I-Appliance BBS The Official Source for Internet Appliance Upgrades and Mods BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.net Reply to Thread | Printer |

Home / I-Opener Areas / I-Opener Technical Stuff FixUSB

FixUSB (modified 0 times) Programmer Profile There is a program that needs to be tested to see if it successfully solves the problems with USB networks. at present it will only run from a real mode dos (autoexec.bat). If it is determined to be successful I will add it to an expansion rom that will load with the bios, making it available to users of all operating systems. 04-25-2001 10:50:55

RE:FixUSB (modified 0 times) Guy2000 Profile This sounds interesting -- any updates? In the meantime, after months of intermittent network problems with a couple of different USB NICs, I finally got a wireless 802.11b USB adapter from DLink. And it works so much better -- no pauses during large file transfers, no system hangs during MP3 playback etc. 05-13-2001 20:07:04

RE:FixUSB (modified 0 times) thbt Profile If one of you has installed the VIA 4-in-1 drivers and USB Filter drivers, can you do me a huge favour? Go into your Control Panel -> System -> Device Manager, and see if your USB Root Hub device or USB Universal Host Controllers actually use device drivers from VIA. I've installed both of the aforementioned drivers, but I don't think they took. The driver files of these two devices (such usbhub.sys, uhcd.sys, etc) as all have "Microsoft Corporation" stamped on them. the VIAUSB.SYS file that comes with the VIA USB Filter drivers aren't even copied into my Windows subtree, much less used by any USB device. Any help would be appreciated. 05-13-2001 23:50:55

RE:FixUSB (modified 0 times) Programmer Profile well that'd explain why noone ever got back to me.. I never posted the link.. oops.. http://bethie.net/~programmer/FixUSB.com so.. uh .. try it out and let me know .. K? I did try this on my wifes machine(which also has a via chipset) (not her IO) and it fixed the problem of her POS flatbed scanner locking up in the middle of a scan. (it's still a POS, now it's a reliable POS.. :)) 05-14-2001 05:41:58

RE:FixUSB (modified 0 times) ckbone Profile Programmer...whenever I try your link, I get a blank white page, with a few letters in the upper left corner. Are these the files? What do I need to access them? I'm using Netscape 4.77. This also happened when I tried to get your memory interleave progs. Thanks.... 05-15-2001 03:46:31

RE:FixUSB (modified 0 times) * StarFish * Profile ckbone, Try Shift-LeftClick 05-15-2001 04:14:42

RE:FixUSB (modified 0 times) Programmer Profile it should work normally now as well.. I didn't have the extension .com registered as application/octet-stream on my web server.. and you didn't have it registered in your browser.. 05-15-2001 06:12:35

RE:FixUSB (modified 0 times) thbt Profile I've tried it, but haven't seen any differences yet. Programmer, how does it work? 05-17-2001 18:19:25

RE:FixUSB (modified 0 times) Programmer Profile it basically just does what is suggested on one of the via tuning sites. ;FixUSB ;load the Double-Word with the PCI Latency Address mov eax,08000000Ch mov dx,00CF8h out dx,eax mov dx,00CFCh in eax, dx ;Set the byte we are pointing to to 00H and eax, 0FFFF00FFh or eax, 000000000h out dx, eax ;load the Double-Word with the PCI Delay/Caching Address mov eax,080000070h mov dx,00CF8h out dx,eax mov dx,00CFCh in eax, dx ;Set they byte we are pointing to to ???? ?00?b And eax,0FFFFFFF9h Or eax,000000000h out dx, eax 05-18-2001 05:48:30

RE:FixUSB (modified 0 times) Programmer Profile thbt: does that mean that you are still experiencing network dropout or problems with USB speakers? or that USB is functioning properly to begin with and you were hoping for an improvement? 05-18-2001 05:51:45

RE:FixUSB (modified 0 times) thbt Profile Win98SE would crash or freeze on me when I accessed the network for too long. FixUSB didn't seem to help. =( 05-18-2001 12:00:58

RE:FixUSB (modified 0 times) Programmer Profile have you tested for this problem on other computers using your usb adapter? are you starting from a clean win98se install with up to date drivers? what you mention has never been a problem for me, so I'm not like to be much help.. The only problem I ever had with an IO locking up is when I was running it too fast and the power supply couldn't keep up with spikes; switching the power supply with another one that I had solved that problem for me.. 05-21-2001 06:29:23

RE:FixUSB (modified 0 times) jbarr Profile This may be obvious to most, but sometimes we overlook the simplest things... The instructions are to PUT THE PROGRAM INTO YOUR AUTOEXEC.BAT FILE. I neglected to do this. I only ran it once thinking that it would permanently patch things. It looks like this needs to be run on every boot. I'll let you know if it corrects some of my problems!!! 05-22-2001 09:26:01

RE:FixUSB (modified 0 times) jbarr Profile Hi! Quick followup... I installed the FixUSB.com into my autoexec.bat file, and here's the results: 1. Ran a looping batch that continuously pinged abother computer on my network overnight. Result: No problems to report 2. Set up an MP3 playlist of several songs on another computer on my network and played the playlist continuously and repeatedly overnight (with the sound turned down, of course!) ;) Result: MP3's still playing clearly and network activity still going. 3. Installed Microsoft Office 2000 over my network. Result: Took a while, but I now have a perfect Office 2000 installation on my i-Opener. 4. Browse 'net for a couple hours. Result: No problems, no worries. Conclusion: While this is definatly not a scientific test, it is a series of real-world stuff that all performed flawlessly and as expected. Seems my previous USB problems are a thing of the past... 05-25-2001 20:58:50

RE:FixUSB (modified 0 times) * StarFish * Profile From the Past,( I-Opener Technical Stuff / D-Link 650 USB ethernet .. does it work.. ) Unit_1 asks, Please try a test. Send a 10+Meg ZIP file FROM the I-O to another computer and then do a CRC check (using the "Test" option) on the file (while at the other computer) to verify that it arrived intact. also Unit_1's Post (5th from the End), is Very Good. Tells How TCP/IP might be a Fix ? I-Opener Technical Stuff / DLink USB Hub + 10/100 Ethernet 05-25-2001 21:44:42

RE:FixUSB (modified 0 times) jbarr Profile Thanks for the links. First, I took the jailbait image file that I downloaded and zipped it. This gave me a very nice 14mb file to play with. Next, I set created a batch file that that would move the file back and forth between my i-Opener and on of my other PC's. (Copy the file, delete the source, copy the file back, delete the source, etc. effectivly moving it back and forth.) I set it up to look 16 times. The result: I ran a zip "test" on the file and it tested out with no errors. No problems, very stable. All problems now seem to be solved!!! 05-27-2001 17:22:22

RE:FixUSB (modified 0 times) YouBecha Profile I was hopeing it would fix my warplink issue. nope. However it was interesting that the first time I rebooted the warplink utility did not see the card, yet windows did. So I unplugged it, rebooted again, once it was up and running I plugged in the card and it started working. But unfortunately warplink still drops the network after a couple of hours of use. Thanks for the try. http://www.geocities.com/mr_bubba_zanetti/ 05-28-2001 06:04:41

RE:FixUSB (modified 0 times) ckbone Profile I'm convinced the Warplink problem is not specific to the I-Opener, since they won't work reliably on other computers either. I think they draw slightly too much current from the USB connection on all computers, causing crashes, lock-ups, etc. Software USB fixes cannot solve this problem....if I'm right about this. 05-28-2001 06:17:24

RE:FixUSB (modified 0 times) YouBecha Profile I will try to not cross post here and on the warplink thread, but the fixusb thing seems to have stopped the lockups...it still isn't 100% but it is very useable now. 05-28-2001 13:33:37

RE:FixUSB (modified 0 times) Wild_Pencil Profile | Email Thanks for testing this patch, and definite thanks to Programmer for providing this USB fix. It will be included in the next BIOS upgrade I'm working on. Should be released in a week or two. Now that the Free-drive account is gone, I'll have to come up with another distribution method. -WP 05-28-2001 19:12:37

RE:FixUSB (modified 0 times) Mitchell Profile | Email YouBecha: My warplinks work perfectly. They have been connected for over a month with no interruption. In fact, the iopener hasn't been rebooted in over a month. Both machines (one iopener and one standard pc) both run win98se. I REALLY wish I could get the warplink working with win2k though. Tried it once and that was it. 05-28-2001 20:54:44

RE:FixUSB (modified 0 times) Rick Profile Niether the BIOS patch nor FixUSB worked for me. I am still not able to e-mail 200k attachments. Another thread mentioned a USB filter driver but the link is broken and I can't find the driver on the VIA site. I have a V3 I-opener with 128mb of Kingston ram and a Pentium OverDrive 200Mhz CPU. I am also using a D-Link 3 port hub with 10baseT eithernet adapter (DSB-H3E) with external power supply. And ideas? 07-16-2001 20:03:15

RE:FixUSB (modified 0 times) Rick Profile OK, I found the filter driver and installed it. I also downloaded the latest V5pmi.bin and flashed that. Still no good. 07-16-2001 20:33:10

RE:FixUSB (modified 0 times) Wild_Pencil Profile | Email Some suggestions.. Try it using your original processor, instead of the overdrive chip. Try it with the stock RAM plus original processor. Try it without the USB Hub (if possible). Try a different USB adapter, or an updated USB-Ethernet driver. I've found Linksys USB Ethernet Adapters (Pegasus chip) to work quite well. 07-16-2001 20:45:02

RE:FixUSB (modified 0 times) ckbone Profile Programmer....I finally tested the FixUSB patch. I waited till I did a clean Win98 install....all MS updates installed.....No VIA drivers and no MS USB patch. Only my 3-Com USB HPNA network drivers to test the LAN. Too bad.....I've got the same USB problems as before. Data is always corrupted going out of the I-Opener...even small files (2mb). Data coming into the I-Opener is fine......even large files of 125mb or more. I tried this data transfer testing on an external USB drive....same result. I then installed the 4in1 drivers v.4.32.....still the same problems. Do you recommend installing the VIA filter driver and the MS USB patch over your program? I am really disappointed about this....I thought sure a clean install and your FixUSB patch would solve my continuing USB problems.... 08-01-2001 16:09:43

RE:FixUSB (modified 0 times) *SF* Profile You might want to see If You can Remove All Protocals Except TCP/IP. That Is Rumored to Help ??? 08-01-2001 16:30:40

RE:FixUSB (modified 0 times) ckbone Profile The only problem with that fix is the network requires Net Beui (?) to operate. Also I can live with the network prob.....I am more concerned with corrupted data leaving the I-Opener, such as to a printer. My USB printer f***s up text pages. I cannot transfer files to any USB external drive....the files are usually corrupted. This has been an ongoing problem with all three I-Openers running Win98 and ME. This last Win98 install had no other stuff on it but the LAN drivers....still didn't work right. 08-02-2001 07:09:09

RE:FixUSB (modified 0 times) Programmer Profile did you put fixusb in the autoexec.bat? (i'll assume that you are using the dos .com file) FixUSB forces some settings in the VIA Hardware (PCI Latency among others) I'll assume that you are not using any other software that modifies these registers. The VIA filter driver (AFAIK) activates messages that are not not enabled on the UHCI USB implementation. http://support.microsoft.com/support/kb/articles/q232/8/89.asp The microsoft patch solves enumeration problems in their uhci driver http://support.microsoft.com/support/kb/articles/Q240/0/75.ASP I do not have the problem that you are reporting with either of my v5s or my v3. It should be safe to rule out power as the problem, as you have managed to transfer large files IN.. OK.. I did a search and turned up this page http://www.caprok.net/dave/FAQ.html which is about a board with the same chipset as the IOpener (even though it is not using it the same way), and it recommends disabling the write cache pipeline. http://bethie.net/~programmer/fixusb2.com force disables said feature. If this works for you, please let me know and I will include the update in my bios patch. 08-02-2001 08:32:20

RE:FixUSB (modified 0 times) ckbone Profile Programmer....it's installed correctly. I tried the FIXUSB2 prog....same result. My testing is not very scientific, but I think the results are valid. I transfer a known-good utility file.....try to install it. If it won't install or crashes the other computer, I know the transfer was corrupted. For a large file, I transfer a 125bmb video. I then watch it.....I can see very choppy video, distorted sound, sometimes have a crash trying to play it. It's also corrupted. No differences noticed between your two versions. Do you think I should try the VIA filter, and the MS USB fix next? I hesitate to do this, without knowing if they cause more trouble than they're worth. I did notice one thing.....I took a CPU performance hit with FIXUSB2 of almost 8%. I used the Sandra 2001te benchmark.....I have a Compaq desktop with the same chipset as the I-Opener. It has no USB problems to speak of. I really appreciate all your help with these strange problems. What's hard to understand why some people report no USB problems at all, while I have all kinds of USB problems with three I-Openers, V2, V4, V4b.....three different CPUs, Three different OS installs. Go figure.... 08-02-2001 18:35:42

RE:FixUSB (modified 0 times) Programmer Profile ckbone: it is rather surprising that you have the exact same issue on all of your IOs.. perhaps it is some configuration problem. (either that or you are the only one using netbeui) in any case, I have written a a program that should make it reasonably plain as to whether the problem is possible to be solved. It configures the north bridge to its default state with regards to PCI config. This is guaranteed to slow down your machine, but if the USB works under this configuration, we can start turning the enhancements back on until we find the one causing the problem. If the USB does not work while using this program, then it is unlikely that any northbridge configuration will ever help you, and perhaps we should look to the southbridge, IRQ configuration, etc.. anyhow let us know.. http://bethie.net/~programmer/ckbone.com 08-03-2001 07:44:27

RE:FixUSB (modified 0 times) ckbone Profile Thanks again....I may get to try it tonight. I just wonder how many people here actually use USB out for anything other than USB speakers, it seems to handle USB sound ok....perhaps I just don't hear the errors. I never used a USB printer till recently.....this canon printer prints fine over the parallel port, but over the USB port pictures have blank spots, text pages get garbled, etc. I have four different USB drives, CF reader, zip, orb, and super disk....they all return errors going out. Going in, they all work fine. My BIOS settings are mostly default....different memory timings make no difference. Now the IRQs are set to AUTO....USB and the Video use the same the same IRQ. I never could seem to separate the two.... We'll see..... 08-03-2001 11:54:16

RE:FixUSB (modified 0 times) Ragnar Profile ckbone, Are you using the printer straight thru the usb port or are you routing using a hub? My printer works fine when it is attached directly to the iopener but produces garbage when I route thru a usb hub, especially if I am using another usb device, even a mouse, at the same time. 08-03-2001 12:35:36

RE:FixUSB (modified 0 times) ckbone Profile Ragnar....Your inability to to print when attached to a powered USB hub is a symptom of the VIA USB problems which many VIA chipset motherboards exhibit. MS specs claim the ability to daisy-chain 127 devices...of course they were using Intel USB controllers when they designed the system. There's no reason you shouldn't be able to do this....except for the USB/VIA problem. I managed to separate the USB and Trident IRQs.....USB is now 11, not sharing it's IRQ. If anyone is interested, I can detail the procedure....it's tricky. I retested Programmer's patches with the same errors going out. The newer patch just posted(ckbone.com), resulted in an inability to boot, either normally or in the safe mode. I was able to boot using the step-by-step conformation mode, not loading the autoexec.bat. I then removed the new patch, and the system returned to normal. What a pisser...... There's an easy way to see if you have a USB problem....try sending a jpeg picture file, say 1 meg, out the USB port....either by LAN, or to an external USB drive. Then open up the picture in another computer.....I can't do it, the picture is always f***ed up. 08-04-2001 03:20:19

RE:FixUSB (modified 0 times) laserfan Profile I have followed this thread w/detached interest (hey, thankfully *I* don't have this problem!) and have thought all along ckbone that you have no doubt shot yourself in the foot w/all your upgrades & updates (my philosophy: if it ain't broke I don't try to fix it, at least not anymore...). But you have been SO helpful for SO long that when you suggested this simple test I gave it a try (and amazingly I do have an 1100Kb JPEG file to test with). So here goes: 1. Download the file from my main PC to my IO over my Netgear EA101 USB/Enet. Lickety-split. A second or two is all it took. 2. Open using IE and observe that it is a thing of beauty. 3. Rename it and send it back to my main PC. Again, only a couple seconds to xsfer. 4. Open it from my IO over the network, again using IE. THE IMAGE HAS GLITCHED-OUT ALMOST IMMEDIATELY! I.e. the top inch or so of the image looked fine then it turned to hash. My goodness! 5. Go to my main PC and open it w/Irfan Viewer (slick viewer btw). Same damn thing. How about that, I've got the problem too. Never noticed it before, never. Recently in fact I have been emailing lots of JPEGs to my family from the IO and they apparently have all worked just fine, but these files are typically less than 200K in size. So for the purposes I use my IO, email and web access, I have never before noticed this problem, nor am I inclined therefore to get too excited about it, but again ckbone as you have been helpful to ME I will look for future opptys to confirm or deny your attempts to solve this w/Programmer. My config is very simple: original v1 IO w/stock Netpliance BIOS, Netgear USB as above, 128Mb RAM, the WinChip2 upgraded mp, and of course an HD, running Win98SE installed w/98lite Chubby. Further, I have deliberately done MINIMAL dinking w/updates & upgrades, having gotten this thing to work PERFECTLY (or so I thought until today!) with video-blanking and hd-spindown and network access including remote VNC and IOLED On/Off (thanks Programmer!)etc. etc.: Trident CyberBlade i7 AGP v6.50.5482-10 VIA Disk Controller VATAPI.VXD v2.0.950.2110 VIA Tech 3038 PCI to USB Universal Host Controller drivers are Microsoft 4-23-99 v4.10.2222 automatic settings IRQ15, error detection disabled Again, since my IO is working *to my satisfaction* I will be reluctant to make any significant changes to it, but will otherwise try to help if I can! 08-04-2001 08:38:23

RE:FixUSB (modified 0 times) ckbone Profile He, He......So I'm not alone after all. Where I really noticed this was when I tried installing an image of a hard drive, to my new larger hard drive. The image was good because I recently crashed, and was forced to restore that image (which was on the hard drive). I burned the image on a cdrom, using an HP USB cdrw. The burn went ok.....But when I tried to put the image on the new hard drive.....the install always failed. However, when I got that same image over the parallel port there was no problem with the install. This is the VIA/USB problem at it's best (worst). 08-04-2001 09:59:18

RE:FixUSB (modified 0 times) laserfan Profile I have an image of my IO Drive C on another partition of the same hard disk. Your comments inspire me to consider: 1. Get that image onto another physical platter somewhere 2. Don't use my USB network device to make the image transfer Obviously the "only way out" of my IO is via USB. Guess I will consider/plan at some future date, when I have the IO cracked-open again, to take the HD out and back it up to another PC using an IDE cable. In any case, since I've said my IO is my primary surfing tool, I need to be wary of stuff I might have DL'ed on my IO and then transferred to my main computer, which is where I generally keep downloaded stuff for eventual backup to CD-ROM. BTW I do also have a USB standalone CD-writer, so I need to be wary of possibly using it w/my IO as well I guess (I haven't tried it to-date). 08-04-2001 10:30:39

RE:FixUSB (modified 0 times) ckbone Profile I've been aware of this problem for some time, but never it's full extent. ANY stuff moved off the I-Opener over the usb is suspect. You can use direct cable connection (parallel port) to get your hard drive image off onto another computer.....that works ok. 08-04-2001 11:53:43

RE:FixUSB (modified 0 times) YouBecha Profile I have been trying to make my Warplink USB wireless NIC work properly with the Iopener for quite a while now (because it was so cheap). So I tried all the VIA usb fixes and FixUSB etc. I recently picked up a Fujitsu 1200 and added a USB port to it. I have noticed that if I run the warplink on the Fuji with battery power only then I get the same problems the Iopener has. If I run the Fuji off AC power, then the warplink keeps on running just fine. What does this mean to you? I have to agree now that a lot of Iopener problems are due to a lack of power, I don't know if it is the brick, or more likely the onboard power supply. Hope that info helps your troubleshooting. http://www.geocities.com/mr_bubba_zanetti/ 08-05-2001 08:30:28

RE:FixUSB (modified 0 times) laserfan Profile Thanks for the tip. I have a parallel transfer cable that I haven't used for a while in favor of a much-faster USB cable, tho I haven't yet used either w/my IO. I had set-up this IO in "standard fashion" ie. loaded the HD w/everything I thought I'd need, including network drivers for my Netgear USB device, and since then have gotten everything for the IO off my network so haven't needed either of these alternative xsfer methods. BTW also as I have re-read this thread I too am using NetBEUI for my LAN i.e. client services and file sharing are bound to NetBEUI but not to TCP/IP. Wonder if this is at issue somehow... 08-05-2001 08:35:35

RE:FixUSB (modified 0 times) ckbone Profile More things to think about.....The power issue makes sense, as does the Netbeui thing. But both issues make no sense in that data transfer into the I-Opener is also affected by power and Netbeui....and this inward data over the USB is unaffected, and works fine. FWIW...I listened closely to an MP3 streamed off one I-Opener to another. You can hear the errors....clicks, pops, etc. If you're using Winamp, you can see the errors using the visual equalizer bar graph. I guess this audio corruption is interpreted as a different sound....and only shows up as a click or pop. I'm wondering if the USB problem could be hardware related....Does anyone who did the second USB port hack have these problems using that second port? We really need some feed-back.... Help us out guys, try that jpeg transfer test...I'm convinced this is a common problem to all I-Opener versions, regardless of the OS configuration. I'd like to see a fix here....All USB add-ons are affected...... 08-05-2001 09:03:35

RE:FixUSB (modified 0 times) El_Kabong Profile | Email I sent some MP3's to my IO using a Belkin USB to USB setup, then I sent it back to the desktop. Results: All files sent were slightly damaged by the transfer. Pops, phase distortion and artifacts were clearly heard and not present before the transfer. I'm using the Fix USB v5 BIOS and an AMD K6-2 333Mhz working at 300Mhz. The IO otherwise works great. 08-05-2001 22:34:42

RE:FixUSB (modified 0 times) ckbone Profile There a lot of previous hints to this problem falling into place. I remember a topic here started by a guy who could never get a Buslink USB hard drive to work properly....he couldn't write to it without crashes. I had him try suggestions on the USBMan website: http://www.usbman.com/ I don't think he ever got it working. I always wondered why when I transfered a MS update folder from one I-Opener to another using a USB device, I could never get the installer to work and always had to download the update again(blaming MS). Yesterday I tried to transfer a folder with 150 small jpegs out of an I-Opener. I can transfer each jpeg, about 50k, one at a time ok. When I transfered the whole folder at once, both computers crashed after a few seconds. On that USBMan site, under VIA suggestions, he concluded that some of the VIA problems might not be fixable....he recommended a new PCI USB controller card....specfically not VIA. All-in-all it doesn't look too good here.....maybe a software patch by Programmer will fix the problem....but it's beginning to sound like a hardware problem, and a difficult fix. 08-06-2001 03:39:43

RE:FixUSB (modified 0 times) laserfan Profile It occurs to me that Netpliance never, to my knowledge, announced any plans/accessories that used the USB port. Maybe they knew there was a problem? Anyone lurking here that knows any techs from NP, or the original mfr? 08-06-2001 04:46:52

RE:FixUSB (modified 0 times) Programmer Profile ckbone: you mentioned ( I Think) that you have another machine with the same chipset as the IO that is not experiencing the problems that the IO has with USB. It very likely is a configuration issue, and to that end I wrote a program that will extract the configuration info into a file so that we can look and see both what the good machine is setting up and what the IO is setting up (so hopefully we can find some differences and take care of them). http://bethie.net/~programmer/writecfg.com is a tiny program (96b) that reads the configuration of the northbridge and writes it to register.bin note: I have only tested this program from a dos box under win98. please run the program both on your machine that has USB functioning and on an IO and send the resultant files to me at programmer at bethie dot net and I'll take a look and see what's up 08-06-2001 09:36:52

RE:FixUSB (modified 0 times) Programmer Profile http://bethie.net/~programmer/fixusb2.com has been rewritten with the applicable data from the via hardware site. Give it a try and let me know.. 08-06-2001 11:34:56

RE:FixUSB (modified 0 times) dvfagan Profile | Email I have been running a stock V5 under WIN98 for two months as a home server. The cable modem access is via a USB <>10 Base T adapter. I have only experienced 3 WIN98 lockups. This is about average, I think. I also use a BAFO USB<>USB adapter to transfer files to/from a desktop PC. The Bafo software has locked up, but always when closing out the last file of a set. This appears to be a protocol violation of some kind, else why would it only occur on the last file? My cable modem access has been fine. 08-06-2001 14:05:20

RE:FixUSB (modified 0 times) Programmer Profile USB in bulk transfer mode does not offer any sort of validation that the data arrived successfully. you're basically just not supposed to HAVE these kind of problems. Things that use bulk transfer: USB Hard Drives USB CDRoms, and Burners Zip Drives USB Ethernet Controllers Some USB Link devices Things that don't mouse keyboard things that absolutely need to be validated (according to manu) The Traffic INTO the IO is fine because the buffers are not allowed to become stale before the data makes it's way into memory The Traffic OUT of the IO can be invalidated if it sits in the PCI buffer for too long before being sent on the PCI BUS. This is what has caused hard drive data corruption on the 686B, and is very likely what is causing our USB data corruption. The TCPIP protocol has it's own packet verification mechanism, this is why TCPIP can reliably transfer files. Bulk IO to the USB port, followed by packet transfer accross the network followed by validation and resend. Latency patches are useful in making sure that devices that are not responding in an appropriate amount of time do not waste system time. This can however cause problems with software that makes no effort to ensure that it's data was delivered successfully as if the device times out before the data was sent the data does not get sent. The only apparant way to get the flaky hardware to work properly is to find the exact balance it needs between load and latency to work flawlessly. Ideally the manufacturers of things like USB hard drives and such would give optional drivers to use when data validation is a MUST. 08-07-2001 08:10:13

RE:FixUSB (modified 0 times) laserfan Profile Verry inneresting, Programmer.... To the point about verification, I had thought before you posted that I would try to switch my LAN protocol from NetBEUI to TCP/IP to see if that changed anything. Honestly I am a Windows networking neophyte and only have my system set-up as I do because of Steve Gibson's Shields-Up site that suggests to NOT bind "file sharing" to TCP/IP cuz this creates a potential opening for hackers. So I'm bound instead to NetBEUI per his instructions. Is there any reason why trying to run my LAN over TCP/IP will NOT work i.e. am I wasting my time trying to re-configure my LAN to use IP so that my file xsfer packets are "verified"? As I re-read your post, it suggests to me this might EXACERBATE the problem because if TCP/IP w/verification turns out to be slower than NetBEUI, then the backup from whence the corruption occurs becomes even worse. Does any of this make sense? Again, I'm not well-versed in networking so be gentle... 08-07-2001 12:28:57

RE:FixUSB (modified 0 times) ckbone Profile Programmer's explanation is the first info that makes sense about this USB problem. I tried the latest fixusb2 patch....still failed the 1.1meg jpeg transfer test. Try it yourself and see....the top 1/2 of the picture is ok, bottom is messed up. Mp3 streaming out of an I-Opener did sound better. I couldn't hear any pops this time. I managed to transfer a 25mb video without a crash.....but playing the file showed errors, not so many as the last time I tried this. Maybe there is some hope here. I just sent programmer the two test files.....If he can pull off this USB fix, well, I'll be astonished......you know there are countless persons who haved tried to fix the VIA/USB problem, including VIA itself...all without success so far. 08-07-2001 18:40:19

RE:FixUSB (modified 0 times) Starman97 Profile Are you running Win 98? If so, that would explain your USB problems, the Win98 USB stack is horribly broken, Win2K is better, but far from correct. One of my partners at work wrote a Win98/2K device driver, there are some very serious bugs in the Microsoft code, most devices instruct you to pull the USB cable if any problems occur, MS cannot recover from missed packets, they leave file handles open on USB resets and other major mistakes. bad,bad,bad MS... We have a CATC USB Chief+ analyser, wonderful device, shows where the errors happen when USB flakes out, it's always Win98. 08-07-2001 20:39:00

RE:FixUSB (modified 0 times) laserfan Profile I am indeed running Win98SE. I nosed-around with this a little more tonight. Grabbed the MS Universal Host Controller "fix" for VIA chipsets (uhcd.sys v4.10.2223) and this actually made the problem WORSE! Upon attempting to send a file from the IO, it would sometimes stop in mid-transfer and alert "network resource no longer available". Other times the file would get sent without a stoppage, but the file (1100Kb) was always hosed. I reverted to the original 4.10.2222 version of uhcd.sys and it worked "better" again (unless of course you consider that a file that was transferred very quickly, but with unannounced errors, is not actually "better"!!!). I had also unchecked the "Disable error checking" box (i.e. I ENABLED error-checking) on both computers w/no help there. This is certainly a turd in the punchbowl isn't it, as it calls-into-question all uploads from the IO to the network... 08-07-2001 21:42:30

RE:FixUSB (modified 0 times) Programmer Profile laserfan: The only reason not to use file sharing over tcp is to protect windows from intrusion. If your internet access is secured through other means, then there's nothing to worry about. (i.e. my internal network at home uses TCP only, and has file sharing etc enabled; my external network (which only connects to my dsl line) has a seperate lan card in my server and will not accept "file sharing" connections.) Using tcpip may take longer depending on how hosed your system is USB wise, but it will verify the packets and ensure that there is no data loss. Using TCPIP will not however solve any of the other USB problems, so after I look through the data that I've been sent, I'll hopefully have some (good) answer.. 08-08-2001 06:57:54

RE:FixUSB (modified 0 times) Programmer Profile OK.. http://bethie.net/~programmer/ckbone.com is now a file that sets up the Northbridge configuration the way that compaq sets up theirs... :) Thanks ckbone!! ckbone: please give this file a try.. it's a whopping 151 bytes.. ouch.. if it works I'll rewrite it to be a bit smaller and set it up in the bios patch. 08-08-2001 08:10:14

RE:FixUSB (modified 0 times) zyxw Profile | Email Hey, You're great! Do you mind one request... can you try to patch in the "del" key patch as well... I use your patches (don't need the CD ROM support) and it'd be real nice to just hit del on boot... thanks! 08-08-2001 09:22:33

RE:FixUSB (modified 0 times) Programmer Profile The Del key patch is Wild_Pencils.. I haven't even looked at it, and I am not willing distribute a file that I have been expressly asked not to distibute. If you want the del key patch (and whatever else Wild_Pencils latest bios contains) and are not interested in having BadFlash reprogram your bios, you could always send Wild_Pencil an email asking for the bios and agreeing (not to distribute it openly) and (not bug him with simple questions) In any case, the bios that is on my site will be updated once ckbone verifies that we have a solution and will be v5Patched + memory interleave + fixusb + video mode switch, and I'll have to assume that Wild_Pencil will put the updated code into his BIOS release, and distribute it to Badflash from there. 08-08-2001 10:54:05

RE:FixUSB (modified 0 times) Lokrien Profile OK... don't kill me for this, but... from what i have tested, my usb adapter works fine. I DO have netBeui on the adapter (linked to my network of god knows how many #@&% computers!) and tried moving several large files TO another computer. I tried a zip executable of 4.3 meg size -- it unzipped and installed correctly on the other computer, a large (1.3 meg) bitmap which came out ok, 3 sweet movies (you know the kind) and they played just fine! I even went so far as to run them across the network on the iopener, no probs. I have: netgear usb network adapter, winchip2, 128 meg ram, V1 iopener, 6 gig Toshiba hd. NOTE: this wasn't not an extensive test by any means, but so far, all is cool. I am surprised by this as so many seem to have issues with their iopeners using the usb networking. I had recently thought of rebuilding my setup, but i think that I just may leave it as-is... ie. working. 08-08-2001 16:17:06

RE:FixUSB (modified 0 times) laserfan Profile Same as mine--do you have the EA101? Adapter sw version from 7-16-99 v2.18.0008.0000? Win98SE? Do you show 2 each of the USB Root Hub and VIA Tech 3038 PCI to USB Universal Host Controller? (See Device Mgr). Inquiring minds want to know. 08-08-2001 19:56:19

RE:FixUSB (modified 0 times) Programmer Profile Lokrien: could you run http://bethie.net/~programmer/writecfg.com on your IO and send me the file it creates.. it only takes less than a second to download and run.. :) .. and are you sure that you are not inadvertantly using tcpip as the transport?? just having netbeui bound will not make it used. 08-09-2001 06:20:23

RE:FixUSB (modified 0 times) ckbone Profile I haven't had time to do a complete test, but the latest fix patch still has problems.....won't pass the jpeg transfer test over the lan, but it transfered ok using a USB CF reader( never did before ). I tried it with and without the 4in1 v4.32 drivers installed.. no difference. I'm wondering if the Yamaha drivers could be a problem. I know SB Live drivers mess up IDE data transfer on a 686b chipset. The I-Opener has a 686A (I think). It's worth a try. Unless many people aren't doing the testing right, some people, like Lokrien, apparently don't have the problem....maybe we're overlooking something easy/obvious...I'll try more tests tonight. I'm running Yahaha drivers on all my machines......And we know an I-Opener with a RISE chip won't boot with those drivers. Lokrien.....do you have any 4in1 drivers installed? Which ones? Also, have you applied the VIA USB filter driver or any other VIA patches? 08-09-2001 09:54:49

RE:FixUSB (modified 0 times) Programmer Profile ckbone: I just updated http://bethie.net/~programmer/writecfg.com to read the USB configuration from the southbridge as well (file size now 120bytes <sniff>).. please run it on your compaq and on one of your IOs (after running ckbone.com ) and mail the files to me.. 08-09-2001 11:03:46

RE:FixUSB (modified 0 times) YouBecha Profile BTW my linksys USB 10 nic works absolutely fine...it is just the warplink that causes USB trouble. http://www.geocities.com/mr_bubba_zanetti/ 08-09-2001 14:55:42

RE:FixUSB (modified 0 times) Lokrien Profile yes Laserfan, same setup as you there. Programmer, ran and sent. lemme know what you find... 08-10-2001 01:22:35

RE:FixUSB (modified 0 times) Lokrien Profile ckbone: my via agp driver (viagart.vxd) reads as file version 2.90.0 I really don't remember when was the last time i updated the driver. in the viagart.inf file, driverVer=08/10/2000,5.00.00.0404 I always do the full driver set, so this should be the same as any other drivers for the chipset. I ~may~ have put the filter drive on. I put it on one of my machines (a few have Via chipsets) and as the USB is so important on the iopener, it was probably on it... but i cannot be sure. i am gettin' too old for this! any more questions... i will check back as work allows. check the time i am posting... this is pretty regular for me. <sigh> 08-10-2001 01:34:34

RE:FixUSB (modified 0 times) laserfan Profile My VIA stuff in Device Mgr/System Devices is from Microsoft. A search for "via" turned-up: viagart.vxd in \windows\system\ and it is v4.04.000 viapfd.sys in \windows\system32\drivers\ and this one is 5.00.2195.100 There are also three INF files: The viagart.inf has Lokrien's same reference "DriverVer=08/10/2000,5.00.00.0404"). The viavsd.inf references DriverVer=03/26/2001,2.0.950.2110. Viamach.inf shows DriverVer=08/01/2000,5.00.00.0120. I don't recall at this moment what I installed--but maybe since Lokrien and I appear to have the same hardware setup a comparison between us might be key? I will send that file off to Programmer (who doesn't have the problem but is trying to help us!!! Thanks!). 08-10-2001 05:06:17

RE:FixUSB (modified 1 times) Programmer Profile Ooops.. The writecfg.com that is up there is not giving me any more information than it was before.. I need someone to look in the registry on their IO at hkey_local_machine/enum/pci/ven_1106&dev_0686????????????? and tell me what the folder under it reads.. should be something like BUS_00&DEV_??&FUNC_?? hopefully, you can get something to me by around 2pm central today or I will not mbe able to work on this again til monday.. 08-10-2001 06:05:42

RE:FixUSB (modified 0 times) Programmer Profile OK http://www.bethie.net/~programmer/writecfg.com is ready to be re-re-re-downloaded and will now give me the usb configuration info from the southbridge as well as the information ferom the northbridge. Right now I am primarily interested in the data from ckbones Compaq and IOpener(after running ckbone.com). Lokrien: The data from your northbridge is very similar to everyone elses. could you redownload and rerun writecfg and send me the file so I can get a look at the data from your southbridge. 08-13-2001 07:03:02

RE:FixUSB (modified 0 times) Kcmjr Profile I've been reading this thread with interest because I too have the formentioned USB problem. Large files transferred over the USB NIC cause the "network resource no longer available" message and corrupted files. I wonder why no one has attempted to use IPX instead of NetBEUI? If TCP/IP and NetBEUI are having problems IPX/SPX is a time tested protocol that while similar to TCP/IP has many differences. NetBEUI is nice because while chatty and non-routable it's auto configuring and fast. Perfect for small home networks. IPX requires some setup like TCP/IP but the differences in the protocol structure might be enough to help with this issue. Only one way to find out...... My setup is as follows: V5 with BadFlash v5 BIOS, IBM Travelstar 20GB hard drive on an IMOD 3 board, 128 meg MICRON SODIMM, AMD K62+ with BadFlash cooling mods, LinkSYS 100TX USB NIC (w/NetBEUI), Running Windows 2000 Pro with SP2. The end result is I'm running stable and cool at 300 Mhz and mostly happy as a clam except for the USB and one stupid problem with the modem refusing to dial out occationally. The modem thing hopefully will go away when I install the right drivers tonight. I've connected to both my desktop machine and my laptop via a rolled ethernet cable with the same results. I hope to get the second USB port fully installed later this week and can test the connectivity on it then. I started to install the port using the RoastBeef mods but dropped one cap on the carpet (I hate it when that happens) and just got the replacement. I'll also try to set up with IPX/SPX and see if that makes a difference. I'll post my results later this week. Since we're on the subject of USB, does anyone know if the polyswitch used with the USB ports on the IO is polarized? I'd hate to have put it in backwards. I'm curious about Programmer's BIOS patches, I noticed some message about a BIOS patch during bootup but never knew what it was. Is what I'm seeing one of these patches riding on top of the BadFlash BIOS? KCM 08-13-2001 13:50:15

RE:FixUSB (modified 0 times) ckbone Profile Even if you have a USB LAN setup working ok, there is still a potential problem with anything else sent out the USB port, say to a USB zip drive, or burned on a cdrom. This USB problem may be fixed by one Programmer patch, which hopefully will fix all of our USB data corruption problems. Programmer.....do you need any more test files from me? I have two other I-Openers I could run the test on. You also might be interested in some info off this link: http://bbs.pcstats.com/viahardware/messageview.cfm?catid=19&threadid=5420 08-13-2001 18:05:51

RE:FixUSB (modified 0 times) Programmer Profile Thanks for the herads up on that thread.. it's pretty cool to see that some other people are making use of the stuff I wrote.. and giving me credit even... :) 08-14-2001 06:36:19

RE:FixUSB (modified 0 times) QiJaH Profile I couldn't have done it without your help. If you have any comments or suggestions, please let me know 08-14-2001 13:39:46

RE:FixUSB (modified 0 times) Programmer Profile OK New for testing: http://bethie.net/~programmer/ckbone.com does everything it used to PLUS, tell em johnny, A NEW TOASTER OVEN!!!!!!! well OK maybe not. but it DOES change the value of the USB latency timer, disables babble on whatever , and will not release a continuous REQ until it finishes.. (hopefully it'll make USB work right.) an interesting note: ckbones compaq has an OLDER revision of the chip we have. you'd think they would've IMPROVED things. oh well lemme know how it works 08-15-2001 11:09:15

RE:FixUSB (modified 0 times) zyxw Profile | Email Hi, Kewl, can we get a version of this in bios form? Also an aside, is there published information on modifying that is writing these code additions into the bios? Compgeeks is selling for $40 a 2.5" to USB enclosure/adapter (allows a notebook ATA/IDE to plug into an external enclosure/adapter and be usb connected to a host), if there is some documentation on how extensions are done, it'd be nice to try to map the usb devices and see if booting from a usb ide drive or cd rom would be possible. Not sure I'm up to this, but it'd be nice to read and maybe code play a bit. Thanks 08-15-2001 11:44:49

RE:FixUSB (modified 0 times) ckbone Profile I hope I'm not the only person trying these different USB patches by programmer....after all, I'm not the only person with USB problems, and just because they haven't worked for me yet, well, they might fix your problems. For those of you who don't know how to run these patches, here's how: Download the patch (the latest is ckbone.com) Put it in the root directory of the C drive. Click RUN, type msconfig, click ok. At the msconfig screen, click the autoexec.bat tab. Click edit, type a line that reads ckbone.com. Click apply, click ok, reboot. The patch is then applied when Windows starts up. Check different USB things, sending files, printing, writing to USB drives, etc. Let us know how it goes...... Programmer....I've tried the latest patch....still fails the jpeg file transfer test. Transfering a bmp gives a much better picture, but with some errors. I tried the patch on a bare Win98 install, with and without the 4in1 drivers (4.32 and 4.29), with and without the VIA USB filter driver, and without netbeui ( I managed to get my USB HPNA LAN working without netbeui....it works with ipx/spx, although much slower). If I'm doing something wrong here, could someone let me know? I've also got my USB controller not sharing it's IRQ, and have tried many different BIOS settings, including the default settings. I'll try all this stuff again tomorrow on a different I-Opener. The one I tested today is the V4b, K62450+ @200mhx, Win98se. 08-15-2001 18:09:15

RE:FixUSB (modified 0 times) laserfan Profile While it slices & dices, and may yet win the war for the Allies, it doesn't help my USB xsfers, unfortunately. Sigh. I appreciate that Programmer and ckbone are expending significant amounts of energy on this, but it appears to be a lost cause doesn't it. 08-15-2001 19:53:25

RE:FixUSB (modified 0 times) Wild_Pencil Profile | Email As far as I can tell, I've never encountered any USB problems on my machines - even before USBFix came along. Not to say that I'm problem free -- just that I've never pushed my IO's as thoroughly as you all have. (BTW, I gotta say.. ckbone's memory-tweaking guidelines are VERY thorough! ) Question. Are we all certain it's a USB buffer overrun problem, and not caused by poor voltages when the IOpener transmits on the USB, or by using poorly shielded USB cables or whatnot? Also.. do "normal" machines (without overclocking,extensive hardware mods, or memory-timing tweaks) display the same behavior? I seem to recall that ckbone has some really tweaked out settings on his machines.. maybe this is related? -WP 08-15-2001 20:42:31

RE:FixUSB (modified 0 times) Kcmjr Profile My IO isn't being tweaked heavily, just a memory upgrade and CPU swapout. The CPU is running at 300 Mhz, I haven't tried to push it yet. The USB cable that came with my LinkSys adapter looks to be well shielded. The cable is clear plastic with what I assume is either tin plated copper or steel braid under it. I also tried IPX/SPX with no luck. I'm curious, everyone seems to be using these patches with Win98, will they work with Win2k? I'd like to know for sure before I try them. It does seem that small files transfer fine, even large numbers of them. The problem is when sustained transmissions occur due to large files so the voltage thing might be on track. I just copied a bunch of files over the LAN connection adn they seemed to go OK. Some were PDF's as big as 2.5 meg. Those look OK but it took a couple of tries to get them across. KCM 08-15-2001 22:13:15

RE:FixUSB (modified 0 times) *SF* Profile WP > Question. Are we all certain it's a USB buffer overrun problem, and not caused by poor voltages when the IOpener transmits on the USB This Problem is with the VIA Chipset , not just I-Openers. I have a few Epox motherboards with VIA Chipsets, they also have this problem. 08-15-2001 23:54:12

RE:FixUSB (modified 0 times) ckbone Profile Concerning testing....These tests I've been running have been on an un-tweaked clean install of Win98se. I've got most of the updates off the MS update installed. No VIA drivers or patches for the first tests, if that test fails I try the patch with the various VIA stuff. I keep an image of the first clean install on the hard drive...and I always restore that clean image before doing any new testing. So I don't think it's the OS that is at fault. This install has no other programs except Sandra for benchmarking, the LAN drivers, and a small image viewer prog. FWIW...I've tried the patches on my two other machines....both tweaked....with identical results. I also wonder about the power problem. I just bought two power supplies with more amps (2.64). When they arrive, I retry the tests. Of course it could be some internal power regulation problem...very hard to pin-point, very hard to fix..... 08-16-2001 03:53:36

RE:FixUSB (modified 0 times) laserfan Profile Re: StarFish comment: >poor voltages when the IOpener transmits on the USB My Netgear hangs just a foot off the back of my IO, but then it traverses 25' of CAT5 cable to my Router/Switch. It also has 3 LEDs on it. I dunno enough about USB & Ethernet but imagine this does require some juice to xmit (versus receiving from the switch, which always appears to work OK). Maybe we are indeed barking-up the wrong tree looking at the software only. I'm going to try a shorter cable into my switch, heck maybe even a xover cable to my PC/Server if I can find one, to see if that makes any difference at all. 08-16-2001 05:30:21

RE:FixUSB (modified 0 times) laserfan Profile Oh, and my IO is otherwise unadorned w/gidgets/gadgets so it will be trivially easy to physically move won't it!! :) 08-16-2001 05:32:39

RE:FixUSB (modified 0 times) *SF* Profile Some AMD-K6®-2, Athlon, and Duron microprocessors cause JPEG images to be incorrectly displayed in some applications. When you view a Web page containing a Joint Photographic Experts Group (JPEG or .jpg) image file in Microsoft Internet Explorer V4 or V5, the image appears distorted. Although the image size and position is correct, most of the pixels are displayed using incorrect colors. This problem is also seen when you use a JPEG image in an Office 97or Office 2000 document. When the option to view the Microsoft Windows Active Desktop as a Web Page is selected, JPEG images appearing on the Active Desktop are also distorted. \ http://www.gordonfamily.com/AMD/ Thanks go to QUE for finding this Interesrting Page. I haven't found a Bad CPU yet. 08-16-2001 05:42:00

RE:FixUSB (modified 0 times) Programmer Profile Lokrien: Please run the latest version of writecfg (redownload it) and send me the output.. you have a different revision of the chipset than ckbone. Laserfan: Please run the latest version of writecfg (redownload it) and send me the output.. you have a different revision of the chipset than ckbone. (and the same as Lokrien) StarFish: .. you have the same revision Northbridge as ckbone, you are experiencing the same issues? send me a file.. 08-16-2001 09:32:24

RE:FixUSB (modified 0 times) Programmer Profile starfish: of all of the images I have only yours does not report having seen a bus parity error of some sort (including ckbones compaq).. what's your configuration??? 08-16-2001 09:46:08

RE:FixUSB (modified 0 times) Programmer Profile I just updated http://bethie.net/~programmer/ckbone.com again, it now adds some stuff for sdram timing control; forcing the chipset to talk to dram as 100Mhz, bumping up drive strength, and slowing the dram refresh. 08-16-2001 10:10:34

RE:FixUSB (modified 0 times) Programmer Profile New to Bethie.net!!! http://bethie.net/~programmer/data.zip contains 2 files 0.bin 8242880 binary zeros 1.bin 8242880 binary ones If you are experiencing network difficulties, maybe we can find a pattern; Download data.zip (2258 bytes) extract it on the IO make a folder on the destination machine named Data copy the 2 files into the folder zip the folder If the resultant zip file is NOT 2258 bytes (or conceivably very close to it) then all is NOT right in wonderland. (send me that zip) If the resultant file IS 2258 bytes, then try it again, if it continues to work fine then the chipset doesn't have problems transfering large amounts of 0s or 1s.. 08-16-2001 11:30:45

RE:FixUSB (modified 0 times) laserfan Profile I sent you Registe.bin based on writecfg.com of 13-Aug, and just tried the experiment with the bin files. I unzipped the data.zip file on the IO first, which resulted in files of 1,048,510. Then I transferred the two files to Data folder and zipped it. When I copied it back to the IO and unzipped it again, the data files were still 1,048,510 in size. But the data.zip file is a different size (2,660 bytes vs 2,258). I thought this might have something to do with: 1. Compression prog/algo used (mine is PowerDesk, what do You use?) 2. Pathname & other options (I did try putting the Data folder on the remote PC at the root c: level, also switched-off "relative path" etc.) 3. 4 different levels of compression--which to use... So I'm not sure this test is valid but will send you my 2,660 byte file nonetheless. BTW why would we test the xmission of 0s & 1s anyway? Zeros and ones, who uses THOSE? Let's get real! Sheesh. [Just what DO I enter to put a big fat smiley face here?] 08-16-2001 17:33:53

RE:FixUSB (modified 0 times) Programmer Profile oho monseur LaserFan, zee test is perfectly valid.. .. it says that the 2 files did not get corrupted in their transfer.. The reason behind all ones and all zeros, is that they compress REALLY well and are easy to compare and test. Have you done ye olde jpeg test while using the last ckbone.com? (you're actually 1 version behind on ckbone.com) The only odd thing left I see with your system is the fact that irq routing for the USB posts shows up as disabled. ckbone: The winchip machine is the same as laserfans, IRQ routing is disabled, otherwise everything (except memory) will be the same on it (after running the latest ckbone.com) as your compaq. one tiny wrench in the works is that even though the northbridge is the same on the compaq and the IO (laserfans and your winchip machine) the southbridge on the IO is a newer revision. Still, givin the amount of changes, I would think that there's not much left to do if an identical configuration works for one machine and not another. I am very interested in the results of the 0 and 1 file transfers on your machine with out any patches loaded.. If they transfer fine, then I think I'll have to make them a bit bigger.. I have just added a change to http://bethie.net/~programmer/ckbone.com which changes #devsel timing to fast as it is on the compaq. and a change to when and how the AGP writes data. 08-17-2001 06:27:11

RE:FixUSB (modified 0 times) *SF* Profile > starfish: what's your configuration??? V4b, AMD K6 III 333 (3x66=200), 128meg, 12gig HD, DSB650 D-Link USB ethernet adaptor (clear) 08-17-2001 08:46:31

RE:FixUSB (modified 0 times) ckbone Profile Jeez....I just tried the latest patch. No improvement over the LAN. The jpeg is corrupted, and when I transfer a small 4mb program to another computer, it will not install...a pop-up screen says the files are corrupted. I'm going to try a transfer by several USB devices I have....I'll also attempt a cd burn of an image file to see if it will install. I haven't tried that since the start of all this. FWIW...I tried a new power supply, 2.64 amp, didn't help. I tried the new patch with and without the 4in1 drivers...no difference. It's hard to believe after all the work programmer did on this, the LAN still is screwed up.....all the while my POS Compaq works just fine with the original BIOS, no patches, OEM Win98se. Do any of you guys who have a LAN working ok, use any special BIOS settings? I've tried every different way I know about, without success. 08-17-2001 17:04:40

RE:FixUSB (modified 0 times) Kcmjr Profile Something just occurred to me, I was adding the componants for the second USB port and a few other final mods when my IO stopped working. Took me a while to find the problem, by removing R130 and adding the SW1 switch the system stopped booting. It's probably a minor glitch and I'll search the board for answers after I send this reply but during the troubleshooting I thought that perhaps my lousy soldering job on the USB #2 polyswitch might have caused problems. I removed it at the expense of the polyswitch and will now need to get another since it wasn't the problem. Here's my thought, perhaps the USB #1 polyswitch is too low in value. If it's tripping during large transfers because of too high a current draw I assume it would effectively reset the USB device and would look like an odd dropout. Has anyone tested this? My IO is still in pieces and my wife is getting impatient to get it back, I doubt I'll be able to test this out unless I can do it from outside the case. Anyone have any thoughts? KCM 08-18-2001 23:50:18

RE:FixUSB (modified 0 times) Kcmjr Profile Something just occurred to me, I was adding the componants for the second USB port and a few other final mods when my IO stopped working. Took me a while to find the problem, by removing R130 and adding the SW1 switch the system stopped booting. It's probably a minor glitch and I'll search the board for answers after I send this reply but during the troubleshooting I thought that perhaps my lousy soldering job on the USB #2 polyswitch might have caused problems. I removed it at the expense of the polyswitch and will now need to get another since it wasn't the problem. Here's my thought, perhaps the USB #1 polyswitch is too low in value. If it's tripping during large transfers because of too high a current draw I assume it would effectively reset the USB device and would look like an odd dropout. Has anyone tested this? My IO is still in pieces and my wife is getting impatient to get it back, I doubt I'll be able to test this out unless I can do it from outside the case. Anyone have any thoughts? KCM 08-18-2001 23:51:23

RE:FixUSB (modified 0 times) Lokrien Profile OK. i took off TCP/IP protocol and now i get errors! %&$^!!!! I guess it was using tcp/ip for my network (never seen that before on a netBeui network... so news to me!) I had a nice large photo that i knew in detail! oh yeah! didn't even get over 1/4 of it without errors. so.. i am in same boat with everyone else. <sigh> 08-21-2001 16:12:31

RE:FixUSB (modified 0 times) ckbone Profile Lokrien.....Thanks for the new info. I was wondering how you got around the problem, guess you didn't. There are still some people who claim no problems with USB. If that's true, there has to be an explanation. There is a Badflash wiring mod that enables more internal power...it's a simple fix detailed in the Technical reference section. Could that be the answer? How about V5s....do they have this USB problem? Any more ideas on this subject? FWIW....I tried burning a 400mb image of my V2, using an HP USB cdrw. After the burn was over, the burn prog reported errors. I know what that means. I picked up a small computer at a show last weed-end. It has an Intel 810 chipset....It works fine on my LAN, but the I-Openers can't transfer anything to it either, without errors. 08-21-2001 16:38:42

RE:FixUSB (modified 0 times) Programmer Profile Has anyone used my transfer test and had it come back UNSUCCESSFULLY yet?? The information contained in the corrupted file make help me make a better educated guess about what to try next.. so PLEASE someone send me a corrupted result from my test. ckbone, copying it to and from a cf or zip drive would qualify as well.. 08-24-2001 08:54:55

RE:FixUSB (modified 0 times) ckbone Profile Jeez....I need help here. I downloaded the latest test files ok, unzipped ok, transferred to the Compact from an I-Opener ok. However, these two test files then refuse to Zip back up..They remain the same size after the zip....Am I doing something wrong? I tried Zip magig and WinRAR. My head hurts...... 08-25-2001 08:35:53

RE:FixUSB (modified 0 times) Programmer Profile nope.. if they're still very large after zip, that's exactly what I want.. please send them on over.. 08-27-2001 07:49:05

RE:FixUSB (modified 0 times) Programmer Profile well.. the files that you sent me were identical to the ones I posted.. which, so far I made a much larger file (10MB) at http://bethie.net/~programmer/data.zip It is not simple 0's and 1's, rather this time it is made up of the 0 to 255 followed by 0 to 60 which provides us with an easily compressable but not repeating(blockwise) file This should be more similar to a real-world test, and just as easy for me to validatate. so, If you are interested in maybe finding the problem, download the file to your IO, unzip it copy it somewhere zip it and email it to me.. note: the download is only 45Kb, but the unzipped file is 10Mb, so make sure you have room for it.. 08-28-2001 08:09:12

Let's kill this thread. (modified 0 times) Georgie Profile This has grown into a very long thread. It's time to kill it, isn't it? Programmer, Ckbone and all the others, You are cordially invited to clean up the small remaining details and finalize the solution to this realy nasty USB problem. Here is the new thread, I'll see you all there: http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2352&Idle=0&Sort=0&Order=Descend&Page=0&Session=georgie.99901550071588 08-28-2001 09:35:43

RE:FixUSB (modified 0 times) laserfan Profile Programmer said (some time ago now): >The only odd thing left I see with your system is the fact that irq routing for the >USB posts shows up as disabled. >ckbone: The winchip machine is the same as laserfans, IRQ routing is disabled, I don't know what this means. Where does one find this info about one's config? Another question I have is: can anyone suggest why I have two (count 'em) USB Root Hubs and 2 Host Controllers? Noticed in Device Manager, and then/now WCPUID shows I have 2 UHCI USB Controllers: Bus Dev Fnc VendorID DeviceID Revision SubSystem IRQ Device Type --- --- --- --------- --------- --------- --------- --- --------------------- #0 #0 #0 1106h 0501h 03h 00000000h - Host Bridge #0 #1 #0 1106h 8501h 00h 00000000h - PCI-PCI Bridge #0 #7 #0 1106h 0686h 1Bh 00001106h - ISA Bridge #0 #7 #1 1106h 0571h 06h 00000000h - IDE Controller #0 #7 #2 1106h 3038h 0Eh 12340925h 15 UHCI USB Controller #0 #7 #3 1106h 3038h 0Eh 12340925h 15 UHCI USB Controller #0 #7 #4 1106h 3057h 20h 00000000h - Host Bridge #1 #0 #0 1023h 8420h 5Ch 84001023h 15 VGA Compatible This strikes me as odd for a device that has only one USB port. Anyone understand why this is the case? 09-02-2001 20:26:48

RE:FixUSB (modified 0 times) Programmer Profile The system has 4 USB ports and 2 host contollers, it is only the matter of having 1.25 exposed that causes consternation.. The available port is the first port on the first controller, the unpopulated port is the second port on the second controller.. Any time 2 PCI devices share an IRQ IRQ routing will be used. 09-04-2001 06:47:35

Reply to Thread | Printer | All times are PST Powered by UltraBoard v1.62

I-Appliance BBS General - NEWS, Breaking NEWS - linux-hacker.net BBS on CD Questions - User Groups in Your Town - E-Bay, Yahoo Auction LINKS ONLY (MODERATED) - Jokes & Humor - usdtv box I-Opener Areas - I-Opener General Posts - I-Opener Technical Stuff - I-Opener Technical Info Reference (MODERATED) WebSurfer Areas - WebSurfer General Posts - WebSurfer Technical Stuff - WebSurfer Technical Info Reference (MODERATED) Other I-Appliances - Virgin Webplayer - WebPal - MailStation - 3Com Audrey - ThinkNIC / Oracle NIC - Gateway/AOL Connected Touch Pad - Philips/AOL-TV - ACER NT-150 STB - MSN Companion - Compaq IA - Emachines IA - IBM NetVista IA - Intel dot.station - MSNTV2 Network Computers - Fujitsu-Siemens Scovery Linux Net Computers - GallantWeb Server Consoles - Dreamcast - xbox hacking Web Pads - Epods One General Post - Panasonic CF-01 - Fujitsu Stylistic 1200 - Cybiko Thin Clients - Thin Clients DigitalVideoRecorders - DTV Recorders, Replay, TiVo, Etc... - Tivo Hacks Linux - Midori Linux General Board - ITX MotherBoards Single Board Computers - Earth Computers mARMalade Elite Member Area MISC Areas - The Swap Meet WE TAKE NO RESPONSIBILITY FOR ANY TRANSACTION HERE AND NO COMMERCIAL POST PERIOD !!! - Wireless Net - Cameras - Robotics - Routers/Gateways - Mattel JuiceBox