Delcom driver question for jp1

Post by **ElizabethD** » Sat Apr 05, 2008 9:12 pm

I wanted to try the USB cable from Tommy for jp1.
During installation of the Delcom USB driver 3.1.0.0 I got a nice BSOD faulting the USBIODS driver.
So I downloaded 4.6.0.0 version. No BSOD. No workie. Half-way through a download from the 6131 it says it failed and creates a debug file https://www.hifi-remote.com/forums/dload ... le_id=5501
.
Windows XP home.
Ideas? I know it'll work, it's just the drivers can kill you.
I'm confused what to use because in the Software section at Delcom they have USBIODS file as well as something called Windows drivers. Am I supposed to use something else?

Post by **ElizabethD** » Sun Apr 06, 2008 7:54 pm

I noticed I've been running an obsolete IR

So I updated to v7.11 yesterday.
Cleaned out the previous Delcom driver(s) using the Remove code Tommy supplied.
Installed v3 again. This time no BSOD. No problems. Unlike yesterday, I can plug it in, take it out, reboot, use external drive in the same slot, change back to jp1 cable, startup the computer with no problems.

But the failure half-way during the IR download from remote issue persists. In diagnosis is a debug file from today, this time on 8910.
https://www.hifi-remote.com/forums/dload ... le_id=5504

On the driver itself, Everest reports:

[ Universal Serial Bus controllers / Delcom USB IO Driver ]
Device Properties:
Driver Description Delcom USB IO Driver
Driver Date 5/6/2005
Driver Version 3.1.0.0
Driver Provider Delcom Enginnering
INF File oem57.inf
Hardware ID USB\Vid_0fc5&Pid_1222&Rev_0010
Location Information USB IO Controller

Is there any other information that could help debug this?
I'm using parallel cable, so this is not any emergency. Still, being the investment in the future, I'd like to resolve it for when USB is the only game in town.

gfb107 · Post by **gfb107** » Mon Apr 07, 2008 5:21 am

I'm no expert on drivers, but I've had success using version 4.6.0.0 on Windows XP Professional SP2.

Post by **ElizabethD** » Mon Apr 07, 2008 7:51 am

Thanks for the feedback, Greg. Confirms that it MUST work.
Home/Pro should make no difference since most of XP is identical except for few network drivers and such.
I don't know HOW to go about figuring when/where/why it stops. Sounds to me like some timeout issue or miscount someplace.

My XP laptop is of hyperthreading variety, but I'm not even sure it it's being used really. I read somewhere on this site there were issues in hyperthreading.

Mark Pierson · Post by **Mark Pierson** » Mon Apr 07, 2008 8:36 am

ElizabethD wrote:I don't know HOW to go about figuring when/where/why it stops. Sounds to me like some timeout issue or miscount someplace.

Check for power save settings that affect the USB ports. If they "go to sleep", they can disable any drivers associated with them. I've seen this problem on a few desktops and most laptops. Disabling any power-saving on the ports generally works.

Post by **ElizabethD** » Mon Apr 07, 2008 7:58 pm

Mark, I looked at the power state properties. I don't understand any of it really. However, wouldn't you think that it wouldn't fall asleep while is busy downloading

Napping at work

In any case, I've collected few screen shots of everything power. Do you see any bad settings? Should I look somewhere else? yes, it's a laptop but running on AC.

Also I compared a USB and Parallel downloads from the same remote and it's all here (power state pics, excel comparison, data from USB and raw 6131 data)
https://www.hifi-remote.com/forums/dload ... le_id=5507

There's no question the download takes place, but 40-50 bytes are mangled by $FF.

unclemiltie · Post by **unclemiltie** » Mon Apr 07, 2008 9:11 pm

I know that Nils had some serious issues with the USB cable a while back and only after a good bit of troubleshooting and "uninstalling" by hand (ie: registry editing) was he able to make the drivers re-install.

Any chance you have another cable to give a try? The Thumb version that Tommy sells and the DIY Gadget version use the same Delcom chip and use the same driver.

Post by **ElizabethD** » Tue Apr 08, 2008 3:42 pm

I have some new information. I played with the transfer rate which, according to the selection popup, isn't even meant for USB but what do I have to gain by not trying?

1. Using 250 it fails in the same way as before.

2. Using 750 different failures
> trial1: IR said that DevType15 4095 is not currently assigned to a device button...XSHIFT-Phantom3 will be ignored. 6131 has no 15 devices, just 8, and I have no idea what 4095 is other than it sounds like couple FF bytes.
> trial2: Download occured, but macro and SP sheets were blank
> t rial3: failed midway as before.

3. Using 1500 I got a flawless transfer twice out of three tries and it was very fast, unlike previously when it took forever to say it failed.

Any ideas?

Bill, in addition to running Delcom removal, I cleaned out the registry to the best of my abilities. I'll never forget Nils struggling with this one.
No I do not have two cables. This is my first ever trial of jp1 via USB.

Post by **ElizabethD** » Tue Apr 08, 2008 7:30 pm

Well, in case anyone looked at the excel comparison with 6 or so bad rows, today I have a download with 48 rows in trouble.
One in four trials worked perfectly with the 1500 setting

I guess I'll wait till Delcom puts their act together and get the darn thing certified by M$. No wonder it's not certified.

Post by **ElizabethD** » Thu Apr 10, 2008 8:30 pm

Continuing this monolog:
Today, the cable is connected. I boot the computer. I get the Blue Screen faulting USBIODS.sys. Quoting, from the lovely blue screen, to the best of my transcription abilities:

Stop 0xC000007E (0xF7CCA8E0,0xF7C40A98,0xF7C40794)
USBIODS.SYS – Address F7C40794 at 7CCC9000, date stamp 427C5648

Anyway, it continues - check with the manufacturer of the driver for updates blah, blah, blah.
I booted to the “last good configuration”. There was a suggestion for twiddling with some advanced settings. I have no skills to do any such thing.
I hope my computer isn’t trashed. I will uninstall that driver. Obviously something is really bad if it breaks booting. I had BSOD soon after the installation of this driver, but then it kinda worked. To get BSOD it now is pretty sad.

Could it possibly be a processor affinity issue?

I wonder if forcing the driver, or IR (?????) to one processor might fix the issue, but I don’t know how or where to do it. Some info here
http://www.microsoft.com/technet/prodte ... x?mfr=true
http://en.wikipedia.org/wiki/Processor_affinity

Maybe these 3 API functions could do the job someplace?
http://msdn2.microsoft.com/en-us/librar ... S.85).aspx
http://msdn2.microsoft.com/en-us/librar ... S.85).aspx
http://msdn2.microsoft.com/en-us/librar ... S.85).aspx

Why? I saw few posts about some Dell laptops (Latitude?) also experiencing problems with the Delcom driver. Some of those Dells are Pentium M, dual core.
I do not have one of those dual-core processors. I do have two processors. Earlier, when I watched the IR download in Process Explorer’s graph, both processor were active. Perhaps the hopping between those chips mangled the downloads. But that still doesn’t explain to me the BSOD at boot time.

unclemiltie · Post by **unclemiltie** » Thu Apr 10, 2008 10:09 pm

Liz

just went and looked. I'm using Delcom's driver version 4.4.0.0 dated 1/30/08. I'm using this both on my Windows Vista (on Mac using Parallels Desktop) and on my Windows XP laptop.

I've had no issues with this driver. I did have issues with the driver version prior to this on Windows Vista (especially the V4.x.x.x versions which just didn't work) I did not have issues in the past on my laptop wth the V3.x.x.x though.

Post by **ElizabethD** » Mon Apr 14, 2008 8:22 pm

I did some checking on several computers in order to rule out the cable and possibly see some pattern. And rechecked again and again on my computer.

1. Recent HP laptop - Pentium M, one processor, 1.73Ghz, 3 USB2 ports. Windows 2000 SP4.
using Delcom version 3.1.0.0 - hopeless. SDA held low during checking the interface. New batteries make no difference. Doesn't work. Period.
Using Delcom version 4.6.0.o - flawless installation and data transfer. On restart of IR, I once got an IR message "DeviceIoControl failed in I2CWrite", but then it worked.

2. Old Compaq laptop#1 - Pentium III, one processor, 597MHz, 1 USB1 port. Windows 2000 SP4.
using Delcom version 3.1.0.0 - flawless installation and data transfer.
Curiously, a mouse and the USB cable were sharing a hub here since this computer only has one, old USB1 port. So hub or no hub made no difference.

3. Old Compaq laptop#2 - Pentium III Mobile, one processor, 1066MHz, 2 USB1 ports, Windows XP-Pro SP2.
using Delcom version 3.1.0.0 - flawless installation and data transfer.

4. My laptop - Mobile Intel Pentium 4, 2 processors, 3.06GHz, 3 USB2 ports. Windows XP home SP2.
using Delcom version 3.1.0.0 - installation is fine. Data transfers unreliable. They do get a bit more reliable with use and a setting of 1500 for speed. But it's not too good, as described in the above posts.

Since that HP laptop v4.6 worked well, I tried my computer again. I uninstalled v3.1,cleaned out the registry, and installed v4.6.
No BSOD yet, but doesn't work. Not even one good download. In one of my trials, 16 bytes are totally missing.

Once 8 byte packets get lost, (not replaced by $FF data as in v3.1), everything shifts.
Those bytes do appear in the repeats IR is doing and I see them via the packet sniffer, but the displacement isn't cured, so I can't use it.

End of a long story. What conclusions? I dunno

Post by **ElizabethD** » Mon Apr 28, 2008 12:05 pm

Update: well, it's still not working well enough to be used by me.
I downloaded a USB sniffer SniffUSB-x86-2.0.0006
http://www.pcausa.com/Utilities/UsbSnoop/
Notepad output is impossible to read. Wordpad is better for looking at the logs.

I watched the downloads fail while the sniffer was writing an incredibly long file.
In one trial, using Excel, I identified packets MISSING DATA and/or control (setup) bytes as well as an extra unexpected packet. I don't know how/if IR reacts to that sort of thing other than the checksum is bad and repeated downloads start on a wrong address.

In case anyone else who can't get the Delcom driver to work wants to try the tedious job of finding what isn't there, in the sniffer log search for this exact phrase:
"TransferBufferLength = 00000000" without the quotes.
There are more of those than the count of missing data packets because of the setup packets.

Whether it's IR or Delcom issue, I can't tell, but I suspect Delcom missing a beat here and there and returns empty packets for some reason.

Post by **ElizabethD** » Sun Dec 28, 2008 7:09 pm

Well, I tried again on my main laptop - Toshiba Satelite A75 S226, XP home SP3, details in previous posts. Using newest IR 7.0.15.1 and Delcom 4.6.0.0 driver. Before that, I used the remove.exe utility Delcom provides and Tommy has in his package. That utility is not good enough, since it left tons of stuff in the registry from v3. So I cleaned it all out, and reinstalled Delcom's sorry driver.
Delcom thing is active as seen in Device manager, but on another bootup causes BSOD -
Stop ox0000007E
values in ( ) c0000005,f620e084,f7b38a98,f7b38794

and have to reinstall driver again and again.

When I run IR, check interface is ok. Downloads from remotes, this time using the newest IR7.0.15.1
6131 missed 40 8-byte segments
7800 missed 25 8-byte segments
8910 missed 41 8-byte segments

This is once again XP-home but this time SP3, using DELCOM v4.6 driver which was supposed to fix hyperthreading, multiprocessor boxes.
It appears that Delcom is unable to certify the driver, therefore, Tommy's USB cool-looking interface for me is doomed since Delcom prefer to remain buggy.
Of course it could be that IR is missing packets, I have no way of knowing.

unclemiltie · Post by **unclemiltie** » Sun Dec 28, 2008 8:20 pm

Liz

I have tried numerous times to get the 4.x drivers from Delcom to work on XP (SP2 and SP3) running in the Parallels desktop virtual machine (on a mac) to no avail. I finally went through the registry, ripped everything out and reverted to the last 3.x version that I could find and all is well.

no idea what is going on with the folks at Delcom, but I'm not touching these drivers again!

-bill