I’ve managed to use blynk to control my A/C unit, by using an IR receiver that I disaasembled from an LED Strip controller to read the signals from the a/c remote. Everything works fine but there’s a problem: the signal sent from the board is pretty bad. I’m using a NodeMCU ESP8266 with an IR led I took from the led strip remote. It only works if I’m pointing the led directly to the a/c receiver. At first I tought the led was too small, but I also tried with a bigger one that I took from an universal remote… still the same result. I might also add that the a/c remote works fine, I don’t need to point it directly to the a/c’s receiver to work. I could also put it in a place facing the ac directly, but I would need to power the mcu from a battery and I’m not sure how do to it or how effevtive this would be. The socket is right under the a/c so the distance is maybe about half a meter from the receiver, but not pointing towards it. Any ideea on how to strentghten the signal? I will post a picture of the circuit asap, I am currently writing on my phone. Thank you
//i’ve also added a dht11 temperature and humidity sensor, the temperature works fine but the humidity is aleays >80%, that can’t be right. Would it be the sensors fault?
Buy a proper IR receiver (photodiode) that you know the specs for, and a matching IR LED them design a circuit that optimises the components you’re using.
@PeteKnight does it matter what kind of receiver I use for reading signals, wouldn’t it read the same signal from the remote? I’m not trying to send the signals to the same receiver I’m reading them with, I’m trying to send them to the integrated a/c receiver, the same one that reads signals from the factory remote of the a/c unit.
Also, the output signal voltage of a salvaged IR photodiode (receptor) may not be compatible with your reception circuitry (NodeMCU) and it’s very difficult to know what you’re working with when you use salvaged components.
It depends on the electrical characteristics of the IR LED, but s you don’t know the specs of the salvaged LED you are using it’s very difficult to say.
Using the wrong wavelength of devices together will often give you some results, but you won’t get the range or reliability.
Okay, so I’d simply connect the LED to earth and whatever pin you’re using, with a 127 ohm resistor in one of those lines.
That resistor value is calculated to give the top-end of the power and voltage values from your table above, so you should err on the side of a slightly higher value if you don’t have exactly 127 ohm available.
Having said that, I drive my IR LEDs without any resistors, as they have a very low duty cycle.
On the left side is my NodeMCU with the IR led pointed towards the ac receiver, wich is somewhere in the round display on the right. The only way I can make it work is by removing the chrome trim, as the display/receiver is about 5mm in depth and the chrome trim seems to block the signal.
The picture above shows the only way I seem to make it work from the spot that I want to, but as you can tell is not really aesthetically pleasing.
I tought about installing my own receiver in parallel with the a/c’s, but I can’t tell wich of the 7 wires above is from the receiver. Any way to find out?
I still can’t tell why the a/c remote works well from any position…
The IR LED that you’re using as your transmitter to the AC may be operating on a different wavelength to the LED of the original remote and the receiver built in to the AC unit. Have you bought different wavelength IR LEDs and tried them?
IR remotes often send the command in a sustained burst that repeats the same command multiple times. If you’re sending it just once then it may not be being received as well. Some systems have some error checking built in, which require the same signal to be received multiple times within a certain time frame, so if you are sending the signal multiple times then either the timeframe may be too long or the number of repeats may be too short.
You can often do tests with the IR sniffing software and the original remote, when used alongside the AC unit to work out how many times the same signal has to be sent before the AC unit will accept the command.
You can use reflectors like the one below to focus the IR light from the LED so that more of it reaches the AC unit. Obviously aiming becomes more critical, but provided your transmitter is in a fixed location then it shouldn’t be difficult to do. You could use a visible light LED to help if you wished.
You could use multiple IR LEDs to give more power.
It looks to me like the IR receiver on the AC is probably behind the little grille at the 12 o’clock position. It looks a little bit obscured to me, maybe there’s a way to make it more accessible without it looking rough?
Personally I think point #2 is worth investigating first.
I have some blinds that are operated by a 433MHz remote and I spent a lot of time sniffing the protocol and reproducing it, only to find that the command needs to be received at least twice for it to be auctioned and sending it four times in quick succession is the most reliable way to make sure that it’s ‘heard’ by the blinds.
Your setup seems quite similar to what I’m doing here:
Since you get the AC to respond occasionally and are looking for greater range, here is a link to a circuit to give you better IR signal. Constant current infrared LED emitter circuit
It is on the AnalysIR site. The site also has a user forum with much info on IR signal decoding and transmission with MCU’s.
IR controllers have many variables beyond just the code itself. Different systems have different IR frequency requirements, the IR signal uses a PWM carrier signal that has defined frequency and duty cycle optimum. While these are often “common” values like 39,000 Hz for the PWM IR signal at 50% duty cycle and many systems seem to accept signals well outside of the nominal spec, these factors being sub optimum can limit your effective sensitivity.
AnalysIR sells a SendIR transmitter board ($10) that should provide a simple pre-built solution to the IR signal strength boost you desire. Trouble shooting the protocols, codes etc for your AC would seem to be a non-Blynk related issue. While interesting – some AC codes are up to 288 digits long and can’t be sent with unmodified libraries – these issues are better addressed elsewhere, I believe.
The remote is at the correct carrier frequency, with the correct IR frequency, with a wide angle LED emitter and likely boosting the signal beyond the current you can muster with your MCU. It is sending clean codes of optimum format and timing.
You did great to get this far with your scavenged parts etc. Reliable response requires defined start and stop signals as well as codes and often minimum timing between either repeats or new signals. Gets easy if you have a model that has been decoded by someone else, harder to do from scratch. Good Luck.
Thank you for the answers. I solved the problem, more or less, without changing the circuit or adding any extra code. I’ve made a cap-visor thingy out of plexiglass that acts like a wall, bouncing the signal to the receiver. Looks way better than completely removing the chrome trim as in the picture I posted in my previous reply. It might seem like an improvisation, but it works great, the a/c receives each signal. For now, I’ll leave it this way. I’ll post a picture below:
