This is the forum archive of Homey. For more information about Homey, visit the Official Homey website.

The Homey Community has been moved to https://community.athom.com.

This forum is now read-only for archive purposes.

Trouble getting IR buttons to be recognised

13

Comments

  • Sommo said:
    what about replacing the "bottom cup" with a new transparent one? could it resolve? could it be cool to see homey naked :expressionless: 
    Why not top and bottom transparent? Then Homey is really naked  ;)
  • honeyhoney Member
    edited June 2016

    Thanks for making that drawing @marcof it gives a good insight on what part of the signal is lost due to the housing.

    Maybe it's about time @emile or @stefan will comment on this...

    Since they have published that questionable video they have ignored every hardware related IR questions. This really upsets me. If they would provide honest feedback earlier than I would choose the Harmony Hub upon the Harmony touch and would not have this situation.
    @Athom Do you really have to wait till a discussion explodes and than wondering why people complains about poor communication? 
  • MarcoFMarcoF Member
    For learning and fun:

  • MarcoFMarcoF Member
    edited June 2016
    I was pointed on a small mistake in my last drawing. The 3,5° beam is not a circle on its own, but a portion of the 5° circle. This makes the beam twice as wide and the square m2 also doubles. The 2 lowest beams are debatable because around that "highpower" beam is still the 5-10° "lowerpower" beam. But that "lowpower" beam is probably not bouncing.

    Here's the new drawing;


  • These are verry nice drawings, im not able to find suitable lenses to "spread" the beam, maybe @Emile @stefan @athom can point us in the direction of a propper widebeam replacement ir led
  • Looks like Athom needs to come up with a Hubble-ish kind of solution. The IR LEDS could use some glasses.
  • veekieveekie Member
    The use of IR for TV control was one of the main reasons i got the homey. Are there workaround for this issue like IR extenders/enhancers?
  • veekie said:
    The use of IR for TV control was one of the main reasons i got the homey. Are there workaround for this issue like IR extenders/enhancers?
    The latest software status update (https://forum.athom.com/discussion/1430/software-status#latest) talks about an IR update in the coming week(s). I would personally wait for that before trying other solutions (although there do exist WiFi-IR 'dongles' and the like, although I do not think any of those have a Homey app at the moment).
  • I hope I can disconnect my Hormony hub asap  :)
  • Trustme2Trustme2 Member
    edited July 2016
    Will Homey support 'all at once' IR transmissions? Normally IR only sends: on, off, 1, 2, 3, blue, green etc.

    the problem is that The remote of my airco sends all The information at once. Like: 19 degrees celcius+cooling+highest fanspeed+swivel in one transmission. Although Homey is seeing this IR transmission in manual mode(after many, many tries), when I try to send this With my Homey to The airconditioner it Doesnt work. Not even The simple(??) on/off. 

    My question is: will Homey support these all at once IR transmissions?

    I have an Electrolux airconditioner.

  • Hmm never thought about this, but never heard about any problems with universal remotes etc.
    I think you airco works like this:

    If temp is 19 and wave is on send  IR feq 1.00
    If temp is 19 and wave is off send IR feq 1.01

    Etc.

    What I mean by this is that you remote create an understandable IR signal through the input what it recieves from the buttons.


  • Okay, I've been reading this topic with high interest, how can I get the top off Homey, no one of my IR works ...:(
    Sometimes my curtains do respond (setting is now 40! pulses from homey) original remote needs only 1 short pulse, so I also want to open he IR LED gaps for better spreading the IR signal, I also use a IR extender for use in my bedroom (airco, TV etc.) 
  • veekieveekie Member
    They are current;y trying to improve the IR quality. Maybe that will also fix your problems. I would suggest waiting for the fix until trying more permanent solutions.
  • daylfydaylfy Member
    Was this another misleading movie they posted ?



  • daylfy said:
    Was this another misleading movie they posted ?



    That video is actually completely and fully legit. The only differences between the Homey in the video and the delivered Homey's are the assembly process, and the software. The Homey in the video was completely hand-assembled by me, while the delivered Homey's are produced at larger quantities. Before packaging, several QA checks are performed, and any Homey that does not meet a minimal range requirement is rejected. The software in that video is just the plain IR driver of Homey, without learning capabilities. I've manually entered the transmitted IR waves of the TV we were controlling.

    The only three things i can imagine to cause these kinds of issues are differences in configured IR-carrier frequency, RGB LED interference and ranges above aproximately 5-6 meters. A difference in sending and receiving carrier could drastically reduce the range, but Homey can send with a variety of different carrier frequencies. For predefined devices this should be correct, but the detection of self learned devices is sometimes a bit off. I think there is a fix for this in the new IR pairing wizard. Secondly, the RGB leds sometimes cause interference on the receiving side of Homey. This can be easily filtered out and should not cause any problems, but i've seen a few device specific issues where the IR beam from the remote was too weak to distinguish. For ranges above 5-6 meters, the angle of the beams are too narrow to provide full 360 degrees coverage, (below that, the reflections of the IR light in the LED ring contribute to 360 degrees coverage) so you may want to rotate your Homey a bit in order to create a direct line of sight. In the weeks before this video was posted, i did a lot of testing with about 20 types of leds and a variety ofexternal lenses. The problem with the lenses is that they decrease the line-of-sight distance in favour of a higher angle, while this improves short range performance (<3m), it actually decreases long range performance. Therefore i doubt wether you'll see improvements if you start adding DIY lenses in front of the IR leds. 

    @Trustme2 ; All-in-one IR transmissions are possible using Homey apps, once the IR API is released (It is going to be very similar to the 433 and 868 API)
  • MarcoFMarcoF Member
    edited July 2016
    At 4m the beam is so extremely narrow and weak that's almost impossible to point it at 1 device and not to speak about multiple device next to each other. 

    I really hope the IR update will be released soon and it brings a big improvement. Otherwise IR is, in my standard homey situation, useless..... 
  • daylfydaylfy Member
    daylfy said:
    Was this another misleading movie they posted ?



    That video is actually completely and fully legit. The only differences between the Homey in the video and the delivered Homey's are the assembly process, and the software. The Homey in the video was completely hand-assembled by me, while the delivered Homey's are produced at larger quantities. Before packaging, several QA checks are performed, and any Homey that does not meet a minimal range requirement is rejected. The software in that video is just the plain IR driver of Homey, without learning capabilities. I've manually entered the transmitted IR waves of the TV we were controlling.

    Now i read that its a different product than the one’s Athom sold to us, or am i wrong?

    So i saw a demo with a special handmade homey running special software made for IR, when in reality its a different product then the one i received. 

    To me it's a bit misleading,

    I really hope you guys mange to fix it, i know you are all doing the best you can, but to show movies like that isnt really honest imo. 

  • casedacaseda Member
    @daylfy ;
    You have interpreted that completely wrong.
    It is the same homey, except for the reason it was put together by themselves, and the ir command was competently hand written, so not a copied IR-code

    The homey's are now put together by a company that mass produce the homey's (which makes way faster assembly)

    The only negative thing is that the attention of putting it together is lower, so a little miss alignment can happen, but doesn't mean it will happen (in my case only 1 IR-led is too low).
  • For ranges above 5-6 meters, the angle of the beams are too narrow to provide full 360 degrees coverage, (below that, the reflections of the IR light in the LED ring contribute to 360 degrees coverage) so you may want to rotate your Homey a bit in order to create a direct line of sight.
    This made me laugh out loud, correct me if im wrong but i think the tv was 10 meters away in the video?
    also it was said that the beam would bounch atleased a couple of times just to provide that "range" so u actualy dont have to "aim" for ur devices.

    also how would i be able to line up 3 devices in the same direction but at different hights.
    homey is no more then 1 meter away from my tv and stereo and there not responding to my homey's IR what so ever even with the hole's being bigger. (previous posts)
  • honeyhoney Member
    edited July 2016
    That video is actually completely and fully legit.
    Yes, In that distance, angle and tilt.
    @JeroenVollenbrock said:
    For ranges above 5-6 meters, the angle of the beams are too narrow to provide full 360 degrees coverage
    Than how narrow it is in the 1 meter range?
    @JeroenVollenbrock said:
    In the weeks before this video was posted, i did a lot of testing with about 20 types of leds and a variety ofexternal lenses. The problem with the lenses is that they decrease the line-of-sight distance in favour of a higher angle, while this improves short range performance

    Seriously? Logitech and all the others using lensed LEDs are wrong? Oh my Logitech works from a 7 meters distance. And the bouncing is perfect. 

    The update might fix the IR learning but wont fix the hardware. The beam, the lack of bounce, the blocked windows will all stay. It won`t work much better than it works now with the `predefined` codes.
  • honey said:
    That video is actually completely and fully legit.
    Yes, In that distance, angle and tilt.
    Nope, it's legit. Period. The only thing different in that video was that we read the codes manually from the remote with some electronics, so the only difference is software.

    honey said:
    Than how narrow it is in the 1 meter range? 
    We don't really have an answer to that as it's more nuanced than just taking the width of the beam coming from the LEDs. The width of the beam of just the LEDs is quite narrow. This was done on purpose, as everything you gain in width of the beam, you lose in actual distance (if you'd visualise it in a graph, the area stays the same). However, the beam bounces partly against the plastic of the half-transparent LED-ring, which means that in short range the beam gets spread out pretty well.

    Bouncing does work, but don't forget that a bounce introduces loss of signal. Because no surface is 100% reflective, a part of the signal gets absorbed into the surface. Furthermore, the signal gets spread out on uneven surfaces, and of course, it has to travel twice the distance (to a wall and back = 2x longer vs direct line of sight). So bouncing should work, but may be less effective. This is the same on every device though.

    So, to sum it up:
    - Close to the Homey (< 3m) the beam gets spread out, and you should have full coverage over 360 degrees.
    - Further from the Homey (> 3m) the beams get narrower, and Homey will have some blind spots (6 of them, in between the LEDs). It depends on the software, but you might have to turn Homey up to 30 degrees to achieve a maximum.
    - The Infrared signal bounces, although the LEDs may or may not be powerful enough to achieve that distance and to overcome dampening of signal because of the walls. Rotating (max 30 deg) is way more reliable than relying on bouncing.

    honey said:

    The update might fix the IR learning but wont fix the hardware. The beam, the lack of bounce, the blocked windows will all stay. It won`t work much better than it works now with the `predefined` codes.
    You're right, the upcoming IR update won't fix hardware, but we're quite sure the hardware does not require fixing. Although the hardware is not ideal, it should be more than enough to achieve a satisfactory performance.

    _____________________________________________________

    In the upcoming update, we're going to fix a lot of stuff in the software, which should increase the transmitting distance. The carrier frequency of the signal was not calculated correctly previously, which causes a lot of difference in transmitting distance (that may cause the signal to only be receivable under 1m, whereas normally 5m could easily be reached).

    Besides that, we've added some pretty cool AI-related algorithms to detect toggle bits. Every time you press a button on your remote, a small part of the signal inverts, so the TV can detect multiple button presses versus one long press. This is hard to detect because the signals are so similar. We've added a really badass algorithm that can differentiate between the two, so even manually learnt signals have toggle bit support. This should improve both reliability (we can just fire one button multiple times without the TV triggering multiple times) and user experience (you can control the device more precisely).

    As a cherry on top of this big IR pie, we're finishing up support for our own IR signal database as well, which means that halfway through teaching Homey the remote buttons it should start autocompleting with information from our database.
  • Thanks @superice for the update and looking forward to it!
  • @superice

    its rather disappointing ur only responding to the last post of @honey ; there is so much testing / adapting done in this topic and not getting a response from any one at athom what so ever on that.

    we are experiencing no coverage at all if its not aimed correctly (almost has to be done with laser in order to line it up right) and thats well within 3meters.

    appalling, cant find a nicer word for it.
  • @superice

    its rather disappointing ur only responding to the last post of @honey ; there is so much testing / adapting done in this topic and not getting a response from any one at athom what so ever on that.

    we are experiencing no coverage at all if its not aimed correctly (almost has to be done with laser in order to line it up right) and thats well within 3meters.

    appalling, cant find a nicer word for it.
    Jeroen already replied to most of it. We're quite sure this is a software issue, as extensive testing has pointed out the half transparent part of the LED-ring shatters the IR beam more than enough to cover 360 degrees, at least close by.

    In individual cases, the LEDs may have been slightly misaligned with the case, causing part of the signal to bounce back into Homey, but seeing as we pump an incredible amount of power into the IR LEDs, the remaining part of the beam should be more than enough. We've adjusted our production process to align the PCB better to the case quite a while ago, so this should not occur anymore.

    We're confident that almost all of the issues will be resolved once the carrier frequency of the signals is configured properly (in software!) even with the early Homeys. Even if only 10% of the beam makes it out, that should be more than enough signal strength. We're outputting (!) more power through the IR LEDs than the other IR hubs use (= input) at all!
  • MarcoFMarcoF Member
    @superice and @JeroenVollenbrock ;
    Thanx for the info. 


    I totally agree with @MarcelKuijper we had to wait to long on a response of athom. 

    @Athom ;
    When could we expect the IR update? 
  • MarcoF said:
    @superice and @JeroenVollenbrock ;
    Thanx for the info. 


    I totally agree with @MarcelKuijper we had to wait to long on a response of athom. 

    @Athom ;
    When could we expect the IR update? 
    Yes, I agree, and I'm sorry. We're working on it ;)

    Anyway, the update will be there somewhere next week hopefully. Should be either the update after 0.9 or the one after that. We're making progress but it just isn't good enough yet.
  • MarcoFMarcoF Member
    OK hopefully is get the "good enough" stamp "asap" :) 
  • well wait n see, sure hope this can be fixed by software.
  • small update,
    TV is working descently at this moment, havent made any changes to homey sinds my last adjustments.

    Im using IR with the rgb led ring off, this could be a factor of interference aswell. 

  • Im using IR with the rgb led ring off, this could be a factor of interference aswell. 
    Do you turn off the LED ring in the setup completely or in a flow?
    I like to do it in the flow itself, but can't find it there...
Sign In or Register to comment.