Any plan for Rm2 RF feature

[tonyzhong20]

Rm2 support RF, would be great to have it

[aldodemi]

You already have it, you can learn/register RF code exaclty as an IR code

[tonyzhong20]

Tried several time, but I'm not able to register RF code

[tyjtyj]

RF 433 and 315 works on RM2.

Are you able to learn IR code ?

[flyingartichoke]

I'm having trouble with this as well. I can successfully learn an IR code with my Broadlink Pro but I cannot get it to work with an RF remote. Are there any suggestions/guides for how to do this?

When learning new RF codes through the Broadlink app itself, it had me do an RF frequency scan followed by learning the actual code. Is there an analog to that process here?

Or should I be using my phone + Broadlink app to generate the codes that are being learned by the python script?

[tyjtyj]

I have 3 rf item
1x 315
2x 433

I use this code

Create a dir/folder name 'codes'

def learn(myrm2):
	while True: 
		codeName = raw_input("DeviceName :")
		if codeName == "":
			break
		while True: 
			myrm2.enter_learning()
			sys.stdout.write('Listening')
			for i in range(5):
				time.sleep(1)
				sys.stdout.write('.')
			ir_packet = myrm2.check_data()
			if ir_packet != None:
				break
			print "Nothing receive, try again"
			# record learned hex code to file
		myhex = str(ir_packet).encode('hex'); 
		print ''
		print str(ir_packet).encode('hex')
		f = open('codes\\' + codeName +".txt",'w')
		f.write(myhex)
		f.close()


myrm2 = broadlink.discover(1)
myrm2 = devices[0]
myrm2.auth()

learn(myrm2)

it will save the codes to codes/[devicename]txt

hex_data = [yourhexdata from file]
myrm2.send_data(hex_data.decode('hex'))

[technicalpickles]

As said, RF support is already built in.

However, when I setup my device originally, I used the e-control app initially. There was a firmware update that said it improved RF learning capabilities.

It is working for me. In practice, I've found it necessary to check_data() in a loop until a code is found.

[NightRang3r]

HI All

I wrote a quick and dirty script to dump all codes from the Android e-Control app db file.

https://github.com/NightRang3r/Broadlink-e-control-db-dump

You can view the full discussion that led to the creation of this script here:

ericmagnuson/rmcontrol#4 (comment)

[tecteun]

I found that learning a 315Mhz code only works through the app, there are two options in the app:

• learn
• sweep / learn
  The 315Mhz RF only works with the sweep/learn option.

How to trigger the sweep / learn from the api? (right now I've worked around this by copying the hex codes from the android db dump)

Note, this sweep/learn was not available in the old rmpro with rm2 hardware, but is on the new rm3 hardware.

[luketoh]

Apparently I have a Broadlink RM3 Pro instead of the usual RM2 Pro. Does this python script work with the RM3 Pro? I tried running a simple python program to scan for it, but I don't think it detects the RM3 Pro. Can someone help? Here's what I got...

pi@hassbian:/home/rmcontrol $ python broadlinktest.py
discover
Traceback (most recent call last):
File "broadlinktest.py", line 8, in
devices[0].auth()
IndexError: list index out of range

here's the program i used from someone else...

#!/usr/bin/python

import broadlink
import time

print "discover"
devices = broadlink.discover(timeout=5)
devices[0].auth()
print "enter_learning"
devices[0].enter_learning()
print "sleep"
time.sleep(5)
print "check data"
ir_packet = devices[0].check_data()
print str(ir_packet).encode('hex')
devices[0].send_data(ir_packet)
print "wait"
time.sleep(2)
devices[0].send_data(ir_packet)

[pavram]

Someone might need to dump the data from the app communicating with an RM3 to get the "Sweep Learning" command structure. It sounds like RM3 pro might have a few new commands that work in a slightly different manner.

The RM3 pro should still be detected by the Discover requests however, since discover is common among all the devices.

@luketoh: it does appear that your RM device isn't being detected.
(the devices[0].auth() command is trying to get the 0'th [first] device from the discovered devices, and that device is "null" (or not found)).

In your case I would make sure the device was on the same network as the computer you are using to attempt to detect it.

Easy way is to use the official app and associated guides to make sure it is on your home wifi, and make sure your computer is on the same network, and work from there.

[lprhodes]

RM3 (and possibly RM2) works by first getting the frequency and then learning the specific code. This code only performs the learning from the look of it.

Working on getting it working on the JS version of the code (broadlinkjs) so perhaps you can get it working in python from that once I'm done - I'll keep you posted

[lprhodes]

Struggling to get the RF sweap learning working @mjg59 perhaps you can see something I'm missing

Wireshark shows the following hex as being sent when I first enable the RF sweep:
5aa5aa555aa5aa5500000000000000000000000000000000000000000000000029d200002a276a00b58128d7e734ea3401000000c8be0000065d76df227ec2b694e98c88215afff3

The command part of the hex is 65d76df227ec2b694e98c88215afff3
Deciphered that's 9fa6f2de6d22144b0b1ce7d551a8c485

When I try to reuse that the code produces the following hex:
5aa5aa555aa5aa550000000000000000000000000000000000000000000000000dd100002a276a00020028d7e734ea3402000000d7c60000065d76df227ec2b694e98c88215afff3

As far as I can see, the the only differences between the two are the id (which I assume is the ID for the device assigned to the device sending the hex codes to the RM device) the count (which is always going to be different between the two) and the checksum - which would of course be different.

The problem is, my own command fails to enable the RF sweep mode

[lprhodes]

Made the tiniest bit of progress - the hex payload for starting the scan is 19 with a command of 6a

[lprhodes]

Cancelling looks to be a hex payload of 1e