Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 06-11-05, 03:53 AM   #25
xorbe
Unbuffered User
 
Join Date: Jun 2005
Posts: 388
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

Quote:
Originally Posted by jsgf
Modeline "1600x1200" 155 1600 1704 1872 2128 1200 1201 1204 1222 +hsync +vsync
Works for me: 59.6Hz refresh.
Thanks! That worked around the issue! (Interestingly it borks on the older driver...)
(**) NVIDIA(0): Mode "1600x1200": 155.0 MHz, 72.8 kHz, 59.6 Hz



btw, how did you generate that? 155MHz maxes at "gtf 1600 1200 57.87" for me, but you managed "1600 1200 59.6"

Oooh, http://xtiming.sourceforge.net/cgi-bin/xtiming.pl the trick is to lower horizontal sync time at the bottom to less than what they recommend I think.

Here's a perfect one for me at 60Hz:
Modeline "1600x1200@60" 154.98 1600 1632 2016 2048 1200 1224 1236 1261

(**) NVIDIA(0): Mode "1600x1200@60": 155.0 MHz, 75.7 kHz, 60.0 Hz


.

Last edited by xorbe; 06-11-05 at 04:11 AM.
xorbe is offline   Reply With Quote
Old 06-12-05, 01:40 PM   #26
phred_sd
Registered User
 
Join Date: Jun 2005
Posts: 4
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

I just got Fedora Core 4 up and running, and the nvidia driver refuses to work. It looked like it might be the same issue here (GF4 Ti4200, DVI, Dell 2001FP) but the modeline fix and the rollback to the previous driver (removing the modeline) don't work. Both had to compile everything, and reported success, but I can only get it to work with the old "nv" driver. Any ideas? TIA.

Here's the error shown in the text window:
Code:
...
(==) Using config file: "/etc/X11/xorg.conf"
XIO:  fatal IO error 104 (Connection reset by peer) on X server ":0.0"
      after 0 requests (0 known processed) with 0 events remaining.
Here's the tail end of the Xorg.0.log after the crash:
Code:
...
(II) NVIDIA(0): Screen initialization complete
(==) RandR enabled
(II) Initializing built-in extension MIT-SHM
(II) Initializing built-in extension XInputExtension
(II) Initializing built-in extension XTEST
(II) Initializing built-in extension XKEYBOARD
(II) Initializing built-in extension LBX
(II) Initializing built-in extension XC-APPGROUP
(II) Initializing built-in extension SECURITY
(II) Initializing built-in extension XINERAMA
(II) Initializing built-in extension XFIXES
(II) Initializing built-in extension XFree86-Bigfont
(II) Initializing built-in extension RENDER
(II) Initializing built-in extension RANDR
(II) Initializing built-in extension COMPOSITE
(II) Initializing built-in extension DAMAGE
(II) Initializing built-in extension XEVIE
(II) Initializing extension GLX
phred_sd is offline   Reply With Quote
Old 06-14-05, 12:06 AM   #27
xorbe
Unbuffered User
 
Join Date: Jun 2005
Posts: 388
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

If you are trying to bring the box up for the first time, stick to the -7174 driver, and you'll know it's not the driver.
xorbe is offline   Reply With Quote
Old 06-14-05, 01:18 AM   #28
phred_sd
Registered User
 
Join Date: Jun 2005
Posts: 4
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

I tried that driver...same problem, even with the modeline and IgnoreEDID lines removed. I've never had an issue with 3D drivers in the past (same card, even), though this is the first time I've tried them with my 2001FP and the DVI interface. Just about every resolution looked like the above ones - the clock was too fast, h or vsync out of range.

http://www.nvnews.net/vbulletin/showthread.php?t=51972

Here's someone else having an issue. Is there something FC4-specific going on here?
phred_sd is offline   Reply With Quote
Old 06-14-05, 03:28 AM   #29
zander
NVIDIA Corporation
 
zander's Avatar
 
Join Date: Aug 2002
Posts: 3,740
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

@phred_sd: this looks like a different problem, possibly Fedora Core 4 related; do you have SELinux enabled?
zander is offline   Reply With Quote
Old 06-14-05, 09:11 AM   #30
phred_sd
Registered User
 
Join Date: Jun 2005
Posts: 4
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

Quote:
Originally Posted by zander
@phred_sd: this looks like a different problem, possibly Fedora Core 4 related; do you have SELinux enabled?
Yes. Maybe that's the issue. It seemed harmless enough when it asked...I tried 'setenforce 0' and that didn't allow the driver to work...though it may have to do with how it apparently modifies the kernel rather than whether it is enabled or not at the time.

I haven't really gotten anywhere, so maybe I'll just reinstall without SELinux and try again.
phred_sd is offline   Reply With Quote
Old 06-14-05, 09:00 PM   #31
phred_sd
Registered User
 
Join Date: Jun 2005
Posts: 4
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

Quote:
Originally Posted by phred_sd
I haven't really gotten anywhere, so maybe I'll just reinstall without SELinux and try again.
That was it, apparently....no SELinux, and the old driver version works fine now. Thanks for the info....you might see if there's a workaround or something that might belong in the installer/readme (unless I missed it ) to avoid this.
phred_sd is offline   Reply With Quote
Old 06-19-05, 11:33 PM   #32
nox771
Registered User
 
Join Date: Jun 2005
Posts: 2
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

Just an FYI for anyone with a Samsung monitor like mine - Samsung SyncMaster 213T. On a fresh install of Fedora Core 4, I had the same problem getting DVI to work at 1600x1200 as everyone else (it worked with the 'nv' driver but not with 7664 'nvidia' one). I finally managed to get a working modeline (pulled from the log when using the 'nv' driver, which seemed to have better diagnostics).

Monitor section is as follows:

Section "Monitor"
Identifier "Monitor0"
VendorName "Samsung SyncMaster 213T"
ModelName "LCD Panel 1600x1200"
HorizSync 30.0 - 81.0
VertRefresh 56.0 - 75.0
# 130.4MHz 74.1kHz 60Hz modeline
Modeline "1600x1200" 130.4 1600 1648 1680 1760 1200 1202 1206 1235
Option "dpms"
EndSection

I also had to add the following line into the screen section, otherwise the 'nvidia' driver defaulted to the analog output (which was strange because the 'nv' driver had no problem defaulting to the DVI output):
Option "ConnectedMonitor" "DFP"
With that modeline it didn't need the "IgnoreEDID" option. Took a couple hours to figure out this one, so best of luck to everyone else.
nox771 is offline   Reply With Quote

Old 06-20-05, 01:22 PM   #33
freelsjd
Registered User
 
Join Date: Feb 2003
Posts: 73
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

This seemed like a great idea (to use the nv driver to determine the correct Modeline settings and then place that into the xorg.conf file and try the nvidia driver). So I did this and got the following Modeline for my Dell 2000FP monitor:

Modeline "1600x1200" 162.00 1600 1664 1856 2160 1200 1201 1204 1250 +hsync +vsync

I also get the following output in my Xorg.log file from the nvidia monitor at version 7174

(II) NVIDIA(0): Clock range: 12.00 to 400.00 MHz

However, under the new 7664 driver, the equivalent information output line is

(II) NVIDIA(0): Clock range: 12.00 to 135.00 MHz

Thus, it rejects the Modeline and I cannot achieve 1600x1200. Note I also tried the NoBandWidthTest option and it did not help.

I can get 1600x1200 using the nv driver, but no dvi only analog screen.

So, for now I am back to the 7174 driver again.
__________________
J D Freels
freelsjd is offline   Reply With Quote
Old 06-21-05, 02:05 AM   #34
nox771
Registered User
 
Join Date: Jun 2005
Posts: 2
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

Quote:
Originally Posted by freelsjd
Modeline "1600x1200" 162.00 1600 1664 1856 2160 1200 1201 1204 1250 +hsync +vsync
I got curious about this modeline and did some checking. Using this page:

http://www.tkk.fi/Misc/Electronics/f...2rgb/calc.html

it looks like that modeline is the VESA timing for 1600x1200. Further, looking at the monitor specs on the Samsung 213T and the Dell, they both support 162MHz clocks (which matches the VESA timing):

http://www.samsung.com/au/products/m...Specifications
http://support.dell.com/support/edoc...lish/specs.htm

This all makes it look like the DVI connection should support at least 162MHz (and higher for larger panels). My log also shows the 135MHz limit. I don't know why it is there, it seems arbitrary, its not a monitor limitation. The modeline I pulled came from the 'nv' log, which had the following:
(II) NV(0): Supported VESA Video Modes:
(II) NV(0): 720x400@70Hz
(II) NV(0): 640x480@60Hz
(II) NV(0): 640x480@67Hz
(II) NV(0): 640x480@72Hz
(II) NV(0): 640x480@75Hz
(II) NV(0): 800x600@56Hz
(II) NV(0): 800x600@60Hz
(II) NV(0): 800x600@72Hz
(II) NV(0): 800x600@75Hz
(II) NV(0): 832x624@75Hz
(II) NV(0): 1024x768@60Hz
(II) NV(0): 1024x768@70Hz
(II) NV(0): 1024x768@75Hz
(II) NV(0): 1280x1024@75Hz
(II) NV(0): 1152x870@75Hz
(II) NV(0): Manufacturer's mask: 0
(II) NV(0): Supported Future Video Modes:
(II) NV(0): #0: hsize: 1600 vsize 1200 refresh: 60 vid: 16553
(II) NV(0): #1: hsize: 1280 vsize 1024 refresh: 60 vid: 32897
(II) NV(0): Supported additional Video Mode:
(II) NV(0): clock: 130.4 MHz Image Size: 432 x 324 mm
(II) NV(0): h_active: 1600 h_sync: 1648 h_sync_end 1680 h_blank_end 1760 h_border: 0
(II) NV(0): v_active: 1200 v_sync: 1202 v_sync_end 1206 v_blanking: 1235 v_border: 0
(II) NV(0): Ranges: V min: 56 V max: 75 Hz, H min: 30 H max: 81 kHz, PixClock max 140 MHz
I don't know where the "additional" mode comes from, but it seems to work on the Samsung (it is not VESA timing). It seems the real fix is to remove the 135MHz limit from the driver.
nox771 is offline   Reply With Quote
Old 06-21-05, 03:14 AM   #35
kulick
Todd Kulick
 
Join Date: Jun 2005
Location: Like, in the Bay Area
Posts: 23
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

I've been fumbling through this problem with my FX5700 Ultra and a Samsung SyncMaster 243T (24" LCD, native 1920x1200). I think I've finally decided that the reduced pixel clock that I see under the latest (7664) nVidia driver is a limitation of my graphics card. Well, not mine exactly, mine will happily run quite a bit faster.

It seems like nVidia may have changed the latest drivers to enforce the expected/supported TMDS clock rates for their cards. (Aside: don't confuse the TMDS (DVI) pixel clock limit with the RAMDAC limit (like I did at first)...they are separate chips/circuitry and probably have different specs!) Since alot of nVidia chips can actually go faster than the specs that they are sold under, people can get things to work that, according to the card's supported capabilities, should not.

The work around that I chose was to go back one rev on the driver. For me the reduced pixel clock modes wouldn't work as my monitor really, really wanted a very specific 154MHz pixel clock 1920x1200 format. The new driver wanted to limit my FX5700 Ultra to 150MHz.

I bet the limited pixel clock that people are seeing is a function of your card. Do ppl want to post their card model along with their pixel clock limit under 7664? I bet that might confirm the pattern. If this fact is true, we may not see this "feature" disappear in future drivers. It sure would be nice if nVidia added a flag to disable this limitation. You can overclock lots of other parts of the card with things like nvTweak...

In general, I do like the strategy of examining your monitor's EDID output (or manual, I guess) to find the exact native format and then setting that up in your xorg.conf. If you have to, tune it a little to reduce the pixel clock. Just remember, sometimes persnickety monitors might not like it much when you deviate significantly (or even minorly!) from the timing format that they expect. My Samsung appears to be one such monitor. In fact, mine is so picky that even using the EDID reported modeline only works if I also specify "ExactModeTimingsDVI" in the configuration file. I wish I knew what the heck the driver was doing with that flag...then I could probably get things working under windows too!

-t
kulick is offline   Reply With Quote
Old 06-21-05, 04:40 AM   #36
arokh
Registered User
 
arokh's Avatar
 
Join Date: Jun 2004
Posts: 287
Default Re: 7664 and 1600x1200 (maximum pixel clock?)

kulick, this is a bug not a limitation in the cards.

7174:
(--) NVIDIA(0): Display device CRT-0: maximum pixel clock at 8 bpp: 400 MHz
(--) NVIDIA(0): Display device CRT-0: maximum pixel clock at 16 bpp: 400 MHz
(--) NVIDIA(0): Display device CRT-0: maximum pixel clock at 32 bpp: 400 MHz

7664:
(--) NVIDIA(0): DFP-0: maximum pixel clock: 155 MHz
(--) NVIDIA(0): DFP-0: Internal Single Link TMDS

I was running 1600x1200 fine with 162MHz pixel clock earlier.
__________________
Asus A8N32-SLI : AMD64 3700+ @ 2.85GHz : 2GB PC4000 OCZ Platinum EB : 2x 7800GT SLI
arokh is offline   Reply With Quote
Reply


Thread Tools

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

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 09:02 AM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.