Nvidia WDM Driver (VIVO)

[alfer]

I have a Geforce 4 TI 4200 card with AGP8x, and vivo. The card works fine, tv out works flawlessly. I have the latest detonator drivers installed (43.45) and the system runs windows xp. My problem is that when i install the latest wdm drivers (1.22) for some reason the "nVidia WDM Video Capture (universal)" device in device manager does not start with windows, I get "This device cannot start. (Code 10)" every time i reboot. If i uninstall the device and let it get reinstalled, the video capture will work for that session, but after i restart, i get the same problem. The other 3 devices installed by the WDM drivers work fine and do not have this problem. I have tried other versions (beta 1.23, 1.19, 1.0 all of them have the same problem. Does anyone else have these drivers installed and working or know what the problem is with them?

[marcocom]

try this. try completely uninstalling the nvidia drivers for your WDM (from add/remove programs) and upon rebooting let Microsoft install its default reference drivers for the produc. tell us if the conflict still exists or if after rebooting, it still fails to start. this will tell us if your conflict is driver level, or IRQ/system/ACPI level

(keep the accelerated 42.x VGA drivers by nvidia installed or reinstall them if needed afterwards)

(if the OS refuses to boot without automatically installing a newer, non-reference driver, your going to need to get in the WINDOWS/INF folder and search for and delete all files named "OEM*.INF" that have "NVIDIA WDM"in its header.read them in a text editor like TextPad4 or even notepad. these are the default drivers for unknown devices so that it doesnt ask you everytime you reconnect a device)

[alfer]

OK, here are some things ive noticed. If I uninstall the drivers from add/remove programs, remove the oem*.inf files relating to the wdm portion, and uninstall the devices in device manager, when i reboot, windows does not detect those 4 devices at all, and will not attempt to install drivers or even detect the devices until i run the wdm driver setup program again. Also, with the built in winxp drivers 31.45 i think, the device does start and it doesnt have the problem, however, atv2000 (my tv tuner program) doesnt display any video.

[marcocom]

wait.

give it time.

reboot and then give it at least 20 minutes. (its silly i know, but microsoft feels that the plug and play process would be best if the user not even know its happening.. not even its status...brilliant)

[muzz]

Originally posted by marcocom
wait.

give it time.

reboot and then give it at least 20 minutes. (its silly i know, but microsoft feels that the plug and play process would be best if the user not even know its happening.. not even its status...brilliant)

Thats pretty strange..... I'm waiting to hear how this turns out.... be back in a half hr.

[alfer]

Still no change. When the drivers are uninstalled, windows xp doesnt automatically detect those 4 devices at all. Does anyone actually have these drivers installed properly on their system? Maybe if i knew a working combination of display drivers and wdm drivers i could get this working. I did a google search also, it seems quite a few people had the same problem back with versions 1.08 and 1.19, but noone ever seemed to offer a solution.

[muzz]

Didn't think so........ and it's been 2 hrs.....

Sorry to hear of your troubles alfer.

[marcocom]

Originally posted by alfer
Still no change. When the drivers are uninstalled, windows xp doesnt automatically detect those 4 devices at all. Does anyone actually have these drivers installed properly on their system? Maybe if i knew a working combination of display drivers and wdm drivers i could get this working. I did a google search also, it seems quite a few people had the same problem back with versions 1.08 and 1.19, but noone ever seemed to offer a solution.

i dont run the nvidia WDM. i have a pixelview for video in. im interested to see if theres a hardware level conflict, i guess. try this, goto START / RUN.. and type "msinfo32" and then drill down from the first node named "Hardware Resources" to the Conflicts/Sharing dialogue. do you see that WDM sharing an IRQ maybe with another device? maybe your soundcard? (see some devices can handle sharing an IRQ and some cant. but that doesnt mean they dont all try it - especially if a driver isnt written properly or isnt updated for you OS maybe. dunno)

whats it tell ya?

[marcocom]

Originally posted by muzz
Didn't think so........ and it's been 2 hrs.....

Sorry to hear of your troubles alfer.

its a friday night, homey. you do have plans...right?

[marcocom]

good news dude.

the problem has been isolated by others and they say they have been able to remedy this by buying a quality PSU with a strong voltage reading on the 3.3v rail .

If you cant afford a nice new one (i recommend Antec. california ripened and built for quality) theres a guy who says you might want to try a hack -


Re: WDM Drivers for Geforce 4 TI 4600 DON'T WORK!!!
Here is a solution that worked for me. the main problem seems to be with the 3.3V supply, the capture device can't start if the 3.3V is too low, it works fine once it is on but it needs a "jump start". my hardware monitor has always reported it to be 3.24V - 3.25V although I measured 3.301 on all the 3.3V wires from the PSU (Aopen), this doesn’t seem too bad but maybe the GFX card uses onboard regulators that are unable to deliver the startup current needed by the capture device if the voltage is too low?
Anyway, the PSU regulates the 3.3V rail according to a sense wire from the motherboard, a sort of feedback indicating if the 3.3V on the MB is ok, if not the PSU will increase or decrease the voltage until the sense wire shows 3.3V, so what I did was to install a variable resistor on the sense wire, that way the PSU thinks that the 3.3V on the MB is too low and cranks up the 3.3V rail.
The sense wire is nr 11 on the atx connector, mine has 2 orange wires in the same location, one is 3.3V supply and the other is the sense which on my PSU is thinner than all the other wires (I tried both).
I cut the wire (there goes the warranty!!) and mounted a variable resistor mine is 2.2K ohm but go for a smaller one if you are going to try this, mine is set at only 14 ohms so a 22 or 47 ohm would be perfect.
Make sure that it is set at the LOWEST RESISTANCE before you plug it in (did you remember to pull the plug before you started?).
This might be enough to make it work, if not turn it up a notch and try again (reboot or turn pc off while adjusting), you don’t have to have 3.3V shown in MBM or bios, mine is only 3.28 but works fine (maybe my GFX will OC higher now!!)
Of cause, I take no responsibility if you try this and your MB starts to smell funny and glow in the dark, all I’m saying is that it worked for me.

tech:
a 0.02V voltage drop over the resistor will give you 3.32V on the 3.3V rail and raise the voltage shown in bios by 0.02V, but not necessarily the same value, dono why.


sniffed it off of this thread d00d - http://www.hardwareanalysis.com/content/topic/2235/

also some people said that they had some decent results with a 3rd party driver from some smart germans try the (www.hauppauge.com) wdm drivers for w2k(WDM 3.31) at http://hauppauge.lightpath.net/software

The IRQ conflict is an issue as well, that some say they have seen.
sounds to me like the WDM drivers havent really been updated in some time by a real engineer. they do this all the time at companies, when a product's excitement wears out and it gets handed off to Legacy Support and its all about patching instead of a fresh build for say...the XP update to the Win2k drivers, for example.

[Fusion]

Hmm, I can't see it being the 3.3 rail.
I've had the same problems on quite a few different PC's I've built in the last few weeks, and even my own at times.
But my 3.3 rail is nearly always on 3.34, and fluctuates between 3.31 and 3.35. But never under 3.3.

It seems like a combination of the latest WDM 1.22, and the 43.45 (and newer) drivers.

In WinME, it's a friggin nightmare as WinME doesn't actually remove the device manager drivers. Instead, you are left with ghost drivers and registry entries that sit there forever.

Here's my own solution that worked in XP.

Remove the drivers from "Add Remove", Reboot.

Run regedit and remove everything with the following entries

NVTUNE
NVCAP
NVTVSND
NVXBAR

Just do a search for each entry, then wipe it.
Do a windows search on your main drive for

*NVTUNE*.*
*NVCAP*.*
*NVTVSND*.*
*NVXBAR*.*

And wipe anything that is found.
Reboot, then re-install the 1.22 drivers.
Worked for me.
Although, on one PC I had to install the 1.19 drivers, then UPDATE to the 1.22. :confused:

[muzz]

Originally posted by marcocom
its a friday night, homey. you do have plans...right?

Whats the matter marco?
Do you feel the need to attack me, because I don't agree with something I feel is foolish?
OS's DON'T need to warm up( this is not an old tube radio afterall).........thats the first I heard of that 1.:rolleyes:

[marcocom]

thanks for contributing , muzz

[muzz]

YW marco....... anytime.:firedevil

[alfer]

FUSION:

I followed your instructions exactly. Same problem. I used 1.08, 1.19, 1.22, and the beta 1.23 from a site i found. Nothing worked. I used the exact steps you suggested. I think the nvidia dev team has given up, but who knows,

aLfer

Originally posted by marcocom
good news dude.

the problem has been isolated by others and they say they have been able to remedy this by buying a quality PSU with a strong voltage reading on the 3.3v rail .

If you cant afford a nice new one (i recommend Antec. california ripened and built for quality) theres a guy who says you might want to try a hack -



sniffed it off of this thread d00d - http://www.hardwareanalysis.com/content/topic/2235/

also some people said that they had some decent results with a 3rd party driver from some smart germans try the (www.hauppauge.com) wdm drivers for w2k(WDM 3.31) at http://hauppauge.lightpath.net/software

The IRQ conflict is an issue as well, that some say they have seen.
sounds to me like the WDM drivers havent really been updated in some time by a real engineer. they do this all the time at companies, when a product's excitement wears out and it gets handed off to Legacy Support and its all about patching instead of a fresh build for say...the XP update to the Win2k drivers, for example.

[Fusion]

FUSION:

I followed your instructions exactly. Same problem. I used 1.08, 1.19, 1.22, and the beta 1.23 from a site i found. Nothing worked. I used the exact steps you suggested. I think the nvidia dev team has given up, but who knows, :

Hmm, strange. Sorry to hear about that.

I'll have a word with a couple of friends I know.

Any chance of a link to those 1.23 drivers ?

[sxotty]

What card do you actually have?

I had a gainward and it worked, I had a leadtek and it never worked. This part 3rd parties have control over so it actually depends on what company you buy it from.

For example on a Firingsquad review of the same Leadtek, they said

"if we were able to get the VIVO feature working, we’d be able to claim doing just this, but alas, this wasn’t the case. A little disappointed over the video-in issue, we’re hoping that all issues have been resolved by the time you read this. "

As when they reviewed it Leadtek had not even bothered to make it work. I think they have it fixed, but when I tried to use it it did not function.

[alfer]

My Card is a chaintech A-GX21 Ti4200, but i dont think it matters since this problem has been reported by many different people using different manufacturers.

[alfer]

Great news for anyone else having this problem. It seems that the nvidia developers have decided to do some work and have fixed the problem, at least for me. After some vigorous searching, I found a beta version of the drivers v1.27 on gainwards site (My card is a chaintek, but chaintek's website is horrible) The link is here (http://driver.gainward.com.tw/VGA/Driver/nVidia/WDM).

I completely uninstalled 1.22 from add/remove programs, rebooted, installed these 1.27 drivers, rebooted, and upon start up xp will present 4 install driver boxes. Make sure to select install from a list (Advanced) and then dont search, i will choose the driver to install IF you have used previous versions, because for some reason xp wanted to install the old drivers rather than 1.27.

These drivers have fixed my problem, the device starts normally, no error message. Hopefully NV releases these to the public soon since I would imagine quite a few people have the same problem.

[undrwater]

I've been having similar problems, although i'd never gotten capture to actually work.

I've noticed that when downloading the WDM drivers, certain files get copied, but the one we are worried about is the nvcap.sys file.

If you have a WDM driver for your particular card, you might want to download that one.

Next, uncompress and install the driver as normal.

What i did next was to compare the nvcap.sys from my uncompressd file with the one in windows\system32\drivers directory. On mine the size of the file was different, so i copied the file from the uncompressed driver over the one in the windows directory, said yes to overwrite that one, and rebooted.

After reboot, checked device manager, no problems with capture driver there, started AsusDVCR and watched the video coming from my VCR! :D

I truly, truly, truly hope this helps some of you...I've been working at this for too long!

Good Luck!