eg stops getting events from usb-uirt

[dbbd]

Something strange is happening from time to time. EG stops getting events from usb-uirt.
Restarting EG issues an error initializing the usb-uirt. indows device manager say that the uirt is enabled and is working fine.

What helps is unplugging/replugging the usb-uirt, then restarting EG and it all works well again.

Any ideas?

Dan

[piert]

Have you, by any chance, attached a wired IR emitter to the USB-UIRT?

If so, try disconnecting it and see if the problem persists. If this solves your problem it means that the wired emitter is running through an electro-magnetic field (such as those around receivers and TV's) and causing the USB-UIRT internal electronics to receive a voltage back through the IR emitter. Possible solutions are re-routing the wire or shielding it, which is what I have done.

Best regards,
Perry

[piert]

Also read through this thread:

http://65.36.202.170/phpBB2/viewtopic.p ... a7b4ce9bea

[ijourneaux]

I have also noted that my new USB-UIRT seems to want to stop providing events to Eventghost. Some times I can get around it by disabling and re-enabling the USB-UIRT in EventGhost. Sometimes I have to unplug/plug the USB-UIRT and restart Eventghost.

I don't have ferite cores on the usb cable to the USB-UIRT. I do have an emitter and I am not using shielded cable.

I will try these things hopefully this weekend. In one of the other threads, I believe someone created a script that would detect the problem and reboot the USB-UIRT. I will also try that this weekend. Unfortunately, as with most of the problems I have been experiencing recently, the problem is not reproducible on demand.

Unfortunately this is getting in the way of WAF (Wife Acceptance Factor) on my new HTPC.

I will post regarding my experience.

[piert]

Hi ijourneaux,

How were your test results?

[kushal_fpost]

I m also facing the same problem, please somebody help me out.

[ijourneaux]

I did add ferite cores to the long IR Blaster line I have to use but that did not solve the problem. I didn't have any shielded wire so I the blaster cable I had installed but the problem persisted.

One question I had is if I did have shielded cable, where would you ground the shield to?

I did notice that when the problem does occur the problem is only on the IR recieving as the blaster function continued to work.

WHen it does stop recieving events, stopping and restarting EventGhost does not resolve the issue. If you restart EventGhost, you get an error when eventghost tries to start the USB-UIRT.

Unplug the USB-UIRT and plug it back in and restart EventGhost and eerything works again (atleast untilthe next time).

[piert]

I did notice that when the problem does occur the problem is only on the IR recieving as the blaster function continued to work.

This appears to be a different issue than mine, then. In my case there is is no way I am able to use the USB-UIRT for emitting or receiving in Eventghost once it gets into 'stuck mode'.

As for the shielding, I am not using a ground: I am using an unwrapped steel wool cleaning sponge and I run the emitter wire through the center of the wool 'tube'.
I assume the way it works is that the (rather voluminous) 'sponge tube' acts as a Faraday cage.

[omaier]

Hello,

I just registered into this forum to tell you how I fixed my occasional problems in using USB UIRT in combination with Windows 7 and EventGhost. My HTPC did not respond on IR-Inputs every 3 to 6 days after starting out of standby. As a temporarily workaround a restart of the driver (via devcon) and a restart of EventGhost fixed the problem.

But I was not really satisfied with this workaround and opened the UIRT case to dejam the circuit. In my opinion the capacitve filtering of the UIRT Powersupply is to small to filter the high currents when the IR Sending Diode is active.

So I replaced the 100┬ÁF Elko by a 200 ┬ÁF type (watch polarity!) and soldered a 100nF ceramic capacitor in parallel. Additionally I soldered a 100nF ceramic capacitor beetween the Pins 5 (VSS) and 14 (VDD) of the Microcontroller. To further improve EMC I made a second ground connection from the IR-Diode shielding case to the PCB and added the a cable ferrit to the USB-cable. Until now (for about 2 weeks) my UIRT worked completely stable and did not need any restarts - so I would call my modifications "successful".

You should not make these modifications and risk to damage your UIRT if you are unexperienced in electronics.

I hope that my description helps some of you in making your HTPC more stable (and improve the WAF).

[dopdahl]

Nice work. I just added wired emitters to mine today and got locked up once...

If it happens again, I'm going to use your hardware mods... Thanks,

[piert]

I just registered into this forum to tell you how I fixed my occasional problems in using USB UIRT in combination with Windows 7 and EventGhost.

Thank you for sharing this information!

Did you ever experience any further problems with your mod?

I am interested in this solution (as the 'steel wool Faraday cage system', though practically working, is not really aesthetically pleasing), but my electrical engineering and soldering skills are, unfortunately, insufficient.

Perry

[dopdahl]

I'm no authority here, but the concept is to increase the ability of the circuit to withstand current draws which result in voltage drops as the emitters are fired.

Fast and/or significant current draws on a circuit can cause voltage drops if the power supply can't respond quickly enough to maintain proper voltage. When proper voltage isn't maintained, it can cause the PIC or other ICs running in the circuit to lockup.

So our friend above has taken a belt and suspenders approach. Increase the size of the 'big capacitor' on the part of the circuit that feeds the emitters and add a high frequency cap in parallel to help with response time. The capacitors act as 'current resevoirs' and are able to instantly supply current to the emitters preventing a voltage drop that would otherwise occur as the voltage regulator is slower to respond to the increased demand for current which results in an unwanted voltage drop as it tries to supply the current.

Secondarily, he's added a high frequency cap across the pos and neg. pins of the IC. Basically, does the same thing. Helps to smooth out the availability of current to the circuit which helps prevent voltage drops on increased current demand.

It should be said that there is not necessarily anything 'wrong' with the USB-UIRT circuit as is. It's an awesome piece of hardware. Different emitters may use more current than others, some people send all commands to both emitters and the blaster at the same time instead of just to 1 of the emitters, etc.

Interestingly, this same exact issue exists in a pinball machine power board that I have. The machine would constantly reset when both flippers were fired simultaneously. Turned out that there was actually an error on the board. A capacitor was incorrectly located on the board rendering it useless. Hacking a cap in the right place, completely fixed the voltage drop and therefore the reset.

[piert]

Thank you for the explanation. It reminds me again of how much there is left to learn