USB-UIRT Forum

Hi Jon,

As I said in another topic about multiroom control, I am trying to set up a remote control system to control five air conditioning appliances from Girder.

It looks that I reached a new dead end. As your previous answers were helpful, I turn to you again.

It appears that Daikin's IR remote control has a different frequency than audio/video (and this seems to apply also to the other AC manufacturers).
I plot the signal of a phototransistor, and the signal was twice faster than for a regular remote control (space between the two IR pulses for a '0' was half the one encountered on an audio/video remote control).

Therefore, the USB-UIRT won't be able to simulate an AC remote control.

Would it be possible to modify your hex code to adapt it to those specifications?
I am willing to help you on this if you agree.

Thanks for all.

ocarton,

Were you looking at an unmodulated IR receiver output or a raw phototransistor output? (I need to know if you were seeing a change in carrier frequency or data rate).

-Jon

Jon,

For my tests, I used a basic phototransistor. I connected it directly to a digital oscilloscope to have a look at the signal which was emited by the remote control.

This allowed me to see the pulses of IR. Each pulse is about 100 nano second long (this applies for AC and audio video remote) but on the AC remote, the distance between pulses was half the one used on the audio video.

However, I don't see the difference you make between carrier frequency and data rate : carrier frequency is doubled, so is data rate, isn't it?

By the way, thanks for you very quick answers. It's a real pleasure!

Ocarton,

Did you manage to get your Daikin A/C working with the USBUIRT? I have the same issue....

I also thought that the remote was transmitting quite a long macro command

Chirpychips

You say the pulses were 100nS long, so how long were the gaps between? I would like to know the actual frequency used.

Data rate vs. frequency is not a constant since there are so many modulation/pulse schemes employed.

ChirpyChips,
what I'm finding is that many A/C systems use a 'smart' remote control which doesn't merely send out a code for the button you press. Instead, the remote itself keeps track of the desired temperature, etc., and instead each time *any* button is pressed on the remote the remote sends out *all* of the current settings to the A/C (i.e., fan settings, temp settings, etc...)

-Jon

If I recall well, there were two lengths for the gaps : 0.9 and 1.4 ms (if I'm not mistaking, the first is a 0 and the second is a 1).
If it helps, I may achieve to get and send a printout of the signal (won't be before next week though ).

So far, I could not verify the fact that the remote sends severall informations at once (temp, fan speed...). But I also believe it behaves that way.
If that's true, it means that the content of the IR messages has to be decoded. Otherwise, we wont be able to recompose a complete valid message when a single order comes from the software. It also means that the appliance's status must be stored in the UIRT : when an 'on' order is sent, the last temperature, fan speed... must also be sent...
Up for some challenge?

Storing in the UIRT would not be the right place to do it -- better to be handled by an upper program.

The IR physical layer encoding sounds odd, however: From what you've said, you see 100 NANOSECOND IR pulses followed by ~1 MILLISECOND spaces...!

-Jon

Hi jon,

Ok for the program. I'll try to see if a Girder's plugin fits the needs.

For the timing, I guess the best thing is to send you a graph of the signal. I'll do that next week.

Hi Guys
Was wondering if this issue has been sorted out ??? If so what was the fix ???
Thanks
Frank

Hello,

I thought I would resurrect this thread because I have been searching for a way to control my Daikin air-conditioner from my computer for years. I got the USB UIRT about a week ago and initial attempts to learn the codes failed (just by learning through girder and " accepting burst" for each keypress).
The learning window indicated that the signal was 33.5 kHz.
So I learnt the code again and accepted burst, then changed number of repetitions to 1, and frequency to 35 kHz (as you can't select lower than that) and voilà, it works!

I have now programmed in voice commands for Dragon NaturallySpeaking and I have voice control of my air-conditioner. I just wish I had got this working about four years ago!

Hello
I have the same problem, and if i follow your email, it doesn't work ...
Could you help me and give me some daikin code from your learning phase
Sorry to wakeup this thread after many months
Kind regards
Domotics