6800le 'Black Screen' - The Complete Guide: Identification, modification and reason

Printable View

Show 100 post(s) from this thread on one page

03-25-2006, 05:33 AM
scorpion77

Some more gami...eh testing ;)

SikaRippa you asked me to post an update if I do get BSS with my new PSU ... I'm sorry haven't seen any, but it might be nice to post my experiences anyway :P

After playing ut2k4 linux amd64 and testing for over a week I think it's safe to say my BSS problems are really gone. Currently I'm running at 400/850 12x6 with the memory timings relaxed a bit (10-21-10-someting) and combined with a regular overclock (amd3200 754 2Ghz@2250 mhz) this is good for 4400+ 3dmarks in 3dmark05. All I have seen so far is:

- 2 reboots while running ati tools artifact test for about 5 mins (non reproducable after I turned off winxp's auto reboot function to see what was causing this)
- some arti' s while playing ut2k4 (actually only once and it could have been just software related)
- one regular lockup while playing ut2k4 (image freeze, pressing reset got my system working right away, unlike with BSS)

If I understand correctly especially the last one is actually a good thing BSS wise?
03-25-2006, 06:10 AM
***Deimos***

Quote:

Originally Posted by dkr09

Thank you for your reply, SikaRippa.
Thanks read my post and give an advice for me in spite of my poor english.

My card seems to made that site,
and almost same that card linked without
cooler and some parts. I think it is OEM products.
The value of green capacitor is 40 470 6E.
I bought this card a week ago and the cards get BS from when I bought.

I tried plug in and out the card several time for test. Changed VGA and
motherboard and power and so on. Now the R47 parts and 3 chips are
not so hot with additional 12cm cooling fan with composed first PC parts.
I don't know why the situation was happened. I don't have tester so
I checked core voltage with Rivatunner monitoring program, but It says
core vid is 2.0v. It is incorrect value, isn't it? Another report tab,
the voltage was 1.2v on the BIOS info.

Anyway all is good.
My card driving with 16/6 pipe and 400/800 clocks without any problem.
And BS situation disappeard completely.
I am very happy. The cap mod was very helpful for me.
Thanks and Good luck SikaRippa!

Sounds great man!
What capacitor did you use for the cap mod?

Vid says 2, because Vid is an "ID". IT is voltage #2 on the list. IT IS NOT A MEASUREMENT. The actual voltage is whatever corresponds to voltage #2 on the list. You can check this using niBitor... is should be 1.2/1.4/1.5.. probably 1.4V

To measure the real voltage you need multimeter. The place to measure is shown on SikaRippa's pictures.

What did you do exactly to make the 3 voltage regulators on the back of the card not make so much heat? Did you just mount a fan blowing at the back?

If you want 900+ on memory try increasing tRP=6 and tRCDRD=6. Your timings should be something like (timingset 6):
060A140A
05010306
00240306

Anyone know if there is something like "VPU recover" and where to turn it off..
03-25-2006, 01:39 PM
SikaRippa

Quote:

Originally Posted by ***Deimos***

auto-detect with 12/5 and 16/6 is about the same. top 3dmark speed also same. Likewise, it was at similar level before I installed NV5. Although increasing Vmem increased top memory speed before artifacting in rthdribl or atitool, autodetect still same or even lower. I think there may be some damage on the card. When I made a mess with some AS5 one time, I used contact cleaner to clean it up. I sprayed it everywhere, and made puddles of AS5 that even went under the chips (GPU, mem etc).. I sprayed a lot, until it was nearly clear... and it vaporizes fast. Dont know how big issue AS5 residue could be.

I will try the capacitor mod again, because last time the legs to the capacitor were too long. I will post some capacitors I have. I'm not sure which has lowest ESR.. I think they are all hybrid, 6.3V-10V and 1000uF-6000uF...

Contact cleaner might cause some problems.. I have read (datasheet?) that electrolytic capacitors are quite sensitive to all chemicals, because the vent at to bottom is not so good. For instance during manufacturing a wrong cleaner (to remove flux) can cause many problems like deterioration of parameters.

You had long wires, possibly poor/cold contacts ? Those both destroy the effect of modification, since the series resistance increases. An example. The original "green capacitors", CV-EX 470 uF have (should have as new!!) ESR of 0.03 ohm each -> 0.015 in parallel. A copper wire with diameter of 0.3 mm and lenght of 6 cm has also resistance of ~ 0.015.....

What capacitor you used, low-ESR ? If you are not sure about specs, post all codes in the capacitor.
03-25-2006, 01:50 PM
SikaRippa

Quote:

Originally Posted by dkr09

Thank you for your reply, SikaRippa.
Thanks read my post and give an advice for me in spite of my poor english.

My card seems to made that site,
and almost same that card linked without
cooler and some parts. I think it is OEM products.
The value of green capacitor is 40 470 6E.
I bought this card a week ago and the cards get BS from when I bought.

I tried plug in and out the card several time for test. Changed VGA and
motherboard and power and so on. Now the R47 parts and 3 chips are
not so hot with additional 12cm cooling fan with composed first PC parts.
I don't know why the situation was happened. I don't have tester so
I checked core voltage with Rivatunner monitoring program, but It says
core vid is 2.0v. It is incorrect value, isn't it? Another report tab,
the voltage was 1.2v on the BIOS info.

Anyway all is good.
My card driving with 16/6 pipe and 400/800 clocks without any problem.
And BS situation disappeard completely.
I am very happy. The cap mod was very helpful for me.
Thanks and Good luck SikaRippa!

Thank for info, and nice you get your card working/cool down :up:

Having BS with a new brand card it not nice, indeed :mad: This means AXLE cards are as-bad-as the first GW and MSI-cards (having "4 3 40 6E" capacitors). Some of those cards have been quite hard to modify by adding capacitor and/or the BS has been appeared again quite soon :( . Therefore, the only way (with those cards) has been changing the capacitors. But, let's hope AXLE will be working better.

--

As ***Deimos*** already said, to only 100 % method to find out voltages is to measure :D
03-25-2006, 02:03 PM
SikaRippa

Quote:

Originally Posted by scorpion77

SikaRippa you asked me to post an update if I do get BSS with my new PSU ... I'm sorry haven't seen any, but it might be nice to post my experiences anyway :P

After playing ut2k4 linux amd64 and testing for over a week I think it's safe to say my BSS problems are really gone. Currently I'm running at 400/850 12x6 with the memory timings relaxed a bit (10-21-10-someting) and combined with a regular overclock (amd3200 754 2Ghz@2250 mhz) this is good for 4400+ 3dmarks in 3dmark05. All I have seen so far is:

- 2 reboots while running ati tools artifact test for about 5 mins (non reproducable after I turned off winxp's auto reboot function to see what was causing this)
- some arti' s while playing ut2k4 (actually only once and it could have been just software related)
- one regular lockup while playing ut2k4 (image freeze, pressing reset got my system working right away, unlike with BSS)

If I understand correctly especially the last one is actually a good thing BSS wise?

Thanks for update :) Nobody likes to see BS's so it is allways nice to get reports "anything to report". I wonder if (hopefully never) you will get BS again it takes time (months). Moreover, you have XFX6800XT with "purple capacitors" (?) which should deteriorate much slower than *normal* 6800le-cards.

Yep, the case you decribed is not BS i.e. it was **normal** shut down.
03-26-2006, 04:07 AM
dkr09

Quote:

Originally Posted by ***Deimos***

Sounds great man!
What capacitor did you use for the cap mod?

Vid says 2, because Vid is an "ID". IT is voltage #2 on the list. IT IS NOT A MEASUREMENT. The actual voltage is whatever corresponds to voltage #2 on the list. You can check this using niBitor... is should be 1.2/1.4/1.5.. probably 1.4V

To measure the real voltage you need multimeter. The place to measure is shown on SikaRippa's pictures.

What did you do exactly to make the 3 voltage regulators on the back of the card not make so much heat? Did you just mount a fan blowing at the back?

If you want 900+ on memory try increasing tRP=6 and tRCDRD=6. Your timings should be something like (timingset 6):
060A140A
05010306
00240306

Anyone know if there is something like "VPU recover" and where to turn it off..

Hi! Thank you for adviced me. I checked the value core VID 2 and the value
was 1.4V. I have understand correctly. I modded voltage 1.1 in 2D mode and
1.3 in 3D mode with Nivitor. That was more good about the heat of 3chips
and R47 part. And tried with the memory setting but couldn't over 900Mhz. It was under stable with 850Mhz. It is enough for me. Thanks!

I used this capacitors. 10V 2200uf caps. http://www.samwhatech.co.kr/shop/ima...7110728_WB.pdf

And the fan... yes just blow 12CM fan 1800rpm from side of the card.
no heatsink or anything.
now it seemed no need fan in 2D mode and light 3d game.
When playing 3dmark or ATItools artifacts test or heave 3dgame,
the chips quite hot but not to concern to broke I think. I can thouch
upto 10sec. The past I cant touch only 1sec without any overclocking.

Good night Deimos!
03-26-2006, 07:07 PM
***Deimos***

Quote:

Originally Posted by dkr09

Hi! Thank you for adviced me. I checked the value core VID 2 and the value
was 1.4V. I have understand correctly. I modded voltage 1.1 in 2D mode and
1.3 in 3D mode with Nivitor. That was more good about the heat of 3chips
and R47 part. And tried with the memory setting but couldn't over 900Mhz. It was under stable with 850Mhz. It is enough for me. Thanks!

I used this capacitors. 10V 2200uf caps. http://www.samwhatech.co.kr/shop/ima...7110728_WB.pdf

And the fan... yes just blow 12CM fan 1800rpm from side of the card.
no heatsink or anything.
now it seemed no need fan in 2D mode and light 3d game.
When playing 3dmark or ATItools artifacts test or heave 3dgame,
the chips quite hot but not to concern to broke I think. I can thouch
upto 10sec. The past I cant touch only 1sec without any overclocking.

Good night Deimos!

I dont understand.
Are you saying that increasing the Vgpu DECREASED the heat output from the voltage regulators are the back of the card?

For the memory, try looser timings. You can use NIBITOR. Auto-detect the timingset, and then you can set a new timingset temporarily (in windows, ie dont need to flash).

If you want 900+ on memory try increasing tRP=6 and tRCDRD=6. Your timings should be something like (timingset 6):
060A140A
05010306
00240306

Furthermore, disable internal clock test in rivatuner. Go to powerusers, global overclocking.. somewhere there... just enter 1 in that field. For me, the difference with these two was from 760Mhz -> 900 (disable internal)->960(loose timings)
03-26-2006, 07:25 PM
***Deimos***

SilkaRippa,

The capacitors I tried using last time:

Tayeh
2700uF
6.3V
LE 105C A117

TEAPO
3900uF
10V
SM105C
09/00

nichican
4700u
HD 6.3V
H0045

I also had a nice fat one somewhere, but I cant seem to be able to find one.. (the last two are de-soldered from old Athlon motherboard btw).

EDIT: although my case does not involve my LCD going blank/black, there is the suddent tearing/flicker followed by slower performance than usual. Do you think its possible this is that voltage regulator stuck in 1-3 clock cycle where there is not sufficient voltage.. ie the C143 capacitor is ruined, but only slightly.
03-26-2006, 08:23 PM
dkr09

Quote:

Originally Posted by ***Deimos***

I dont understand.
Are you saying that increasing the Vgpu DECREASED the heat output from the voltage regulators are the back of the card?

For the memory, try looser timings. You can use NIBITOR. Auto-detect the timingset, and then you can set a new timingset temporarily (in windows, ie dont need to flash).

If you want 900+ on memory try increasing tRP=6 and tRCDRD=6. Your timings should be something like (timingset 6):
060A140A
05010306
00240306

Furthermore, disable internal clock test in rivatuner. Go to powerusers, global overclocking.. somewhere there... just enter 1 in that field. For me, the difference with these two was from 760Mhz -> 900 (disable internal)->960(loose timings)

No,no. My card's Vgpu value was 1.4v at first. When 1.4v, it doesn't so
heats up too. yes the regulators back of the card were pretty hot.
But at the first, my PC was in abnormal. and in the state,
the voltage regulator chips and R47 part were very very hot.
I coudln't touch the chips at all. I don't know why these were occures.
But while installing the VGA card many time repeatedly,
the situation was disappeared.
And now I set the Vgpu value to 1.1v in 2Dmode and 1.3v
in 3Dmode with Nibitor and the heat little more decreased.
No need additional fan blowing in 2Dmode.

And and I tried tRP and tRCDRD values 5 to 6 with your advice.
I couldn't pass internal cheking over 860 in the past.
I disabled internal cheking and tested again with your new advice.
now the memory clock goes up endlessly!
Rivatuner is very good app. It has many wanderful functions I didnt know.
It seeems it is stable under 900Mhz or more.
I didn't test many time. I set 400/800 safely.

Sorry for making you confusing. I can't speak English well.
I wish my message could understanding you well...
Have a good day.
03-27-2006, 07:13 AM
SikaRippa

Quote:

Originally Posted by dkr09

Hi! Thank you for adviced me. I checked the value core VID 2 and the value
was 1.4V. I have understand correctly. I modded voltage 1.1 in 2D mode and
1.3 in 3D mode with Nivitor. That was more good about the heat of 3chips
and R47 part. And tried with the memory setting but couldn't over 900Mhz. It was under stable with 850Mhz. It is enough for me. Thanks!

I used this capacitors. 10V 2200uf caps. http://www.samwhatech.co.kr/shop/ima...7110728_WB.pdf

And the fan... yes just blow 12CM fan 1800rpm from side of the card.
no heatsink or anything.
now it seemed no need fan in 2D mode and light 3d game.
When playing 3dmark or ATItools artifacts test or heave 3dgame,
the chips quite hot but not to concern to broke I think. I can thouch
upto 10sec. The past I cant touch only 1sec without any overclocking.

Good night Deimos!

OK .. the high temperature was due to (too) high Vcore (?) There was a thread in this forum earlier, where that coil was dropped off (!!!!) due to excess temperature after Vmodding. It might be safe to not use higher than say 1.3 V EXCEPT with good cooling.

--

The capacitor you have used has ESR of 0.025 ohm so it should be VERY GOOD for this purpose :up:
03-27-2006, 07:27 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

SilkaRippa,

The capacitors I tried using last time:

Tayeh
2700uF
6.3V
LE 105C A117

TEAPO
3900uF
10V
SM105C
09/00

nichican
4700u
HD 6.3V
H0045

I also had a nice fat one somewhere, but I cant seem to be able to find one.. (the last two are de-soldered from old Athlon motherboard btw).

Nichicon is manufacturing good capacitors, but:

Tayeh & TEAPO ARE KNOWN TO MAKE VERY BAD CAPACITORS

.. those can be even dangerous (!!!), see following page (e.g. post #3):

http://www.badcaps.net/forum/showthread.php?t=388

I have myself changed capacitors to two MOBO's and half of those were TEAPO's :rolleyes:

So, with those capacitors you propably can not get any performance increase at all, maybe decrease...

Quote:

EDIT: although my case does not involve my LCD going blank/black, there is the suddent tearing/flicker followed by slower performance than usual. Do you think its possible this is that voltage regulator stuck in 1-3 clock cycle where there is not sufficient voltage.. ie the C143 capacitor is ruined, but only slightly.

I think that is not possible, because stucking of voltage regulator causes BS. The mechanism is explained in datasheet (and post #1) in detailed, but in short: if regulator is not able to raise output voltage within 1-2 clock pulses it is considered a short condition. All outputs are disables -> card's power off -> BS
03-28-2006, 11:34 PM
Maz_D

I have a aopen 6800le, with 45 470 6E capcitors. The card has been working fine for a while, then suddenly, the card started messing up in games after a few minutes or longer depending on what game.

Anyway i think i might have the same problem, so i read this thread, and discovered it might be something to do with the capacitors. Because I can't find anyway to get these capacitors in the UK (maybe someone can tell me where to get them) i tried something else. I put a fan leaning ontop of the capacitors and this reduced the crashing in all games except for chronicles of riddick (very 3d intensive).

I was wondering could the problem with the capacitors be heat related? Can you maybe put a heatsink on them, or could that destroy the card?
03-29-2006, 07:33 AM
SikaRippa

Quote:

Originally Posted by Maz_D

I have a aopen 6800le, with 45 470 6E capcitors. The card has been working fine for a while, then suddenly, the card started messing up in games after a few minutes or longer depending on what game.

Anyway i think i might have the same problem, so i read this thread, and discovered it might be something to do with the capacitors. Because I can't find anyway to get these capacitors in the UK (maybe someone can tell me where to get them) i tried something else. I put a fan leaning ontop of the capacitors and this reduced the crashing in all games except for chronicles of riddick (very 3d intensive).

I was wondering could the problem with the capacitors be heat related? Can you maybe put a heatsink on them, or could that destroy the card?

Hi,

Thank you for proving for new info :) There have been some (unaccurate) reports that Aopen-cards were BS-affected, but not yet 100% confirmed (?).

You are sure you have BS-problems:

- A hard reset (=power switch off/on) required to get PC working ?
- 'Standard BS-programs' like SkiAlpin2005 demo or HL2 or F.E.A.R etc. cause easily problems/BS ?
- Setting 'vsync=on' decreases BS-propability ?

--

You are talking about replacement capacitors ? I suggest to test with extra (low-ESR !!!) capacitor at first, because it is soooo much easy to obtain/install.

--

Heat & capacitors is a complex issue. Higher temperature = shorter life, so in general the better cooling the better performance => fan & heatsinks are good idea. BUT when testing BS-feature the high instantenous temperature DECREASES ESR of electrolytic capacitor => BS-propability decreases !! (I have demostrated that myself by freezing capacitors, which caused BS ASAP).
03-30-2006, 02:45 AM
Maz_D

Hi, it seems like the bs, but instead of a black screen, it artifacts heavily then crashes. Maybe I'm wrong. I did the enhanced capacitor mod and that helped a lot, but it still gave me artifacts, but this took a lot longer than usual.

Could this just be overheating? Please help

ps. the artifacts just make the screen get covered with lots of rectangles, but only happen in very 3d intensive games
03-30-2006, 02:56 AM
Maz_D

Is it safe to put a heatsink on a capacitor?
03-30-2006, 11:44 AM
***Deimos***

use nibitor, typically timingset6 (auto-select).. test the timings first, and try overclocking, then modify your bios, save, and flash using nvflash.

Code:

tRC tRFC tRAS tRP tRCDRD tRCDWR tRRD original16 17 10 5 5 3 4 loose 10 20 10 6 6 3 4 hynix datasheet HY5DU283222AF 2.8ns 16 17 10 5 5 2 4 2.5ns 18 21 12 6 6 3 4 2.2ns 21 24 14 7 7 3 4 2.0ns 23 26 16 7 7 4 4

tRCDRD=6 and tRP=6 make the biggest improvement ;) even if you have 2.8ns Hynix like me, you should hit 900Mhz easy. BTW, I dont advise doing Vmem mod.. only 10-20Mhz improvment.
03-30-2006, 12:05 PM
SikaRippa

Quote:

Originally Posted by Maz_D

Hi, it seems like the bs, but instead of a black screen, it artifacts heavily then crashes. Maybe I'm wrong. I did the enhanced capacitor mod and that helped a lot, but it still gave me artifacts, but this took a lot longer than usual.

Could this just be overheating? Please help

ps. the artifacts just make the screen get covered with lots of rectangles, but only happen in very 3d intensive games

In my opinion, in general BS and artifacts are not closely releated to each other. Artifacts etc. visual *effects* on screen are caused many/most cases due to:

- Too high clock speed memory and/or core
- After opening locked units which does not work properly
- Too high temperature due unproper cooling e.g heat sink loosen

If card is in 'a stock condition' (clocks/voltages/pipes/cooling OK), artifacts might indicate card is broken OR some kind of software problem (uncompatible drivers for certain game etc.)

BS is appeared due to fail of the switching regulator of the card (caused usually ruined/weak capacitors in the case of BS-affected cards).

--

Is your card in stock condition ? What are temperatures ? Cooler sited properly ?
Are the capacitors running very hot (test carefully with finger tip) ?
03-30-2006, 01:43 PM
Maz_D

Its all at stock, it doesn't have a temperature monitor. But I don't think its overheating, it doesn't even seem hot. I've never seen anything happen like this before, it was working for a while then all of a sudden it starts producing artifacts in certain games, after a few minutes of playing, then crashes straight after.

I think I'll try to get a replacement if i can.
03-31-2006, 04:33 AM
SikaRippa

Quote:

Originally Posted by Maz_D

Its all at stock, it doesn't have a temperature monitor. But I don't think its overheating, it doesn't even seem hot. I've never seen anything happen like this before, it was working for a while then all of a sudden it starts producing artifacts in certain games, after a few minutes of playing, then crashes straight after.

I think I'll try to get a replacement if i can.

You could use e.g. rivatuner to test temperatures. I have heard stories about loosen heatsink of core or too few thermal paste causing several different problems.

I suggest to test with finger those capacitors: if those are cool (=cooler/equal than other capacitors) the ESR is still quite low level and replacement does not help (much).

The reason why I think there are some other promlems in addition to *normal* BS, is fact that I have never seen artifacts just before BS (and not heard such reports).
04-13-2006, 12:49 PM
SikaRippa

Final Update 2.2

Version 2.2 of Document posted !!

Finally, after precicely 1 year after the first version, I posted a hopefully final version of the BS-document, see post #1. Please, take a look at least the measurements chapters:

- Oscillosscope fotos of AC-current/voltages rippels measured directly on 6800le-card
- Capacitor ESR -measurement results shown and setup fully explained (last Chapter)

=> Finally I was able to confirm the CV-EX capcitors are really ruined, since ESR-values are completely out of range .. or just 'BS'

------

This has been a hard project, but I think it has been quite usefull at least for me. I have learn my new things, hopefully helped some (many?) people and got many new virtual friends around the world: Finland, Sweden, Norway, Denmark, UK, France, Netherlands, Switzerland, Poland, Estonia, Russia, Spain, Italy, Croatia, Greece, Israel, USA, Canada, Australia (I hope did not miss too many..)

Thank you for all for solving out the BS-problem !!!
04-13-2006, 08:32 PM
***Deimos***

http://www.xtremesystems.org/forums/...ad.php?t=95954

read the part about replacing heatsink 18-20 times and the screw... really makes you wonder.

After failing to get any substantial gains with Vmem mod (again), I will give capacitor a try again. Please recommend specific values for Vmem and also Vgpu (read my measurements)..

also interesting... about half-year ago, when I first contacted you, my Vgpu (with same BIOS) was 1.33V... now 1.32V.
04-14-2006, 01:14 AM
sunn

Quote:

Originally Posted by sunn

Hello, i thought i'd update you on my LE, working nearly 4 months BS free :woot:

I haven’t posted in ages, but the same as above. My AOpen LE is still working BS free, that’s 11months now, a bigger :woot:

You deffo helped me with my first Gainward as I remember googling for '6800LE crashes' and coming up with nothing useful until I was on the Gainward forum and saw a link to this thread, so finally found the problem and confirmed it was a faulty card and was able to RMA, so BIG thanks and applause
:clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap:
04-14-2006, 01:17 PM
SikaRippa

Quote:

Originally Posted by ***Deimos***

http://www.xtremesystems.org/forums/...ad.php?t=95954

read the part about replacing heatsink 18-20 times and the screw... really makes you wonder.

After failing to get any substantial gains with Vmem mod (again), I will give capacitor a try again. Please recommend specific values for Vmem and also Vgpu (read my measurements)..

also interesting... about half-year ago, when I first contacted you, my Vgpu (with same BIOS) was 1.33V... now 1.32V.

I took a look that thread: you have indeed encountered several problems and hard testing..

Yep, installing/uninstalling NV5 20 times does not sound very healthy for the card. Since you got it working better (=perfectly?) at last, I believe there can be problem at contacts e.g. some memory chip is slightly 'loosen'. The bad condition of BGA balls beneath the chip can not be seen, and by having screws in little tighter could make them contact again. If that is true, I suggest not to uninstall NV without GOOD reason and keep screws in current positions precisely.

If I remember correctly, you used bad capacitors (or having unknow quality) before ? It is clear that the better (=lower ESR) added capacitors are the better results can be expected. The minimizing the length of the leads is also very critical, as can be seen those ESR-measurement results I posted.

About voltages. I saw (in that thread) you have large fluctuations in Vcore. That could indicate ruined Vcore capacitors, but also might be due to too high Vcore for the regulator of the card. Could you test (maybe hard with NV5) the temperature of the Vcore & Vmem capacitors e.g. with finger tip ? Warm/hot ?? How about IC's back side of the card ? There are some measured temperature data also in the document to get idea temp values with Vcore 1.1 and 1.2 V. I do not suggest to increase Vcore from 1.3 V, except you'll take care of cooling of those components. There were posted some tested Vmem values that thread I posted about Vmem modification. I wonder that values below 2.95 V should be OK, if the temperatures do not increase too much. The absolute maximum for Vmem is 3.3 V, but I do not recommended that.

The Vcore fluctuations could be analyzed in the same methods I did in the case of Vmem. Did you take look at the oscilloscope fotos I put in the end of post #1 ? Not too clean power supplies !!
04-14-2006, 01:26 PM
SikaRippa

Quote:

Originally Posted by sunn

I haven’t posted in ages, but the same as above. My AOpen LE is still working BS free, that’s 11months now, a bigger :woot:

You deffo helped me with my first Gainward as I remember googling for '6800LE crashes' and coming up with nothing useful until I was on the Gainward forum and saw a link to this thread, so finally found the problem and confirmed it was a faulty card and was able to RMA, so BIG thanks and applause
:clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap:

Nice it is still working :up:

Also my GW 'version 1b' still working after ~ year, but there are some reports showing that model will eventually ruin :(

--

BTW In the first place I had an idea to post BS-document to GW's forum
:slobber: :slobber: Hopefylly, I noticed quite soon the *sensored* attitude of moderators and made a wise decision: document to a GOOD forum and only an advertisement to GW's forum. The thread with the 'advertisement' was deleted quite soon by moderators, and also two similar after some months ...
04-14-2006, 07:21 PM
***Deimos***

No it is not working perfectly at all. Just this morning it was back in "bad" mode. Always refering to 425/950.

mode A:
1-4 fps.. really horrible articacts.. cant even tell 3DMark2005 running.. sometimes blank screen.. happens typically when I tilt the heatsink off altogether.

mode B:
20-27fps.. artifacts galore, usually techno-color, or everything black, or weird stripe effect on textures.

mode C:
~27fps.. but no artifacts at all. It all seems to be running perfectly fine. Unforunately, at 425/950, get about 4200-4300 3DMark2005 points.. only about 800 less than expected :\ When in this mode, it stays in this mode, which is different than mode D.
I've gotten this mode when torqueing the card around.. trying to make sure there is good contact in AGP slot.

mode D:
The mysterious 33-34fps, running perfectly without artifacts, but sometimes, flash, and brief stutter, and then continues in modeC. The places where this happens is typically where the voltage dropped to 1.308V (lowest recorded). Sometimes it freezes altogher, and once I had internal driver error popup.

typical mode D problem:
It could be running perfectly fine at full framerate through all of game test 1, and then in game test 2, like where the little green light goes into the log, flash, stutter, and continue, but at substantially lower fps. No artifacts of any kind. When it begins game test 3, it will be back at full fps. But, ofcourse it can sometimes throttle in that game test 2 (captain looks overboard, or monster opens mouth).

Normally I expect 3DMark2005 run to give the following results:
GT1 20.4fps
GT2 15.0fps
GT3 27.7fps
all +/- 0.1fps
TOTAL = ~5050+ 3dmarks

when it is throttling, it will be something like (20.4, 13.2, 27.8 - GT2 only), (20.5, 13.1, 26.8 - throttled in GT2, and at end of GT3 so didnt affect its score much) or even (16.9, 13.2, 22.5 - running "27fps mode" in all 3)

I hope this explains a bit of whats going on.
04-14-2006, 07:31 PM
***Deimos***

http://service.futuremark.com/orb/pr...jectId=1956543
04-16-2006, 07:15 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

http://service.futuremark.com/orb/pr...jectId=1956543

The project id is not valid - please check the URL.

http://service.futuremark.com/compare?3dm05=1856608 :woot:
04-16-2006, 07:18 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

Normally I expect 3DMark2005 run to give the following results:
GT1 20.4fps
GT2 15.0fps
GT3 27.7fps
all +/- 0.1fps
TOTAL = ~5050+ 3dmarks

when it is throttling, it will be something like (20.4, 13.2, 27.8 - GT2 only), (20.5, 13.1, 26.8 - throttled in GT2, and at end of GT3 so didnt affect its score much) or even (16.9, 13.2, 22.5 - running "27fps mode" in all 3)

I hope this explains a bit of whats going on.

Compared to my 5430 with 19.5 - 17.5 - 30.2 50xx is low :confused:
04-16-2006, 04:17 PM
***Deimos***

http://service.futuremark.com/compare?3dm05=1914193

there, that should work.
5118
20.6
15.0
27.8

Well, first of all, your card is clocked at 450, not 425, so naturally you have higher scores. You also have much higher clocked memory and CPU, which should mean higher GT1 scores.. but you dont. Did you use LOD tweak, and set driver image setting to "high performance"? Like the description for my project says, I did neither.. exactly as-is when you install driver.. default = quality setting.

Jimmor sent me a few BIOS files, and my problem seems to have lessened. I ran 5+ 3DMark2005 in quick succession at 410/900 without force performance level, and it didn't drop to 3D throttle except in between runs. Chaos theory is the ultimate test, and if it works well there too, then I'll be really happy. BUT, this card is tempermental, and several times I had thought all problems were gone only to return next morning.. I wouldn be surprised if the problem was some cold solder joints.
04-16-2006, 11:22 PM
Marvin_The_Martian

Run was with lod.. idk I always seemed to have low gt1 and high gt2 gt3, each run. And btw I thought your card was also a le, but it's an 6800 vanilla. Vanilla's always score higher ( well usually ) with identical clocks compared to le's IIRC.

I got caps sikkarippa send me but I got scared off doing the cap mod before I have another good ( read beter ) vc which will be later today or tomorrow :)
04-17-2006, 11:46 AM
***Deimos***

If you still have the capacitors, you can send them to me. Or at least let me know which brand, model, and capacitance value.

I tried doing the capacitor mods before, but according to SilkaRippa, the caps I bought are P O S (no surprise considering how cheap they were)
04-17-2006, 01:50 PM
vortex240

Guys I need help Please!

My BFG 6800 vanilla just died, it worked fine for 2 years...I lapped the heatsink when I first got it, also the screwes from the heatsink show sings of wear(screwing on and off)

So I cant send it in like that...is anyone willing to help me?

Lend me your heatsink or sell it to me if you want, I'll pay for shipping if you want to lend it to me.

thanks a lot

Wojtek
04-17-2006, 03:50 PM
***Deimos***

SilkaRippa!
I have a very important question regarding my 6800 problem. Do you know how to increase the frequency of the switching power regulators on a 6800?

EDIT: also, do you get high pitch sound when running coolbits auto-detect. Or in certain games or parts of games? It sounds like its pulsing. So I suspect thats the voltage regulators, with improperly tuned capacitors. If I increase the frequency it shouldn't be audible, and might improve the energy supply from power regulators. What do you think?
04-17-2006, 04:01 PM
***Deimos***

Quote:

Originally Posted by MartianInvader

Run was with lod.. idk I always seemed to have low gt1 and high gt2 gt3, each run. And btw I thought your card was also a le, but it's an 6800 vanilla. Vanilla's always score higher ( well usually ) with identical clocks compared to le's IIRC.

I got caps sikkarippa send me but I got scared off doing the cap mod before I have another good ( read beter ) vc which will be later today or tomorrow :)

@433/936 (using 1.4V in BIOS), driver=high performance, LOD=+1.8
5505
21.7
17.2
28.6

compare roughly to:

425/950
5100
20.5
15.0
27.6

The LOD and high-performance help the most in GT2. Part of the increase with GT1 is due to higher CPU and RAM speed used for 5505.

Definetly try lower clock speeds, since you might get higher scores. 90% of "artifacts"/errors are invisible. Check in rivatuner hardware monitor that its not throttling.. other than that, its probably just matter of different system config. ex I got BH5 running at 2-2-2-5.
04-17-2006, 04:30 PM
***Deimos***

Although I dont think I have the black screen problem described here:

"A very interesting aspect about temperature effect was reported in this same thread (see posts #3, #6, #7, Thank you ZeroHero !). It was found that by cooling the ambient of the card ~ 10 degrees causes immediately BS under suitable conditions. This phenomena can be caused by many factors e.g. cooling of the switching transistors. However, the on-resistance of the transistors reduces when temperature is lower and the operation of the switching power should be better ! On the other hand the ESR-value of the aluminium electrolytic capacitors INCREASES when temperature REDUCES, which LOWERS the performance of the switching power supply (See another document about the aluminium electrolytic capacitors if you are very interested [10].). In order to make this even more complex the higher resistance (ESR) causes higher voltage drop over capacitor and in the same time the power 'burn' in the resistor increases, which increases the temperature (would be nice to make electrical model for this ). "

this would explain why several times on cold mornings, high overclock of GPU and MEM wouldn't work.
04-18-2006, 01:49 AM
SikaRippa

***Deimos***, I had a vacation (=no www) for some days, I'll try to figure out your questions...

Cases A-D corresponding your card. I think the most *features* can be explained as too high temperature in certain condition (like in some game test) which causes throttling. I did not get whether there were any differences in the setup except tighter screws ? I mean you get those cases quite random manner (?). Maybe temperature variation gives the random effect, since you said later you had more/different problems in some cases having a cold card.

I read an article about NV5 installation some months ago. There was said that the copper pads for memories and core in the NV5 are NOT allways in the same height. Therefore, NV5 copper pads should be checked beforehand and possibly grind/polish using very fine sand paper etc. If the height difference of the pads is too high, some memories (core) is under higher mechanical stress, which can cause the contact of the BGA ball loosen. Did you have any such problems (cases A-D) before installing/uninstalling NV5 several times ? Do you have any problems with little bit lower Vcore/Vmem ?

---

About capacitors. 'POS'-capacitors do not help too much, so it is good idea to test with known good capacitor. Unfortunately I do not have any proper replacement capacitors. I had some Chemi-con PXA-series, which have been successfully used by several users. Another good candidate is Sanyo SVP. Those capacitors are hard/impossible to find from 'local electronic shops' , and only solid solutions are 'Digikey' or 'Farnell'.

http://us.digikey.com/scripts/dksear...189480&Site=US
http://uk.farnell.com/jsp/endecaSear...=9188770&N=401

.. drawback is high unit price due to postage & package cost.

---

About noises. I have briefly considered that issue in the document. There are some explanations and results (ripple current, noisy audio recordnings). In short, noises are due to rapid changes in current consumption of the card (memory& core). These current spikes either 'vibrates' some components in the cards (e.g. coils) or adds noise to normal audio devices (e.g. via molex-connector of the CD-player). The noise vs. switching regultor freuency is complex issue: it COULD have some effect. I suppose the effect is not huge because the frequency of the ripple current (=noise) is in audio band ~ 100 - 10000 Hz and the switching frequency is 300 kHz at stock.

---

About changing switching frequency of the regulator. Yes, that can be altered by a resistor. I considered also to play with that to see effect to BS, but have not (yet) tested. The idea how to alter it can be seen here ('Oscillator, pages #14-15 ):

http://www.intersil.com/data/fn/fn9134.pdf

By adding a resistor between FS_SYNC (pin #13) and GND the frequency can be increased up to 1MHz. However, I have not tested this and suggest to be very carefully. Higher switching frequency does cause extra problems..
04-18-2006, 09:16 AM
***Deimos***

Thanks for the link SilkaRippa!

I have switched power supply (again).

Code:

410/900 min max okia 450W atitool 11.92 11.94 oscilating wave 4.94 4.96 oscilating wave GT1 11.87 11.91 little impulses 4.98 5.02 little impulses enermax365W atitool 12.25 12.27 oscilating waves 5.08 5.1 little impulses GT1 12.21 12.23 little impulses 5.14 5.16 little impulses

even though with this new power supply all the voltages are higher, and thus so should the Vgpu, and Vmem, the problem is just as bad and even worse!! I get even worse artifacts, and it doesnt just pause for a second when throttling but sometimes even a few minutes.

Another interesting discovery:
http://www.xtremesystems.org/forums/...41#post1401141

Where the card throttles (and sometimes artifacts), is also where the "non-local memory usage" as shown in riavtuner has nearly increased to the size of the video memory, or where it quickly drops. This was taken while playing BF2.

Re: Audio. I dont even have a cable plugged into DVD drive.. it happens regardless of how many hard drives are connected. Its hard to pull all the data together because I have over a dozen excel spreadsheets with results. Although when it happens is pretty random, it is rather consistent in which scenes in 3DMark2005 it occurs in (corresponding to drops in Vgpu and Vmem), as well as changes in the game scene (either all memory filled or flushed). I have NF3, and fast writes doesnt work with my card. I get lockups and stuttering even at stock speeds.
04-19-2006, 12:23 PM
SikaRippa

Quote:

Originally Posted by ***Deimos***

Thanks for the link SilkaRippa!

I have switched power supply (again).

Code:

410/900 min max okia 450W atitool 11.92 11.94 oscilating wave 4.94 4.96 oscilating wave GT1 11.87 11.91 little impulses 4.98 5.02 little impulses enermax365W atitool 12.25 12.27 oscilating waves 5.08 5.1 little impulses GT1 12.21 12.23 little impulses 5.14 5.16 little impulses

even though with this new power supply all the voltages are higher, and thus so should the Vgpu, and Vmem, the problem is just as bad and even worse!! I get even worse artifacts, and it doesnt just pause for a second when throttling but sometimes even a few minutes.

Another interesting discovery:
http://www.xtremesystems.org/forums/...41#post1401141

Where the card throttles (and sometimes artifacts), is also where the "non-local memory usage" as shown in riavtuner has nearly increased to the size of the video memory, or where it quickly drops. This was taken while playing BF2.

Re: Audio. I dont even have a cable plugged into DVD drive.. it happens regardless of how many hard drives are connected. Its hard to pull all the data together because I have over a dozen excel spreadsheets with results. Although when it happens is pretty random, it is rather consistent in which scenes in 3DMark2005 it occurs in (corresponding to drops in Vgpu and Vmem), as well as changes in the game scene (either all memory filled or flushed). I have NF3, and fast writes doesnt work with my card. I get lockups and stuttering even at stock speeds.

I see you have been working & testing in a VERY HARD manner to get problems solved. Quite interesting observations regarding stuttering/memory usage etc. I'll try to get some time later to play with RT myself and test especially those memory usage variations. I have proposed that BS will be appeared just after a rapid memory usage change (=increase).

--

PSU: "enermax365W" is new one ? How you have characterized oscillating waves, oscilloscope, multimeter ? Any idea about frequency ? The higher 5V/12V voltage levels (within specs) should give better performance at least in respect of 'standard' BS. However, DC voltage of Vcore and Vmem should be exactly the same level set by the regulator as far as PSU is capable to provide enough current (ripple voltages of Vcore&Vmem may vary when using different PSU's)

--

Stock frequencies etc: Did you have the same (near to) problems when card was new or did you problems started later ?

--

Noisy audio: Is it coming directly from card ?
04-20-2006, 02:48 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

If you still have the capacitors, you can send them to me. Or at least let me know which brand, model, and capacitance value.

I tried doing the capacitor mods before, but according to SilkaRippa, the caps I bought are P O S (no surprise considering how cheap they were)

I'm on backorder for a 7900GT, but for brand model ect sikkarippa posted somewhere which type's where good iirc and you can always ask him. If I decide not to go ahead I will send them to you in the near future when I get my new card.

That 55xx score looks good, got a compare? I think the main diffrence is my low clocking tccd vs your bh5, and maybe to high clocks idk.
04-20-2006, 06:18 AM
***Deimos***

If by chance a nVidia driver software developer comes around here...., what by chance causes the driver or the BIOS to trigger changing clocks from 3D perfromance to low-power-3D.. and why would the video card ever throttle down to 2D clocks when running 3d app?
04-20-2006, 10:51 AM
Marvin_The_Martian

I think that answer would be temperature but I could be wrong ;)
04-23-2006, 11:33 PM
Firechicken

Hey guys been a while scince i started posting on this thread (page 18 and up) just thought i would let you guys know how my cap mod turned out.

I went ahead on the caps i got from sikarippa (a huge thanks to you sir !!)
and soldered them in and have been running the heck out of this card for quite a while now and have been able to do some extreme mem timmings with custom bios's ect, and have gotten no black screen she's running stable at 335 gpu and 905 mem or so, with everything i run, 3d mard 05,games ect..
too bad i cant unlock the last 4 pipes but sure is a big diff with new timmings and 12 pipes instead of the stock shut down version.

Just wanted to say a big thanks to you all that have helped put thier own unique experiences,talents,pictures and info here for us all to do what we love best tweaking and playing games :D. Also I want to express my gratitude to sikarippa for his ongoing fight and hard work into making all our lives a little easier,and helping to give us a little more performance out of our old faithfull 6800 Le's Thanks a million. M.L.Dovell.
04-24-2006, 12:34 AM
jlovrek

Quote:

Originally Posted by Firechicken

Hey guys been a while scince i started posting on this thread (page 18 and up) just thought i would let you guys know how my cap mod turned out.

I went ahead on the caps i got from sikarippa (a huge thanks to you sir !!)
and soldered them in and have been running the heck out of this card for quite a while now and have been able to do some extreme mem timmings with custom bios's ect, and have gotten no black screen she's running stable at 335 gpu and 905 mem or so, with everything i run, 3d mard 05,games ect..
too bad i cant unlock the last 4 pipes but sure is a big diff with new timmings and 12 pipes instead of the stock shut down version.

Just wanted to say a big thanks to you all that have helped put thier own unique experiences,talents,pictures and info here for us all to do what we love best tweaking and playing games :D. Also I want to express my gratitude to sikarippa for his ongoing fight and hard work into making all our lives a little easier,and helping to give us a little more performance out of our old faithfull 6800 Le's Thanks a million. M.L.Dovell.

you have right, sikarippa is the man, yeah! hm, can you write your memory timings and what memory do you have, cause i have 2.2ns from hynix and i can run 860mhz without artifacts, maybe 870... i have my card at 445/860 and i have 5050-5100 in 3dmark05... i have zalman fatality cooler.... tnx
04-24-2006, 03:52 AM
Marvin_The_Martian

SikkaRippa is the man but I'm glad I have the money to upgrade now, allot off very cheap cards available atm which blow any le out off the water ( 7600GT or even GS anyone? ).

I will keep the le though, never know if with a better cpu/ram I could climb a few spots on the orb and it's a fun card to bench with.

Hope my 7900GT will be a good clocker since I enjoy benching more then gaming lol and with my pos cpu/ram I can't do any damage with a full field off players compared with us LE owners :)
04-24-2006, 07:47 AM
Firechicken

Quote:

Originally Posted by jlovrek

you have right, sikarippa is the man, yeah! hm, can you write your memory timings and what memory do you have, cause i have 2.2ns from hynix and i can run 860mhz without artifacts, maybe 870... i have my card at 445/860 and i have 5050-5100 in 3dmark05... i have zalman fatality cooler.... tnx

Hi, I Just pulled the card out last night as i wrote this and dont know the mem timming's off the top of my head, I dont want to give you somthing
that might be wrong as you can cook a card pretty quick if you get it wrong
but I am putting it into my pop's computer maby tonight or in the next few days (he only has an ati 9600-xt) and is looking forward to being able to play his farcry and medal of honor with some decent settings, soon as it's in I will get the timmings along with what else i did to it's bios ( with nvflash) along with a run with 3d mark 05 to see where are two cards sit, his puter is only running an athalon xp 2800+ (not 64) with the older socket A with the 333 mhz Front side bus and 1x pc3200 and 1x pc2700 ram,so if the bench in 3d 05 dosnt come close then i will re-install it in my machine for a few runs to try and show you and me the diff in scores on the mem.

I do know for sure however that my mememory is the same as yours Hynix 2.2ns,but i have the stock cooler, I really have a crazy amount of air being pumped into the case i think it keeps the card cool enough.
M.D.
04-24-2006, 08:03 AM
Firechicken

Quote:

Originally Posted by MartianInvader

SikkaRippa is the man but I'm glad I have the money to upgrade now, allot off very cheap cards available atm which blow any le out off the water ( 7600GT or even GS anyone? ).

I will keep the le though, never know if with a better cpu/ram I could climb a few spots on the orb and it's a fun card to bench with.

Hope my 7900GT will be a good clocker since I enjoy benching more then gaming lol and with my pos cpu/ram I can't do any damage with a full field off players compared with us LE owners :)

Yeah I have had an ATI All in Wonder x-800-xt for about 9 months now and was really ticked when ati/nvidia came out with thier x1800/x1900...7800's/7900's as they doubled thier performance, but my old high end x800 dose some really nice overclocks and never ever whines at me when i push it harder. (6814 3d05 not pushing tooo hard). There's somthing that has me drawn to these little budget 6800's though I like the tinkering and tweaking and trying to make em run better and faster, and if you burn one up heh you learn somthing and you might not cry as hard as they could be replaced in a resonable amount of time on a small budget (like me lol). But please let me know how the 7900 runs when you get it I am very interested in buying one this year and selling off my x800-xt,
I really would like to see some scores on 3d mark 05 along with any other bugs/ feelings you have on it. Take care M.D.
04-24-2006, 08:25 AM
Marvin_The_Martian

Lol I should have them in sig but there's an easier way, just browse orb for le scores for hwa MtM which is me ;) I'm first page in every 3dmark with 6800le last time I checked so it's easier then letting look up all those compares and post them here :)

05 btw isn't my favourite because I can only get a measly 5430 ;) 03 is much better, and I think I stand allot higher to :)

Edit, pcmark05 6800le I'm not bad either, #3 atm and the two above me are a dual xeon which will kick the ass out off any venice in multithreading and the other says P4 but seeing it's above the dual xeon I recon it's also a dual cpu rig.

Edit: Ok lol I read that back and that what a pretensious litle :banana::banana::banana::banana::banana:!!! ;) So I'll link the compares anyway :)
pcmark http://service.futuremark.com/compare?pcm05=284589
3d01 http://service.futuremark.com/compare?2k1=8923405
3d03 http://service.futuremark.com/compare?2k3=4650814
3d05 http://service.futuremark.com/compare?3dm05=1856608
3d06 http://service.futuremark.com/compare?3dm06=35494
04-24-2006, 09:34 AM
***Deimos***

Quote:

445/860 and i have 5050-5100 in 3dmark05

I certainly hope thats with LOD and high performance in the driver. After about 380-400, my 6800 becomes very bandwidth limited, and thats with memory running at 900, 950, or even 972 (after voltmod). Go figure. With core only at 375, the GT runs memory at 1000, and Ultra with core at 400/425, memory at 1200.

for everybody who is only getting 800's with 2.2ns Hynix
get NIBITOR 2.8 from mkvtech
you can test timings in windows without flashing!!
for 900+, I found I MUST have tRCDRD=6 and tRP=6.

Intructions:
1. Tools-Read BIOS- Select Device-OK
2. Tools-Read BIOS- Read into NIBITOR- OK
3. Go to Timings tab, and click autoselect to change to proper timingset- ok
4. click detail timings, and you will see your timings.. make changes as desired. When you click ok, you will notice, that the Timing0, Timing1, Timing3 have changed. Write these down!
5. Back at the Timings tab, Click test timings, get timings, and insert the values you wrote down, click set timings, and go run atitool artifact tester (highly recommended) or some other benchmark.

EX:
Timing0: 060C1512
Timing1: 05010307
Timing2: 00240306

correspnd to:
tRC=18
tRFC=21
tRAS=12
tRP= 6
tRCDRD=6
tRCDWR=3
tRRD=4
04-25-2006, 01:40 AM
jlovrek

Quote:

Originally Posted by ***Deimos***

I certainly hope thats with LOD and high performance in the driver. After about 380-400, my 6800 becomes very bandwidth limited, and thats with memory running at 900, 950, or even 972 (after voltmod). Go figure. With core only at 375, the GT runs memory at 1000, and Ultra with core at 400/425, memory at 1200.

for everybody who is only getting 800's with 2.2ns Hynix
get NIBITOR 2.8 from mkvtech
you can test timings in windows without flashing!!
for 900+, I found I MUST have tRCDRD=6 and tRP=6.

Intructions:
1. Tools-Read BIOS- Select Device-OK
2. Tools-Read BIOS- Read into NIBITOR- OK
3. Go to Timings tab, and click autoselect to change to proper timingset- ok
4. click detail timings, and you will see your timings.. make changes as desired. When you click ok, you will notice, that the Timing0, Timing1, Timing3 have changed. Write these down!
5. Back at the Timings tab, Click test timings, get timings, and insert the values you wrote down, click set timings, and go run atitool artifact tester (highly recommended) or some other benchmark.

EX:
Timing0: 060C1512
Timing1: 05010307
Timing2: 00240306

correspnd to:
tRC=18
tRFC=21
tRAS=12
tRP= 6
tRCDRD=6
tRCDWR=3
tRRD=4

no, that's with default settings in driver... i don't know what is lod... hm, i have test this timings and still i can't have 900 without artifacts.... i don't know why?
04-25-2006, 11:33 AM
***Deimos***

Are you using default heatsink? How are the mem chips being cooled? I have a NV5 making firm contact with them. While running ATITool (with 5 min warmup) you can touch the back of the PCB where memory chips are. If ts hot, you need better cooling for memory. Its also a good idea to have at least a large fan blocking at the back where all the chips are in the corner. They get very very hot.
04-26-2006, 01:30 AM
jlovrek

Quote:

Originally Posted by ***Deimos***

Are you using default heatsink? How are the mem chips being cooled? I have a NV5 making firm contact with them. While running ATITool (with 5 min warmup) you can touch the back of the PCB where memory chips are. If ts hot, you need better cooling for memory. Its also a good idea to have at least a large fan blocking at the back where all the chips are in the corner. They get very very hot.

i have zalman fatality cooler and passive heatsinks from zalman for memory... on cpu i have titan vanessa s-type with 92mm fan that is placed vertical and it is blowing air to the other side of pcb... maybe the voltage is to low? i have gainward 6800le agp, and somebody on other forum said that gainward has memory at 2.55v, but i don't know what is the true? i can't measure it because i don't have equipment... i think that temperature shouldn't be a problem...
04-26-2006, 02:28 AM
Marvin_The_Martian

If you get 51xx without lod, that's actually very good :banana:
Pm me if you want some help beating me on the orb I won't mind being passed by ;)

2.2ns hynix needs 2.7v afaik, no way the voltage is that low. Did you do bios mod to increase vgpu? If not that corespeed is fantastic to :banana:
I can do around 420 on 1.1v, but up to 470 on 1.4v with zalman vf700cu. I also do have that fan blowing over pcb ( vreg on back off the card, caps near molex connector ) without that I get random blackouts. The ram get's warm to the touch after a bench on 970 and artifacts like hell even with 2.9v, score was even lower then with 950.
04-26-2006, 04:27 AM
jlovrek

Quote:

Originally Posted by MartianInvader

If you get 51xx without lod, that's actually very good :banana:
Pm me if you want some help beating me on the orb I won't mind being passed by ;)

2.2ns hynix needs 2.7v afaik, no way the voltage is that low. Did you do bios mod to increase vgpu? If not that corespeed is fantastic to :banana:
I can do around 420 on 1.1v, but up to 470 on 1.4v with zalman vf700cu. I also do have that fan blowing over pcb ( vreg on back off the card, caps near molex connector ) without that I get random blackouts. The ram get's warm to the touch after a bench on 970 and artifacts like hell even with 2.9v, score was even lower then with 950.

i have 1.5v and 445mhz on core... can you post picture of your case near pcb?
04-26-2006, 06:32 AM
SikaRippa

Firechicken, jlovrek and MartianInvader ; Thank you for your support :toast:

Those replacement capacitors I have sent for severel users (~ 15 sets !!!) were obtained as samples from local supplier of Nippon Chemi-con. Thus, the most credits should be given to Chemi-con !! Once I got those I decided that the best purpose for them will be definately repairing of BS-cards. It was for me quite small task to send some capacitors, in the respect of PAIN of user having a BS-affected card without RMA possiblity !!!
04-26-2006, 06:37 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

I certainly hope thats with LOD and high performance in the driver. After about 380-400, my 6800 becomes very bandwidth limited, and thats with memory running at 900, 950, or even 972 (after voltmod). Go figure. With core only at 375, the GT runs memory at 1000, and Ultra with core at 400/425, memory at 1200.

for everybody who is only getting 800's with 2.2ns Hynix
get NIBITOR 2.8 from mkvtech
you can test timings in windows without flashing!!
for 900+, I found I MUST have tRCDRD=6 and tRP=6.

Intructions:
1. Tools-Read BIOS- Select Device-OK
2. Tools-Read BIOS- Read into NIBITOR- OK
3. Go to Timings tab, and click autoselect to change to proper timingset- ok
4. click detail timings, and you will see your timings.. make changes as desired. When you click ok, you will notice, that the Timing0, Timing1, Timing3 have changed. Write these down!
5. Back at the Timings tab, Click test timings, get timings, and insert the values you wrote down, click set timings, and go run atitool artifact tester (highly recommended) or some other benchmark.

EX:
Timing0: 060C1512
Timing1: 05010307
Timing2: 00240306

correspnd to:
tRC=18
tRFC=21
tRAS=12
tRP= 6
tRCDRD=6
tRCDWR=3
tRRD=4

Thanks for accurate timing tweak info !

One question releted to the test results you provided earlier. How did you got the 'local video memory usage statistics' active in RT ? I could not found those settings (or maybe I have too old version...)
04-26-2006, 06:46 AM
SikaRippa

Quote:

Originally Posted by jlovrek

i have zalman fatality cooler and passive heatsinks from zalman for memory... on cpu i have titan vanessa s-type with 92mm fan that is placed vertical and it is blowing air to the other side of pcb... maybe the voltage is to low? i have gainward 6800le agp, and somebody on other forum said that gainward has memory at 2.55v, but i don't know what is the true? i can't measure it because i don't have equipment... i think that temperature shouldn't be a problem...

The memory voltage of these cards is normally 2.72 V (+/- ~ 0.02 V). However, the speed of Hynix chips is set/limited by supply voltage and 2.95 is maximum acceptable value.

If you want to test whether increase of Vmem will have some effect there is an easy way to make mod:
http://img216.imageshack.us/img216/4...mod16ir.th.jpg

.. in theory it can be done without measuring actual voltage, BUT still I highly recommended measuring voltages before/after (and resistance of used resistor) !!
04-26-2006, 07:08 AM
Marvin_The_Martian

Quote:

Originally Posted by jlovrek

i have 1.5v and 445mhz on core... can you post picture of your case near pcb?

1.5v and 445 isn't great. Is that 1.5v flashed or 1.5v measured? I flashed 1.4 before and though I was getting that, but when I measured it it was still 1.1v. I flashed again at 1.6v and I now measured 1.4v. With 1.4v I can go to 470, but the scores are lower. If I had wc on the chip it would respond better I think :confused: because off less electro thermal dissapation or something like that.

I will put up pics off the pcb but I'm not sure what you would want to see?

Edit: the mod in the post above is truly easy, I done it like I saw it done by someone else with a fan connector molex so I can remove the mod when I'm not benching. I'm thinking next time I'll use a dipswitch though because getting a vr's legs to connect with the molex isn't allways 100% certain. I put small blobs off solder on the legs so I'm sure the contact is good, but that means I have to remember to power everything off because some tin could come loose and kill something. Which would have happend if I didn't have rememberd to power down lol, loose tin on mobo traces near some caps wasn't something I liked seeing :)
04-26-2006, 07:15 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

I certainly hope thats with LOD and high performance in the driver. After about 380-400, my 6800 becomes very bandwidth limited, and thats with memory running at 900, 950, or even 972 (after voltmod). Go figure. With core only at 375, the GT runs memory at 1000, and Ultra with core at 400/425, memory at 1200.

for everybody who is only getting 800's with 2.2ns Hynix
get NIBITOR 2.8 from mkvtech
you can test timings in windows without flashing!!
for 900+, I found I MUST have tRCDRD=6 and tRP=6.

Intructions:
1. Tools-Read BIOS- Select Device-OK
2. Tools-Read BIOS- Read into NIBITOR- OK
3. Go to Timings tab, and click autoselect to change to proper timingset- ok
4. click detail timings, and you will see your timings.. make changes as desired. When you click ok, you will notice, that the Timing0, Timing1, Timing3 have changed. Write these down!
5. Back at the Timings tab, Click test timings, get timings, and insert the values you wrote down, click set timings, and go run atitool artifact tester (highly recommended) or some other benchmark.

EX:
Timing0: 060C1512
Timing1: 05010307
Timing2: 00240306

correspnd to:
tRC=18
tRFC=21
tRAS=12
tRP= 6
tRCDRD=6
tRCDWR=3
tRRD=4

Also thanks for the timings info :toast: , never tinkerd with that before. Was hoping it will get me to 1gig ddr lol I know it has been done before but still it's a nice thing to archieve. Most untill now is 970 passing 05, and 03 always passed higher for me then 05 so I know I'm very close :)
04-26-2006, 07:28 AM
***Deimos***

For the Hynix memory chips, overvolting doesnt help at all.. just more heat, and more stability problems. But, you definetly need at least 2.7V. I found highest overclock at 2.8-2.84V: higher voltage was worse.

You definetly need a multimeter to measure the Vgpu and Vmem properly. Otherwise, you dont really know. The value in the VGA BIOS or Rivatuner (for Vgpu) is often wrong.
04-26-2006, 09:07 AM
Marvin_The_Martian

Strange mine did scale up to 2.9v, after that oc didn't increase. It is harder to keep stable but I don't blame that on my mem chips but my modding skills lol!
04-26-2006, 11:18 AM
***Deimos***

Wrong. It certainly wasnt your modding skills. Doesn't matter what temperature how well modded, there is a finite voltage limit, and a threshold beyond which you see sharp diminishing returns. Most ICs nowadays are made on very very fine lithography process with very small channels. When over-volting, you can quickly run into speed of light limit. Difference of voltage at the drain vs voltage at the source (ie, what you modded to) divided by the distance (typically 90-110nm) results in very high Volts/m gradient in the range of thousands and even millions of volts/meter gradient (velocity saturation and mobility degredation). Such highly energy electrons proportionally cause more damage to the silicon crystal lattice as hill:banana::banana::banana::banana: and other defects, shortening the life of the IC. Numerous other short channel effects also come into play like channel length modulation, subthreshold leakage, body effects, gate tunneling, and various temperature dependencies (carrier mobility decreases with increased temperature) to name a few.

Long story short, once you notice that increasing volts isn't helping much, its a sure sign that you have more than enough volts and that there are other limiting factors.
04-27-2006, 01:07 PM
SikaRippa

Quote:

Originally Posted by ***Deimos***

Wrong. It certainly wasnt your modding skills. Doesn't matter what temperature how well modded, there is a finite voltage limit, and a threshold beyond which you see sharp diminishing returns. Most ICs nowadays are made on very very fine lithography process with very small channels. When over-volting, you can quickly run into speed of light limit. Difference of voltage at the drain vs voltage at the source (ie, what you modded to) divided by the distance (typically 90-110nm) results in very high Volts/m gradient in the range of thousands and even millions of volts/meter gradient (velocity saturation and mobility degredation). Such highly energy electrons proportionally cause more damage to the silicon crystal lattice as hill:banana::banana::banana::banana: and other defects, shortening the life of the IC. Numerous other short channel effects also come into play like channel length modulation, subthreshold leakage, body effects, gate tunneling, and various temperature dependencies (carrier mobility decreases with increased temperature) to name a few.

Long story short, once you notice that increasing volts isn't helping much, its a sure sign that you have more than enough volts and that there are other limiting factors.

I agree most of that story, it is clearly seen you HAVE knowledge about submicron CMOS technology.

However, some comments: at first a short investigation about speed vs. voltage&temperature and at end more specific comment about this case.

In general the higher supply voltage the higher speed is obtained with a CMOS logic WHEN the junction temperature is kept constant. Self heating of circuit increases gate delays (decreases mobility inside channel), so in practice acquired speed is not linearily increased as a function of supply voltage.

I made a following electrical SPICE-simulation:

- ring-oscillator , 3 CMOS inverters
- Transistor parameters 0.13 um CMOS by IBM, used for GF6800/LE/GT chips
- Parameters from MOSIS (www.mosis.org)
- Transistor sizes PMOS: W = 2 um, L =0.13 um, NMOS W = 1 um, L =0.13 um (just an engineering guess..)

I simulated oscillation frequency as function of supply voltage at different temperatures. The frequency in the figure shown below is normalized (1 = 8.7 GHz).

http://img103.imageshack.us/img103/1...13ro2sz.th.jpg

Results:

1. Increasing the supply voltage increases the maximum operating frequency quite linearily (self heating is not considered)
2. Cooling is very important factor when a maximal operational speed is required.

Notes:

1. Accuracy of the models is not guaranteed
2. The minimum/maximum voltages (1.0 V, 1.6 V) might not be usable for GPU, since GPU is using much more complex logic.

---

Speed of Hynix-memories used in 6800-cards is of course more complex case than the investigation above. I think that those chips can operate in a reliable manner having supple voltage at least up to 2.95 V. This voltage level is specified in datasheet as a maximum operating voltage, and there is specified that 3.6 V as an absolute maximum voltage at any pin in any condition. Therefore, I suppose they have utilized a CMOS technology that can withstand such voltage level (2.95 V) without gate oxide degeneration (leakage etc.). Since, degeneration effects take place in long term I think those can not explain why speed is not increased anymore after some voltage level. An instantenous breakdown of e.g. gate oxide or source drain junction has of course catastrophic consecuenses and usually means a burn chip. But in short term as long as you keep below 'high voltages' and take care of cooling (=current/power does not 'melt' chip) I think the more voltage - the more speed should be expected. The same logic can be seen in datasheet, since they specify different supply voltages to chips with different speed specification (propably they are using the same IC-process for all versions and measure/select best out).

Another point is application of the logic, here the interface of the memory chips. The high supply increases I/O logic levels to/from GPU, and maybe after some level the I/O circuitry speed of GPU 'saturates'. Moreover, the observed speed of VGA-card is fuction of several different aspects and it is possible that after certain memory clock frequency some other bottle-neck will appear.
04-27-2006, 01:53 PM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

Wrong. It certainly wasnt your modding skills. Doesn't matter what temperature how well modded, there is a finite voltage limit, and a threshold beyond which you see sharp diminishing returns. Most ICs nowadays are made on very very fine lithography process with very small channels. When over-volting, you can quickly run into speed of light limit. Difference of voltage at the drain vs voltage at the source (ie, what you modded to) divided by the distance (typically 90-110nm) results in very high Volts/m gradient in the range of thousands and even millions of volts/meter gradient (velocity saturation and mobility degredation). Such highly energy electrons proportionally cause more damage to the silicon crystal lattice as hill:banana::banana::banana::banana: and other defects, shortening the life of the IC. Numerous other short channel effects also come into play like channel length modulation, subthreshold leakage, body effects, gate tunneling, and various temperature dependencies (carrier mobility decreases with increased temperature) to name a few.

Long story short, once you notice that increasing volts isn't helping much, its a sure sign that you have more than enough volts and that there are other limiting factors.

huh I thought I used a smiley in that post lol....
Guess I didn't. Point being I knew the short version, long version is way way way to xtreme for me but if my oc doesn't increase when upping the volts.... yeah I dig that :)
04-27-2006, 06:17 PM
***Deimos***

I dont know SikaRippa.. I always thought the capacitance was the major limitation for DRAM (like the Hynix 2.2ns GDDR sitting on my 6800). Afterall, as featuresizes get smaller, if you only had tiny area for the capacitor in each cell, you would need very fast refresh, or just wouldn't work. I believe they use various "roughing" techniques to increase the effective area of the capacitor. It has been a while so I'm not sure.

I would certainly be interested if somebody could confirm if large DRAM suppliers like Samsung use low-K dielectric between the interconnects to reduce interconnect to interconnect capacitance (it certainly seemed to be a big issue on complex R400 chips where the low-K really helped, but want to know if also used for high speed memory). As for the gate oxide, that can be improved with high-K dielectric. I dont think any range close to the supply voltage would be sufficient to cause any serious damage to the gate oxide. The kind of breakdown you descriibe I think is more typical of electro - static shock (typically hundreds to thousands of volts).
04-28-2006, 01:05 PM
SikaRippa

I am try give a short answer. The technology used for DRAM's is quite different to logic-CMOS procesesses, that's clear. The big issue with DRAM's is indeed: how-to-make-as-much-as-possible-capacitance-in-a-very-tiny-area. All methods are used as you already listed, and the CMOS part can be quite poor compared to state-of-art CMOS. However, once technology (=capacitance value/square um) is fixed and memory cells are designed & laid out the speed of a memory cell vs. supply voltage should have similar characteristics as in standard CMOS i.e. the higher supply voltage the higher possible access rate. The capacitance keeping DRAM state should remain quite constant while supply is higher level, so it does not have influence to speed. Another point, why Hynix specify higher supply voltage for higher frequencies, if supply does not have effect to speed ?

Breakdown voltages of submicron CMOS-transistors are usually in magnitude of ~2 times of supply voltages. Generally the weakest part is gate oxide, which starts to tunneling heavily ~ 2*VDD voltage level ((yes, I have measured myself)). The source and drain connections can handle higher voltages.

---

Summary (IMHO):

In theory: higher supply voltage gives higher speed, when temperature is low enough and supply voltage is lower that absolute maximum ratings.

In practice: The obtained speed in a high APPLICATION LEVEL (like running 3d benchmark) is sum of several factors: GPU/memory interface, temperature of all components, software etc. As a result some users do not get significant improvement to 'memory frequency' while having higher voltages.
04-28-2006, 02:57 PM
***Deimos***

Oh I am not arguing at all that higher supply voltage will increase the speed of the DRAM. I can confirm that easily with my BH5 memory! However, I am saying that there are voltage threshold beyond which adding more voltage will not help increase Mhz, regardless of cooling etc, and that this voltage threshold is dependent on the manufacturing process and design of the IC (some are more tolerant like famous Winbond BH5)
04-29-2006, 06:51 AM
jlovrek

ok, today i have replaced zalman fatality cooler with nv5 silencer in hope to get +900mhz on my 2.2ns memory... my temp. went down for about 5-10degrees at load, but my memory isn't more overclockable then before... i'm so :mad: !!!! it stil at 860mhz without artifacts! what should i do? voltmod? hm, i still can't measure my voltage of memory... sikarippa, with your caps, is there any change in voltage of memory? cause they have higher capacity...
04-30-2006, 01:48 PM
SikaRippa

Quote:

Originally Posted by ***Deimos***

Oh I am not arguing at all that higher supply voltage will increase the speed of the DRAM. I can confirm that easily with my BH5 memory! However, I am saying that there are voltage threshold beyond which adding more voltage will not help increase Mhz, regardless of cooling etc, and that this voltage threshold is dependent on the manufacturing process and design of the IC (some are more tolerant like famous Winbond BH5)

I think the case about Vmem & OC-capability is now fully considered. In the case if there was (?) a 6800-card having separed VDD/VDDQ voltages some kind of extra tweaking/testing could be possible by changing the VDD/VDDQ -ratio.
04-30-2006, 01:58 PM
SikaRippa

Quote:

Originally Posted by jlovrek

ok, today i have replaced zalman fatality cooler with nv5 silencer in hope to get +900mhz on my 2.2ns memory... my temp. went down for about 5-10degrees at load, but my memory isn't more overclockable then before... i'm so :mad: !!!! it stil at 860mhz without artifacts! what should i do? voltmod? hm, i still can't measure my voltage of memory... sikarippa, with your caps, is there any change in voltage of memory? cause they have higher capacity...

Voltage mod might give some boost and the replaced capacitors should also enable better OC-capability. However, if the solderering of the capacitors is not proper (=cold solder) this has a huge effec: how well you soldering proceeded ?

It is not suggested to make Vmod's without measuring. The mod I proposed is very easy to do even without mesuring, but allways something can be happened and it that case you might have running card (for a while) and too high voltage levels (short life).
04-30-2006, 02:50 PM
jlovrek

Quote:

Originally Posted by SikaRippa

Voltage mod might give some boost and the replaced capacitors should also enable better OC-capability. However, if the solderering of the capacitors is not proper (=cold solder) this has a huge effec: how well you soldering proceeded ?

It is not suggested to make Vmod's without measuring. The mod I proposed is very easy to do even without mesuring, but allways something can be happened and it that case you might have running card (for a while) and too high voltage levels (short life).

well, the soldering has passed very well, that was doing a tv-service man... i think that solder of caps isn't a problem cause everything works great, i don't have bs any more, just my memories couldn't pass 900mhz withous artifacts... here are numbers from my memory chips:
HY5DU283222
AF-22 338A
04-30-2006, 02:59 PM
SikaRippa

Quote:

Originally Posted by jlovrek

well, the soldering has passed very well, that was doing a tv-service man... i think that solder of caps isn't a problem cause everything works great, i don't have bs any more, just my memories couldn't pass 900mhz withous artifacts... here are numbers from my memory chips:
HY5DU283222
AF-22 338A

OK, soldering should not be problem. The point of my question was, that users have been complaining not to get good performance (or remove BS) even when capacitors replaced. In many cases (AFAIK) the problem was due to cold soldering which increases series resistance = similar effect than ruined capacitors.

Your memories should be capable to run >900 MHz, so Vmod might help (900 MHz operation requires 2.8 V instead of 2.72 V according to datasheet).
05-01-2006, 08:13 AM
***Deimos***

jlovrek... if you only use Vmem mod and capacitor mod you shouldn't expect more than 900Mhz. You need to change the timings. I have posted detailed instructions and several sets of timings to try in previous posts here.
05-01-2006, 11:55 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

jlovrek... if you only use Vmem mod and capacitor mod you shouldn't expect more than 900Mhz. You need to change the timings. I have posted detailed instructions and several sets of timings to try in previous posts here.

I think that luck (=memory chips are unique) plays important role: some users have managed to get ~ 1000 MHz even at stock voltages (and even more with higher voltage). Take a look post #11:

http://www.xtremesystems.org/forums/...ad.php?t=62651
05-01-2006, 12:08 PM
***Deimos***

Most 6800 nu/le have Hynix GDDR (typically 2.8ns or 2.2ns). 2.8ns~DDR700, 2.2ns~DDR900.
I got the biggest boost in memory speed on my card by changing the timings. Before I could barely get up to 900. Afterwards, 960+ (60Mhz increase just by chnaging tRP=6, and tRCDRD=6). With Vmem almost 984 stable.
05-01-2006, 10:03 PM
jlovrek

Quote:

Originally Posted by ***Deimos***

Most 6800 nu/le have Hynix GDDR (typically 2.8ns or 2.2ns). 2.8ns~DDR700, 2.2ns~DDR900.
I got the biggest boost in memory speed on my card by changing the timings. Before I could barely get up to 900. Afterwards, 960+ (60Mhz increase just by chnaging tRP=6, and tRCDRD=6). With Vmem almost 984 stable.

i have tried with all sets of timings and still nothing... damn!
05-04-2006, 12:12 PM
SikaRippa

Quote:

Originally Posted by jlovrek

i have tried with all sets of timings and still nothing... damn!

I think the Vmem mod posted is 99 % safe even without measuring actual voltage. However you should be absolutely sure that:

- You have a correct value resistor !!! (see that guide, e.g. 3300 ohm should give ~2.85 V and 2200 ohm ~2.9 V)
- Do not make any shorts !!
05-04-2006, 05:53 PM
***Deimos***

I find reading values of fixed value resistors a bit ambiguous. Sure, you can try to memorize the whole colour scheme with some rhyme, but its sometimes hard to tell apart brown from orange, and its so easy to get mixed up in the head. And ofcourse you still have the 10%, 5% or whatever, tolerance.

Nothing beats have a multimeter. Its the only way to measure the value for sure. Its nice to know the resistor you hold in your hand is nearly precisely 9.981KOhm
05-05-2006, 10:07 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

jlovrek... if you only use Vmem mod and capacitor mod you shouldn't expect more than 900Mhz. You need to change the timings. I have posted detailed instructions and several sets of timings to try in previous posts here.

Ehm even with no cap mod and with stock voltage I bench at 950, with 2.9v 970 and that's with default timings :stick:

Haven't treid iff the timingsmod will get me over 1ghz though but I certainly will, again :toast: for the info
05-05-2006, 10:40 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

I find reading values of fixed value resistors a bit ambiguous. Sure, you can try to memorize the whole colour scheme with some rhyme, but its sometimes hard to tell apart brown from orange, and its so easy to get mixed up in the head. And ofcourse you still have the 10%, 5% or whatever, tolerance.

Nothing beats have a multimeter. Its the only way to measure the value for sure. Its nice to know the resistor you hold in your hand is nearly precisely 9.981KOhm

Yep, the color coding system is not for humans... I agree also that multimeter is essential for every mods to be sure what you have done. My point was that this mod is quite safe to do in sense of how easily the resistor can be added (=risk to make mistake is low) and if you buy e.g. 3300 ohm resistor (low tolerance, say +/- 1%) the resulting Vmem is quite accurately within correct range.
05-05-2006, 01:58 PM
***Deimos***

Today was the day. Last night I broke off a fin from a copper heatsink. Today, I went to dollar store and got some super glue. Using very sharp hunting knife I scored the fin and broke off 3 little pieces of copper. One side looked like it was sand blasted, so I sanded it down first. I taped off the surrounding area, just in case (there are some pretty tall ceramic capacitors around). I applied a tiny bit of super-glue on each IC, spread around some AS5, attached the copper shims, and then repeated the process to attach the heatsink. Originally, I was going to salvage an old P3 heatsink, but I found a nice small chipset heatsink which seems to fit well.

Results:
ambient temperatures are actually HIGHER. But, I think its working. After running some benchmarks the homemade heatsink gets pretty hot.. good 50-60C I guess.. but thats much better than the broiling hot before. Unfortunately it did not improve my throttling situation... but at least it didn't make it worse.

EDIT: The diode is probably behind the tape.. before when I had a 80mm case fan blowing at the back of the card, it probably caused diode to read lower temp.
05-05-2006, 06:42 PM
***Deimos***

Quote:

5.5 Temperature measurements

I have measured temperatures of the cards: 'Gainward 6800le version 1b' and MSI 6800le (BS-affected). I used a Pt100-sensor (high accuracy platinum temperature sensor) glued at top to wooden stick (see foto below), which I gently put top of each component.

Figure 33. Temperature sensor (Pt100) attached at the tip of wooden stick

The accuracy of the sensor is high, but due to poor thermal contact the measured temperatures are too optimistic (=one or two degrees too low). Results for both card can be found below:

Figure 34. Measured temperatures of GW6800le and MSI6800le-cards

Hmm.. by the looks of things I definetly have temperatures similar to those in the 2nd picture. Is the 1st picture idle temperatures, and the 2nd one under load.. I sure hope so, because thats how warm they are idle, and thats how hot they are under load.
05-06-2006, 06:21 AM
Firechicken

Quote:

Originally Posted by ***Deimos***

Today was the day. Last night I broke off a fin from a copper heatsink. Today, I went to dollar store and got some super glue. Using very sharp hunting knife I scored the fin and broke off 3 little pieces of copper. One side looked like it was sand blasted, so I sanded it down first. I taped off the surrounding area, just in case (there are some pretty tall ceramic capacitors around). I applied a tiny bit of super-glue on each IC, spread around some AS5, attached the copper shims, and then repeated the process to attach the heatsink. Originally, I was going to salvage an old P3 heatsink, but I found a nice small chipset heatsink which seems to fit well.

Results:
ambient temperatures are actually HIGHER. But, I think its working. After running some benchmarks the homemade heatsink gets pretty hot.. good 50-60C I guess.. but thats much better than the broiling hot before. Unfortunately it did not improve my throttling situation... but at least it didn't make it worse.

EDIT: The diode is probably behind the tape.. before when I had a 80mm case fan blowing at the back of the card, it probably caused diode to read lower temp.

lol I feel your pain, send her to me and i'll braze it back together with brass,
are you able to solder it back together? if it's a really poor alloy only brass can save you ( but I like the super glue the only handy man's way for fixing some things hehehe) oh and dont forget lotrs of duct tape :).
05-07-2006, 09:13 PM
***Deimos***

Quote:

Originally Posted by Firechicken

lol I feel your pain, send her to me and i'll braze it back together with brass,
are you able to solder it back together? if it's a really poor alloy only brass can save you ( but I like the super glue the only handy man's way for fixing some things hehehe) oh and dont forget lotrs of duct tape :).

huh?
ok, so I know superglue is quite a good heat and electrical insulator, but I only put a tiny bit on one corner of only 2 of the ICs.. I used AS5 everywhere else.. due to my reluctance to splurge nearly $10 at Radio Shack. As fasr as I can tell.. its all working 100% fine.. its just that I now regret using the small chipset heatsink when I could have slapped on a Pentium Pro bad boy (one step away from SLK90.. certainly excessive)

Upon further testing, I have really noticed the effect of the fan. Before, with only the weak fan blowing. Temps would go up considerably under load.. but not as much as without the fan (ofcourse). Now, with the fan blowing directly at the chipset heatsink on the back of the 6800, even after 15min+ on a rather warm day, the ambient temperature monitored by rivatuner only went up to 28-29C! I still dont like how hot it gets without the fan.. and ultimately will probably change the heatsink if it really bothers me, but for now, I guess I can scratch off MOSFET/IC cooling as my throttling culprit ;)
05-09-2006, 10:53 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

Today was the day. Last night I broke off a fin from a copper heatsink. Today, I went to dollar store and got some super glue. Using very sharp hunting knife I scored the fin and broke off 3 little pieces of copper. One side looked like it was sand blasted, so I sanded it down first. I taped off the surrounding area, just in case (there are some pretty tall ceramic capacitors around). I applied a tiny bit of super-glue on each IC, spread around some AS5, attached the copper shims, and then repeated the process to attach the heatsink. Originally, I was going to salvage an old P3 heatsink, but I found a nice small chipset heatsink which seems to fit well.

Results:
ambient temperatures are actually HIGHER. But, I think its working. After running some benchmarks the homemade heatsink gets pretty hot.. good 50-60C I guess.. but thats much better than the broiling hot before. Unfortunately it did not improve my throttling situation... but at least it didn't make it worse.

EDIT: The diode is probably behind the tape.. before when I had a 80mm case fan blowing at the back of the card, it probably caused diode to read lower temp.

Sounds very nice mod, indeed !

You are most propably right with the diode (sensing temperature). After that kind of modding temperature MUST be lower in the critical areas: it is not important how the diode is if the core/MOSFET's etc. are cool.

Quote:

Hmm.. by the looks of things I definetly have temperatures similar to those in the 2nd picture. Is the 1st picture idle temperatures, and the 2nd one under load.. I sure hope so, because thats how warm they are idle, and thats how hot they are under load.

The most important (?) difference between those figures is core voltage: 1.1 V and 1.2 V. Those temperatures are measured from different cards, but using the same setup/chassis in the same day etc. so those results should be highly comparable. There are two different cases in both fotos idle/load (in parenthesis ) as explained in text.
05-09-2006, 09:49 PM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

Now, with the fan blowing directly at the chipset heatsink on the back of the 6800, even after 15min+ on a rather warm day, the ambient temperature monitored by rivatuner only went up to 28-29C!

My rig is on 24/7 folding with my le, and now while it's starting to get real hot ambient temp readout from the vc is 33c to 35c but I haven't noticed any bad side effects. This is with a 80mm fan blowing over the pcb to cool some off the caps near the molex connector and the two black chips on the backside which regulate current, which were smoking hot after I flashed a higher core voltage. I thought cooling those two regulator chips was the most important?
05-10-2006, 04:34 AM
***Deimos***

Quote:

Originally Posted by MartianInvader

My rig is on 24/7 folding with my le, and now while it's starting to get real hot ambient temp readout from the vc is 33c to 35c but I haven't noticed any bad side effects. This is with a 80mm fan blowing over the pcb to cool some off the caps near the molex connector and the two black chips on the backside which regulate current, which were smoking hot after I flashed a higher core voltage. I thought cooling those two regulator chips was the most important?

well, ofcourse a lot of it depends on the ambient temperature (room temperature). Assuming based on where the memory voltage reading points are and the capacitors on the front of the card, the two upper 8 leg SOP
IC are the two transistors for the memory buck converter, and the ones at the bottom of the card, near where the AGP tab is, are the transistors for the GPU voltage buck converter. That big rectangular 32 pin TSOP is the voltage regulator for both of them... there is a really interesting pdf with all its specs that SikaRippa provided.

On the first page he also has some pics with temperature measurements. My card corresponds to the 2nd one.. ie its quite hot.
05-10-2006, 09:58 AM
Marvin_The_Martian

Quote:

Originally Posted by ***Deimos***

there is a really interesting pdf with all its specs that SikaRippa provided..

The two vmem reg's I described earlier where the one's getting the hottest, after I found out they were fine with a fan I never looked at the pdf again. Not an electrician here, I only digest what I need :toast:

But I'm moving my pc's to a cooler room anyway, just wanted to comment on temps being that high without bad side effects here.
05-10-2006, 10:52 AM
SikaRippa

Quote:

Originally Posted by MartianInvader

The two vmem reg's I described earlier where the one's getting the hottest, after I found out they were fine with a fan I never looked at the pdf again. Not an electrician here, I only digest what I need :toast:

But I'm moving my pc's to a cooler room anyway, just wanted to comment on temps being that high without bad side effects here.

Yep, those THREE MOSFET's close each other (2 with 8 pins, 3rd usually a little bit different but in some cards also 8 pin) are in any condition (=Vmods) running at highest temperature AFAIK. Usually no problems appear even GPU voltage is increased a little bit, but these MOSFET's seem to be the weakest link e.g. burn at first => cooling of those = good.

In finnish forum has been reported that also other MOSFET's (for Vmem regulator) are vulnerable to burn e.g. after of capacitor modification. Since, this kind of problems are rarely reported I think in those cases the MOSFET's have been more poor quality than average.
05-10-2006, 04:33 PM
***Deimos***

Quote:

Originally Posted by MartianInvader

The two vmem reg's I described earlier where the one's getting the hottest, after I found out they were fine with a fan I never looked at the pdf again. Not an electrician here, I only digest what I need :toast:

But I'm moving my pc's to a cooler room anyway, just wanted to comment on temps being that high without bad side effects here.

Now that is quite interesting... since in my case, they are barely warm.. and this is after 5min of atitool, or rhdribl.

Once again, my situation is almost EXACTLY like the 2nd temperature measurement picture posted by SikkaRippa in his first post.

As for replacing that mosfet.. we'll see.. my brother went by that electronics store today, but got confused with all the ripple current, leakage current etc... I also forgot to tell him to get SMT type, since I dont think I would be able to fit NV5 overtop if it was typical radial lead package.
05-11-2006, 07:37 AM
Firechicken

Sorry about that deimos, I read your post wrong and thought you broke your heat sink by accident. I have welded back together some pretty expensive
0.03 deg c/w heatsinks ( a measure of heatsink dissipation) along with a few aftermarket video card coolers people have had me fix up. Must have been a late night when I read your post sorry lmao...MD.
05-16-2006, 11:17 AM
SikaRippa

Quote:

Originally Posted by ***Deimos***

As for replacing that mosfet.. we'll see.. my brother went by that electronics store today, but got confused with all the ripple current, leakage current etc... I also forgot to tell him to get SMT type, since I dont think I would be able to fit NV5 overtop if it was typical radial lead package.

I think here SMT MOSFET's are MUST also in the electronic point of view: the inductance of the leads etc. parasitics ruins easily performance.

'Local electronic shops' do not have too many SMT versions and high performance capacitors (solid electrolytic, extremely high max ripple current at stock, since that kind of stuff is usually not asked by a standard 'hobby electronics'.
05-19-2006, 11:06 AM
ya_master

yeah. :banana:

i did the cap mod at TP10 & TP9 with the black screen affected gainward 6800LE of my sister and it seems to have worked. no bs within about 2 hours of gaming. that should do the trick.

thank you for the guide. was just a matter of minutes to get the show running. :) :) :)

i almost forgot to mention:
the card's strange beep noise while playing is gone too.
05-20-2006, 10:51 AM
jlovrek

Quote:

Originally Posted by ya_master

yeah. :banana:

i did the cap mod at TP10 & TP9 with the black screen affected gainward 6800LE of my sister and it seems to have worked. no bs within about 2 hours of gaming. that should do the trick.

thank you for the guide. was just a matter of minutes to get the show running. :) :) :)

i almost forgot to mention:
the card's strange beep noise while playing is gone too.

everything seems great for now. but with a time bs will be back 99% more often and often...
05-20-2006, 11:46 AM
SikaRippa

Quote:

Originally Posted by ya_master

yeah. :banana:

i did the cap mod at TP10 & TP9 with the black screen affected gainward 6800LE of my sister and it seems to have worked. no bs within about 2 hours of gaming. that should do the trick.

thank you for the guide. was just a matter of minutes to get the show running. :) :) :)

Nice :up:

Quote:

Originally Posted by ya_master

i almost forgot to mention:
the card's strange beep noise while playing is gone too.

Noise from CARD or SPEAKERS ? You mean whistling etc. *modulated* somehow by framerate (changes) e.g. high noise during loading screens of benchmarks ? I am asking because of this kind of noise is not usually fully disappeared after cap mods.

Quote:

Originally Posted by jlovrek

everything seems great for now. but with a time bs will be back 99% more often and often...

It depends.. with some (most/all/?) MSI cards it has been seen that added capacitor do not help too long (weeks/months..). However, GW-cards and especially those NOT having "4 3 470 6E" capacitors seems to usually work quite well/long with added QUALITY capacitor. The poor design of the card is reason that propably all cards will be BS-affected after some years, but hopefully none of us is not anymore using these cards in that day.

I hope your sisters GW will work long enough without extra mods (replacement capacitors are much harder to obtain/install).
05-20-2006, 12:29 PM
***Deimos***

I definetly have that noise, during games, but most noticable during loading screens. I've also successfully gotten fast writes "kind of" working. If the GPU frequency is <370, it just works, and is transparent. Raising GPU higher, on first benchmark or timedemo, the whole screen flashes... all textures flash.. but when you close the application, and restart it, everything is fine.. UNTIL... some 10-12min looping 3d01 Nature later (at 425/950 lets say), it suddenly gets stuck in 3/4 speed mode. Now, an interesting thing that happens in this "mode", is vertex shader performance is way down to about half, and high polygon tests, especially the one with 1 light source, go down to like 1/10th. But, single and multi texturing fill rate tests still show near identical results.. interesting no?

There is no way to get my GPU unstuck from this. Chaning clocks, AGP, fastwrites, CPU speed, re-starting the driver, nothing works to unfix it. I always have to restart the computer.
05-20-2006, 01:45 PM
ya_master

nah. sister reported a blackscreen yesterday while running skialpin demo. but just one time and everything else seems to run quite fine. maybe BS is not completely gone but it's a real big improvement anyway.

would it be possible to avoid BS by increasing vmem a little?

or by using a better psu?

Quote:

Originally Posted by SikaRippa

Noise from CARD or SPEAKERS ? You mean whistling etc. *modulated* somehow by framerate (changes) e.g. high noise during loading screens of benchmarks ? I am asking because of this kind of noise is not usually fully disappeared after cap mods.

noise from the card. i can't tell exactly because i just took a look at the sh*t for about one hour before i grabbed that cap and soldered it to the card.
i don't know if there was noise while showing the 2d windows desktop but while running a 3d game it was definitly really noisy. now it seems to be completely gone. but i will have a closer look tomorrow.
05-21-2006, 11:25 AM
SikaRippa

Quote:

Originally Posted by ya_master

nah. sister reported a blackscreen yesterday while running skialpin demo. but just one time and everything else seems to run quite fine. maybe BS is not completely gone but it's a real big improvement anyway.

would it be possible to avoid BS by increasing vmem a little?

or by using a better psu?

noise from the card. i can't tell exactly because i just took a look at the sh*t for about one hour before i grabbed that cap and soldered it to the card.
i don't know if there was noise while showing the 2d windows desktop but while running a 3d game it was definitly really noisy. now it seems to be completely gone. but i will have a closer look tomorrow.

Too bad :( Some questions/comments.

1. What codes there are top of green capacitors e.g. "4 3 470 6E" ?
2. What kind of capacitor you used, exact coding capacitance ... did you solder it ?

Increase of Vmem propably makes things worse ! Better PSU might help but usually BS will be back no-matter-how-good-PSU you have. So, check capacitor specifications and post numbers/codes to get an idea whether the added capacitor is good enough.
05-21-2006, 11:55 AM
ya_master

yes i soldered the cap but it looks like i will have to remove that one and replace it with another one. the ones i used for my card (just capmod. i had no bs). a better one than the one i soldered i hope. that one came from the guy i bought the card and maybe nood good enough.

got blackscreens again today while running alpinski and elite force 1. one each.

but after that everything was fine for hours. strange sh*t.

i will have a closer look at it when the zalman vf 700 for the card arrives. then i'll mount the cooler, add some heatsinks, look at the caps, the strange noise, the voltages, the temps, solder the other cap, try enabling the last quad and take a look at overclocking.

this is really getting on my nerves.

i did some soldering and the heatsink was removed and replaced by another one for a time but is it possible to rma the card directly to gay-nward? or will it fail cause to the removal of the stock cooler anyway?

card is 1 1/2 years old.
05-23-2006, 04:08 AM
ya_master

things seem to get better.

- new zalman vf 700 arrived today and i mounted it with arctic silver 5 and the ram heatsinks
- i didn't add other heatsinks. the little 2 things in the card don't get too warm and the big silver box from my leadtek card is represented by a thing with open wires in the gainward card of my sister. that one gets really hot and i can't apply a heatsink to it. damn. they should use these boxes in the future... :mad:
- psu voltages seem to be ok. no new one needed i think.
- found out the gpu is running with 1.3v. looks like it's really working on gainward cards.
- removed the old cap i added before and redid the capmod with one of my caps.
- ran 15 games of alpinski :stick: and no blackscreen appeared. could be gone now but i need my sister to do more 'testing'. :D
- card is now running at 440 / 850. 460mhz core gave erros with atitool and so did 870mhz mem. i will leave it at 440 / 850 and everything will be fine. was already flashed to the bios anyway.
- ran atitool for 30 minutes without errors and no BS at 12/6. now trying the same with 16/6. that would be nice if it works at 16/6. my sister's card is getting better than my own. argh! ;)
- markings on the stock caps are ... 470 6E. that's sh*t right?
- the noise from the card only appears while loading screens get faded out. you know? this fade to black / fade to white. that gives some noise. anything else doesn't. but when i enabled the card to run at 16/6 the noise was back and playing all time while unter 3d load. but not as loud as before the capmod.

i think that was all i wanted to mention. :confused:
05-23-2006, 04:18 AM
***Deimos***

Quote:

Originally Posted by ya_master

things seem to get better.

- new zalman vf 700 arrived today and i mounted it with arctic silver 5 and the ram heatsinks
- i didn't add other heatsinks. the little 2 things in the card don't get too warm and the big silver box from my leadtek card is represented by a thing with open wires in the gainward card of my sister. that one gets really hot and i can't apply a heatsink to it. damn. they should use these boxes in the future... :mad:
- psu voltages seem to be ok. no new one needed i think.
- found out the gpu is running with 1.3v. looks like it's really working on gainward cards.
- removed the old cap i added before and redid the capmod with one of my caps.
- ran 15 games of alpinski :stick: and no blackscreen appeared. could be gone now but i need my sister to do more 'testing'. :D
- card is now running at 440 / 850. 460mhz core gave erros with atitool and so did 870mhz mem. i will leave it at 440 / 850 and everything will be fine. was already flashed to the bios anyway.
- ran atitool for 30 minutes without errors and no BS at 12/6. now trying the same with 16/6. that would be nice if it works at 16/6. my sister's card is getting better than my own. argh! ;)
- markings on the stock caps are ... 470 6E. that's sh*t right?
- the noise from the card only appears while loading screens get faded out. you know? this fade to black / fade to white. that gives some noise. anything else doesn't. but when i enabled the card to run at 16/6 the noise was back and playing all time while unter 3d load. but not as loud as before the capmod.

i think that was all i wanted to mention. :confused:

woah.. backup for a second there partner..
you are getting 440 running in all games stable without any artifacts what so ever, and no throttling... ie
if you use rivatuner to set perfromance3d to 440, and lets say low-power-3d=410, and 2d=400.. that it will remain at 440 while its running the game/benchmark. If so, you truelly have a remarkable chip, or mine is fubar, and was defective from the get-go (btw: I get my problems regardless if I'm running 16/6,12/5, or even 8/3.. yes I was THAT desperate)

Show 100 post(s) from this thread on one page

All times are GMT -8. The time now is 01:07 PM.

XtremeSystems