Creative Labs Infra (PC-DVD) Remote - URC 8811 not learning?

[The Robman]

It was a different poster that had a weak battery. All my batteries are fresh. Girder responds the same with learned signals or the upgrade file. If I train Girder with the original remote it works ok. The main problem is getting my 15-1994 to reproduce those codes.

I think you may have missed the point that John was trying to make. He was suggesting that you try training Girder with the 15-1994 when it has the upgrade installed, rather than training it with the original remote and then trying to use the 15-1994.

[johnsfine]

It was a different poster that had a weak battery.

It was a different poster who ALSO had a weak battery.

You posted learned signals. I am quite sure the remote used to send those signals for that learning had weak batteries.

Girder responds the same with learned signals or the upgrade file

I don't believe that, at least not for the learned signals you posted. I believe Girder has trouble with both, but I don't believe the problems would be similar enough to be described as "responds the same".

If I train Girder with the original remote it works ok.

I guess that means Girder can learn the bad signals you get from weak batteries. A JP1 remote can't do a decent job learning such signals.

[zaphod7501]

The JP1 remote is also going to output a much, much more intense signal then the credit card remote (ie the JP1 range may be much greater) and it could be overloading the remote receiver's decoder chip (which can only decode NEC1 protocol, I believe, which is why you can't use the receiver to pick up any old remote's signal).

There is an IR speed setting in the "settings" section for the receiver's plugin (in Girder) that could be adjusted, too. Of course, I'm still using 3.2.9 version of Girder. If you look at the "eventstring" you could compare the perceived command from both remotes to see if Girder sees the same signal. I can also tell you the "eventstring" that my Girder sees with different button presses if you want.

Another note, the "eventstring" that Girder sees is not the signal from the remote, but rather a signal generated by the receiver after it decodes the remote. So the problem might not be in Girder but in the receiver that's not decoding the signal properly. You might want to check/adjust the com port settings also.

[bluerider]

The original remote's battery is a 3.3V cell which measures 3.29V so it should be fine. The IR file that I posted was one where I just learned 5 signals to show that the 15-1994 can't learn most signals from the credit card remote properly. The signals that it does learn can be used to control girder but with the same unresponsiveness.
I have tried training girder with the upgrade installed on the JP1 remote and it does work but again it's unresponsive (misses button presses).

[zaphod7501]

This'll be for the experts to figure out but I looked at the output of a credit card remote and an upgrade in my 6131. Even the shortest press of a button on the Credit Card Remote produces multiple outputs, while the upgrade will send a single pulse (on a short press). The receiver (which is a decoder) apparently knows how to discard the extra pulses and send a single string to the PC. In other words, they built a debounce circuit into the receiver (to ignore a too long or multiple keypress). This receiver does not pass the remote signal to the PC, it decodes it first.

Since my 6131 (with extender) has no trouble with controlling the CIMR100, I would suspect the timing to be slightly different between remotes (6131 vs 15-1994). This device may be similar to the RCA VH920 A/V switch that requires a duration increase to work properly.

If adjusting the IR Delay in the settings for the Girder plugin doesn't help, then using the Device Combiner would give you the ability to adjust the duration of the signal which might aid the receiver/decoder to recognize the signal better.

Again, it's probably not Girder that is having trouble, it's the receiver. You can watch the Girder Icon (which can be set to flash on signal) and see if it even flashes on the undetected commands.

P.S. 3.29V might indicate a low battery since a good one usually measures over 3.3. You would have to measure the voltage when you press a button to determine if it is good or bad (very difficult on this remote).

The OBC codes in the RM upgrade should be correct since they have been working for me for a long time on several PC's and several receivers. (I have 5 Hollywood Plus cards and 6 receivers)

[The Robman]

Steve, were you able to figure out how many repeats of the signal the credit card remote gives with the shortest button press? (Keep in mind also, that your upgrade uses the NEC1 protocol so there's just a 2-word repeating pattern, it's not the whole signal that's repeating).

For the non-Java people, here's a KM version of the upgrade...
https://www.hifi-remote.com/forums/dload ... le_id=2273

[zaphod7501]

I didn't know how many repeats but I watched the output with a camcorder. They see into the near infrared so you can see the LED blinking. That's how to determine if a remote control has a coherent output when someone brings a remote into the shop that doesn't work on their TV.

I'll try to capture the signal with my Hauppauge PVR250 and run it through an editor and see if I can count the repeats. ................................................Wasn't sure that would work, but the shortest press of the credit card remote produces 2 distinct output pulses. ................................................Further, learning with my 8910 show both bursts with identical signals. Longer or continuous presses show up as NEC1 while the shortest show as just NEC, which would always be the first burst.

It looks like the original remote does repeat the entire signal twice. The first lasts 3 video frames, then off for 1 video frame, then on as long as the button is held down. This is as close as I can measure within the limits of my editor and capture system. I'm trying to convert from frames to seconds in my head. There are 30 frames per second so if I am doing this right then it's first burst is 1/10 second followed by a gap of 1/30 second then the continuing signal. Those times could be slightly longer, but within 1/30 second.

Sorry this took so long but I kept thinking of something else to try (or I had to move the laundry around).

[The Robman]

The NEC executor can be made to send the original signal twice before sending the repeating portion by editing the parm value (as described here). In this case the parm value should be "22".

I've just updated the KM file that i created to reflect that:
https://www.hifi-remote.com/forums/dload ... le_id=2273

Steve, the best way to capture the sort of info I was talking about would be to use the URC-8910 to capture a very short press of the credit card remote. Then, in IR, you wouldn't be as interested in the decode information as you would the raw timing information. This should all appear in the "Sent Once" panel but if you break it down by hand you should be able to figure out exactly what was being sent.

Bluerider, either give my upgrade a try, or re-edit your own upgrade and enter a parm value of "22" and see if that fixes it.

[zaphod7501]

but if you break it down by hand you should be able to figure out exactly what was being sent.

That's easy for you to say, but, sorry to say, I haven't gotten to that level of proficiency yet. I've been letting others do the dirty work.

On the other hand, I did change my upgrade and it responds much better with any kind of keypress. Maybe the remote had trained me to press in a certain manner so that it worked all the times I tried it. I actually had trouble getting the original remote to work consistantly. I am just setting up the remote receiver on a faster PC right now so my old upgrade might not have worked as well as before (Sempron 3000+ vs K6-2 300).

[The Robman]

That's easy for you to say, but, sorry to say, I haven't gotten to that level of proficiency yet. I've been letting others do the dirty work.

I didn't necessarily mean that you personally could figure it out, but that if the IR file or the timing data was posted, people like John Fine and myself could easily have figured it out.

[bluerider]

Thanks guys!

My JP1 able is at work so I'll give this a try tomorrow & report back.

[bluerider]

Thanks again!
I've used the new upgrade file and it's working much better now!
I really appreciate the time you guys have spent on this.