JP1 Remotes

The Robman · Posted: Sat Mar 10, 2012 4:18 pm Post subject:

Have we tried building a Slingbox file based on the UEI upgrade yet?
_________________
Rob
www.hifi-remote.com
Please don't PM me with remote questions, post them in the forums so all the experts can help!

eferz · Posted: Sat Mar 10, 2012 4:55 pm Post subject:

3FG · Expert Joined: 19 May 2009 Posts: 3367

I've uploaded a zip file which contains a protocols.ini fragment, and a class file called BitExpander. This allows the use of RMIR directly, without the need for a manual protocol, and accepts the natural OBCs. It uses the UEI executor, and seems to send IR signals identical to the learns eferz obtained from his Harmony.

Here's the RMDU file (for a RCA RCRP05B--I don't own a Slingbox). Of course this RMDU file won't load unless the above protocols.ini and class file have been incorporated.

vickyg2003 · Posted: Sun Mar 11, 2012 6:30 am Post subject:

The Robman · Posted: Sun Mar 11, 2012 11:11 am Post subject:

3FG · Expert Joined: 19 May 2009 Posts: 3367

Vicky,
Unfortunately, we are at the edge of the Widget's capability. This particularly applies to the frequencies it displays. The shortest On duration is 105uSec for the Humax IR protocol. However, I think Kevin's comment that 200uSec of On time as being the absolute minimum duration** is probably conservative for most waveforms, because the Off duration also is important. In the Humax case, the shortest off duration expected is 211uSec. If it were shorter than 200uSec, we'd surely be getting some errors. As it is, the Widget (and accompanying software) is reporting erroneous frequencies.

A typical IR signal on burst is modulated (rapidly turned on and off) at 38KHz. This particular signal is modulated at 56.9KHz***. The Widget counts the number of modulations that occur within 100uS, and reports this for every 100uS period. Since the shortest on burst lasts for 6 modulations (105uS), the Widget will usually see two consecutive 100uS periods with modulations distributed 1/5 or 2/4 or 3/3 or 4/2 or 5/1. Sometimes, either it or the sending remote will mess up a bit, and see a single 100uS period with 5 modulations.

Bottom line: If we want to see these IR signals in detail, learning into a UEI remote will have advantages, at least in determining frequency, and probably in measuring the durations.

I'll take a look at IRScope to see if we can use more of the recorded signal, and get a better estimate of frequency. Edit: IRScope is already doing the best it can with these signals. It finds all the 100uS or longer intervals for which the IR was on during the entire interval, adds up all the transitions and durations of those intervals, and computes the frequency. The issue here is that the longest such interval is 100uS, because the longest On duration is 211 uSec. During one 100uS interval, a 57KHz signal will show 5 transitions, which would compute to 50.000KHz. If the remote wanders 5% in frequency, 6 transitions are possible, corresponding to 60.000KHz. In Earl's first learns for the Harmony, the calculated frequency was always 50KHz. Subsequent learns evidently have some instances of 6 transitions, and the average frequency is calculated to be a little higher than 50KHz. By comparison, a UEI S3F80 remote can in principle measure frequencies to a fraction of 1KHz. One of the counters has a 0.125 uS period.

**I assume that he based that statement on the Nyquist theorem which states that one has to sample at twice the frequency of the highest frequency to be observed in a signal, but the sampling approach in the Widget is more complex than assumed in the theorem, and it should be somewhat better.

***37.9KHz is 455KHz divided by 12, and 56.9KHz is 455KHz divided by 8. A long time ago, somebody started using 455KHz filters in AM radios, and it is just about the cheapest frequency to make, because of economy of scale. Of course it isn't necessary to to make IR operate at these frequencies, but most equipment does.

binky123 · Expert Joined: 14 Feb 2004 Posts: 1292

I'm not sure if it is relevant here but I recall that johnsfine concluded that the original slingbox was running 2.5% slower than in the Dreambox remote(HCS08) when shooting the XMP protocol. This was way back when we first started learning about the XMP protocol and people wanted to control their Dreambox STBs with the slingbox.

eferz · Posted: Mon Mar 12, 2012 6:17 am Post subject:

vickyg2003 · Posted: Mon Mar 12, 2012 6:41 am Post subject:

I gather that to open the OARRI06G I need to have the experimental version of RMIR and the new RDF. I think I can find the RMIR, but I looked in the RDF folders and can't find anything for this remote.

Also never heard of BitExpander, but I admit that I haven't been very active in a while.

I learned a great deal from your Sling ICT files though. I was under the impression that there wasn't a hold capability on the slingbox, and that if the protocol like this one was being sent, the hold test would fail, but in this case the hold is being detected and there are three frames from the official upgrade, even though when reading the code I expected that only the first frame would be sent because subsequent frames are subject to the hold test.

The ICTs from the OARI06G are a perfect, but I don't know why we asked for them, as the ICTs from the slingbox clearly showed that the slingbox is shorting us on blinks for the various pulses.

bnordman · Joined: 04 Mar 2012 Posts: 17 Location: OH / USA

eferz · Posted: Mon Mar 12, 2012 8:11 am Post subject:

vickyg2003 · Posted: Mon Mar 12, 2012 9:01 am Post subject:

eferz · Posted: Mon Mar 12, 2012 1:26 pm Post subject: