USB-UIRT randomly dies.

[zackcorn]

TechBill, I only had to install the driver once per USB port.

I just kept moving mine between 2 ports, and after installing the driver for both of the ports, it never came up to ask me after that.

[yaccri]

If you plug it back to the *same* USB port, it shouldn't ask for installing the driver.

[dooferlad]

I am told by my physicist wife that the most EM radiation from a CRT is out the sides of the screen, so if like me you have your USB-UIRT stuck to the bottom of the TV you may want to screen it. How you screen it is another matter! Earthed metal is what you need, but I certainly don't have any nice thick copper hanging around (thick and low resistance is best for your Faraday cage). Of course I could go and buy some copper pipes to put the cables in that and get some copper from a real of cable I have upstairs to wrap around the main unit, but that may be a bit over the top!

Unearthed aluminium foil is probably the best most of us can put together in a short space of time. I have no idea if it works, but it makes me feel better Of course I have re-routed my cable to be nice and short and out of the way now so any change in how often my USB-UIRT crashes could be because of that.

[rmk1139]

I have this same issue. Mine would die about every hour or so. I don't want to speak too soon on this one, but I added a Ferrite Ring onto the emitter cable a few inches away from the jack. I have only had it on for a few hours, but it seems to have worked really well. So if any of you happen to have a ferrite ring laying around or know of a good place to get some, then that might help resolve your problem. I will try and post again in a few days or so to give an update on this fix.

In case anyone is wondering what a ferrite ring is:
http://en.wikipedia.org/wiki/Ferrite_bead

[yaccri]

Nice idea!
I hope it works well.

When you post the update, can you also attach a photograph?

Thanks,

Yoram

[jmattioni]

Did the ferrite ring cure the problem??? Thanks.

[rmk1139]

Sorry it took me so long to reply.

It appears as if the ferrite bead had solved all of my problems. Scientifically speaking it is hard to say whether it was for sure the ferrite bead, or some other factor (the cable being moved even slightly?), but since I started using the ferrite bead, I have not had a single crash. Sorry I don't have any pictures, but the installation is pretty simple. Just loop it around a ring a couple of times (5-7 I would say) a few inches off of the jack ( ~ 3in). If you are having trouble finding a bead, then just take one off of an old USB cable. If it is enough for a USB cable, then it should be enough for a USB device (... maybe). I hope this is helpful. If anyone has anymore specific questions about how I installed it, then email me at rkraus (at) purdue.edu. But remember, I'm not an expert, so I may not know anymore than you about it. When in doubt, google.

Like I said, I hope this was helpful.

[dinki]

- As a last resort, I use 'devcon' to automatically restart the device

I realize this is a long shot on such an old thread, but can you or anyone else explain how to use devcon to do this? Will EventGhost handle the reset and continue on? I'm guessing you schedule this restart every so many hours?

[yaccri]

I realize this is a long shot on such an old thread, but can you or anyone else explain how to use devcon to do this? Will EventGhost handle the reset and continue on? I'm guessing you schedule this restart every so many hours?

I do it in the code. When transmission fails (returns an error), I call devcon and resend. To do this, you need to access the code that sends the command to the unit.

[bskip]

After fighting with this for weeks now I finally got a setup that works all the time for me. Maybe this will help someone else. I’m using the USB-Uirt with and MCE remote and two blasters to control my TV and my home theater receiver with Windows 7 RTM and EventGhost 0.3.7.r1194.

My USB-Uirt would randomly stop working and never worked after suspending the computer with Windows 7, Vista, or XP. I went to Radio Shack and bought a bunch of ferrite beads for a couple bucks and put two on my USB-Uirt USB cable, one near the USB connector and one near the USB-Uirt itself. I also put one around the blaster cables, probably overkill.

This stopped the random USB-Uirt failures while Windows was running but not when resuming from sleep. I tried setting EventGhost to disable the USB-Uirt plugin OnSuspend and enabling the plugin again OnRusume but that didn’t work. I tried having EventGhost run Devcon.exe to restart the USB-Uirt driver OnResume before enabling the driver but that didn’t work. I got it to work by having EventGhost run an AutoIt script OnResume and then EventGhost would close itself. The AutoIt script would wait for EventGhost to close, then run Devcon.exe to restart the USB-Uirt driver, then start EventGhost back up. I have not had a single issue since! If you’re using Windows 7 or Vista you would need to turn off UAC for this to work.

The Python command for Event Ghost to close itself is “wx.CallAfter(eg.app.Exit)â€

[Tieske]

I recently had some setup changes, which made my problem reoccur. It never really went away, but now frequency is up again to approx once per day. So 2 years after, my issue still hasn't been solved. Regarding earlier comments in this thread;

I have now rerouted my emitter cable (which were lying around the amplifiers) in another way. Only for 4 days now, but so far so good, I'm hoping this will solve my issue but need some more time to be convinced its been resolved.

That didn't work, problem stayed.

Either way though, the UIRT's receiver light still blinks when it receives IR. So it may be still receiving...it just stops sending.

Never noticed this before, but indeed mine shows the same. Girder throws error messages, and no led blinks when sending commands. But the led still blinks when it receives IR.

Use Ferrite beads on your emitter wires (It helps sometimes reduce EMI, and in weaker cases should solve the problem)

I tried this as well, also to no avail.


Recalling my issue: the USBUIRT fails when I send a power-on/off command to my 5.1 amplifier (randomly offcourse, but always on one of these commands).
I went to a local electronics shop to try and find something that I could use to shield the emitter wires (as suggested in some post, using alu foil). But they didn't have anything I could use for it, but the guy at the counter told me that he didn't think the EMI was the issue, he figured that it had to be noise on the power circuitry.
Anyway; The amplifier always makes a big 'boiinng' sound when switched on, so I insisted on the EMI and he sold me a resistor that only lets power come-up slowly (when its cold, it is highly resistive, and it reduces while it gets warmer. Effect; when power is switched on the power is only slowly comming up to the full power). It worked on the amplifier as it now no longer makes the big "boing" noise when it power up. So EMI now has to be a lot less. But USBUIRT still dies. So another option down.

But it brought me to the point of considering the noise on the powerline to be the cause. After my setup changes I also have a lot of X10 failures. X10 was unreliable already, but now nearly stopped functioning at all. That was another reason to suspect the power supply. In the past I also had trouble with the server being unstable, this was resolved by properly grounding the power supply. Since the server has been stable.

The solutions by yaccri also lead me this way

I found the following:
- Replacing the computer may help. Some computers are better at handling USB.
- Reducing the number of devices that are connected by USB may help.
- Increasing the delay between commands helps
- As a last resort, I use 'devcon' to automatically restart the device

The first 2 items indicate a power related cause. A longer delay didn't change anything for me.
I never used devcon. I tried disabeling and reenabeling the device from devicemanager, but that doesn't work to get it back alive.

For now I've added ferrite beads on the USB cable, and no longer on the emitter cables. I added 2, near the USB connector and the other as close as possible to the USBUIRT.
Other options I'm considering;
- using a powered USB hub inbetween
- adding a noise filter to the power supply

[Tieske]

For now I've added ferrite beads on the USB cable, and no longer on the emitter cables. I added 2, near the USB connector and the other as close as possible to the USBUIRT.
Other options I'm considering;
- using a powered USB hub inbetween
- adding a noise filter to the power supply

Well, the ferrite beads didn't work. Just bought myself a powerstrip with built in surge protection, noise filter, etc. Only the server is now connected to it, so it should be relieved of any noise from the powerline.
For now I left the 2 ferrite beads on the USB cable.

[justinm001]

Ive tried the ferrite bead and it didn't help me either.

I can only send about 10 commands before I need to pause and restart girder but I can recieve 100's without any problems. It seems like I can send about 5-10 commands then if I wait a bit I can send more. this is a pain though if I want to lower the volume 15+ times. Is this the same problem and is there any type of fix?