar positions. A picture might be helpful.
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #14 (permalink)  
Old 12th February, 2004, 04:23 PM
Gizmo's Avatar
Chief BBS Administrator
BassTeroids Champion, Global Player Champion, Aim & Fire Champion, Puzzle Maniax Champion, Othello Champion, Canyon Glider Champion, Unicycle Challenge Champion, YetiSports 9: Final Spit Champion, Zed Champion
 
Join Date: May 2003
Location: Webb City, Mo
Posts: 16,178
Send a message via ICQ to Gizmo Send a message via AIM to Gizmo Send a message via MSN to Gizmo Send a message via Yahoo to Gizmo Send a message via Skype™ to Gizmo

Can you perhaps read the numbers on top of the resistors attached to pins 15 and 16?
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #15 (permalink)  
Old 12th February, 2004, 04:38 PM
chrisbard's Avatar
Benchmarker
 
Join Date: March 2003
Location: Earth
Posts: 8,252
Send a message via Yahoo to chrisbard

Question

Hello guys,

I also have 2500 Barton on my 8RDA+ ver.1 and I still have my old DDR 2*256 at 333 not 400. My max FSB is now 175*12 and there without voltage modification it's stable. Above that it'll do nasty things. What can I do to go higher? I'll change my memory today with DDR 400.

Tx for help and excuse me for barging in.

Chris
__________________
I've heard that linux community came up with better implemented security in it's latest Linux Mint Gold version, it's actually preventing the user to log in, thus posing 0 risk in contamining the computer with malware! Well done to the open source community!

Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #16 (permalink)  
Old 12th February, 2004, 09:07 PM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

I've taken some high definition snaps, I'll get a magnifying glass n try n read off some numbers and post links to pictures on my uni webspace tommorow (as I've borrowed the camera from uni, and they don't give out the usb cable with it).

Chrisbard, I had the same problem (couldn't get past 180mhz), some of the 1.x boards are just like that (down to luck), and I think you have to vmod to get higher. I sold my ver 1.0 for £45 and got a 2.1 for £50 from komplett.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #17 (permalink)  
Old 12th February, 2004, 10:32 PM
MIP's Avatar
MIP MIP is offline
Member
 
Join Date: May 2003
Posts: 23

Before I get these pictures up, there are a few more questions I have:

a) Does the bios version have an impact on my particular voltage/current problem? If so, what would be the best bios to use?

b) I'm using memtest_86 to work out if the speed at which I can run my memory/fsb. I've currently upped it to 225mhz x 10.5 (2300mhz cpu, ram: 2.5, 3, 3, 11), and I get two or three errors in a pass (normally on a later cached test in the first 0.1 to 0.5mb). Is this acceptable, can any error correction the board/cpu/memory might have, cope with an error or two if there are very few of them? Because its in a cached test withing the first 0.1 to 0.5mb, does this mean the errors are occuring in the l2 cache?

edit: I've answered my own question to (b). 3D mark crashed with an error. And my guess is that as my machine will boot all the way to 230mhz fsb with the memory on 2.5, 3, 3, 11, without any errors in memtest bar the cached test from 0.1 to 0.5mb, then it is the l2 cache that is producing these errors. I only get errors in the rest of the memory after 230mhz. Look like my CPU and cache need more voltage
__________________

Last edited by MIP; 12th February, 2004 at 10:42 PM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #18 (permalink)  
Old 13th February, 2004, 01:13 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

In respect to a), the BIOS can set the output voltage from the regulator, as this is software controllable.

However, the current limits are set by a few resistors, and the over voltage protection is completely internal to the voltage regulator itself.

As no BIOS is advanced enough to get out a soldering iron and change components on the motherboard for you, I think you'll find that the answer is a no, it doesn't make any difference from the perspect of the voltage regulator.
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #19 (permalink)  
Old 13th February, 2004, 01:22 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Quote:
Originally Posted by MIP
b) I'm using memtest_86 to work out if the speed at which I can run my memory/fsb. I've currently upped it to 225mhz x 10.5 (2300mhz cpu, ram: 2.5, 3, 3, 11), and I get two or three errors in a pass (normally on a later cached test in the first 0.1 to 0.5mb). Is this acceptable, can any error correction the board/cpu/memory might have, cope with an error or two if there are very few of them? Because its in a cached test withing the first 0.1 to 0.5mb, does this mean the errors are occuring in the l2 cache?
Unless you have ECC RAM (and a board that supports ECC), there is NO error correction on the memory bus. That means that any memory errors are totally unacceptable.

Without knowing too much about how memtest86 works internally, if you're only getting errors on a cached operation, it'd suggest that there *might* be a problem with the cache or Northbridge. There's too many ifs and buts to really be sure about where the error might have been occuring. Which test(s) is(are) causing the failures?
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #20 (permalink)  
Old 13th February, 2004, 02:01 AM
Member
 
Join Date: January 2004
Posts: 20

Exclamation

Hy !
Interested in this thread too, because I have too a 8RDA+ 2.1 board, and the same problem here...
All is OK on this board except that if I set in BIOS a Vcore of 1.875 or higher. If I do that, P80Port shows some numbers (at most of 1 second), and then it switch to FF. To resolve -> INSERT pressed while pressing also RESET on the case.
And the interesting point:
If I use 8RDA Vcore utility, I can set 1.875 and 1.900 with no problem (and it's stable to overclock more the processor too).
PS: My board increases the voltage by 0.1v automatically (I don't know why). For example, I need to set 1.500v, and PC HEALTH indicates 1.590-1.600v. So for the 1.900v set by 8RDA Vcore utility I have almost 2.000v. If I go one step further on setting the voltage upper, the screen goes black and computer freezes.
Another interesting problem:
Well, I can set 1.900v with 8RDA Vcore utility, but when I restart windows, computer will not boot... It does the same like setting 1.875v in BIOS. I need to use the above routine to make it work.

So I will try to do this "mod" too, as I want 2.200v to work. And I'm ready with a good cooler for that. Hopping of a good camera from guys with 8RDA+ 2.1 board.
Thanx so far for help ppls.

PS [offtopic]: Did you increase the Vdd also when the FSB was over 220MHz ? If you have 2 sticks of RAM, try with one of them at a time. Or try with a different RAM(s) from some friends. Cached test are the most important, as you will run the computer with cache enabled. And doesn't mater where the error is (begining...end). At mine the errors is located to almost 490M in memory. Those are the weaky "bits" from memory, and you cannot do nothing, except to increase the RAM voltage. Definetly you shouldn't get errors for the things to be stable.
__________________

Last edited by dizzy; 13th February, 2004 at 02:13 AM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
Reply



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
NF7-S voltage issue. Tech head question. toaster AMD Motherboards & CPUs 8 23rd May, 2006 11:25 AM
Issue with 8rda+ :-/ Clipse EPoX MotherBoards 2 30th May, 2003 06:22 PM
8RDA+ & 3DMARK2k1 Serious Issue neosho EPoX MotherBoards 28 22nd April, 2003 08:34 PM
8RDA+ power off issue Yitch EPoX MotherBoards 6 7th April, 2003 03:00 PM
8RDA+ voltage issue Forgiven EPoX MotherBoards 9 4th February, 2003 04:33 PM


All times are GMT +1. The time now is 02:30 PM.


Copyright ©2001 - 2010, AOA Forums
Don't Click Here Don't Click Here Either

Search Engine Friendly URLs by vBSEO 3.3.0