• 10 dec 2017: forum version update. In case of issues use this topic.
  • 30 nov 2017: pilight moved servers. In case of issues use this topic.
Hello There, Guest! Login Register


Band pass-filter (new version)
#1
You can discuss all your experiences with the new band-pass filter here.

To create one yourself, you can use the instruction found here
 
Reply
#2
Is it possible to use this with mercuri0s attiny_433_prefilter?
And how should the hardware.json look like?


lcd4linux
 
Reply
#3
It actually is the mercuri0 code but adapter a bit by me. I will post the source later.
 
Reply
#4
I've received the new LPF today. CPU drops from 30% to about 3% using this new LPF. It works OK for my normal Elro switches, but my doorbell (Kaku) is not received anymore. But since I only use receiving for my doorbell, not very usable for me yet.

This is the code for the doorbell:

Code:
    "code": {
        "id": 9906274,
        "all": 1,
        "state": "up"
    },
    "origin": "receiver",
    "protocol": "arctech_screens",
    "uuid": "032C-00-00-2C-000300",
    "repeats": 6
}
Let me know if you need anything else.

I just connected the other pins for the SPI firmware upgrade, and now it works for my switches as well as the doorbell. CPU is now about 1%. Thanks!
 
Reply
#5
I received my new (attiny) filter today. Seems to be working fine. I did,however, not notice anything about better range or something, i did notice a cpu drop from 30% to about 3-5%. Thats a nice improvement.

One thing i did notice was i still see a lot of 'static' when i check pilight-raw.
I would have suspected it to be silent when nothing was coming in, but i guess it just filters too short pulses or something?
 
Reply
#6
Quote:I did,however, not notice anything about better range or something
That's not what a filter is meant for. For a better range you need an antenna or better receiver.

Quote:One thing i did notice was i still see a lot of 'static' when i check pilight-raw.
The current filter is programmed to support the impuls protocol. That's the hardest one to receive. If you are not using impuls, the filter can be more strict (and more silent). However, the current noise is no issue for the Raspberry Pi. When i programmed it for KaKu, the CPU usage didn't change. The current settings are the best balance.

The future firmware should be able to change the filter settings dynamically based on the protocols included.
 
Reply
#7
Bug 
I received and hooked up the new ATTiny based filter, but am experiencing problems. I have the following remotes to test with : KaKu YCT-102 and Elro AB440R. The problem I experience is the same for both senders.

The filter seems to do its job, because I see pulses in pilight-raw when I press a button and a I see significantly less noise with the filter hooked up. However, pilight-learn does not recognize any button it asks me to press when the filter is hooked up. Without the filter it works as expected and outputs:

http://pastebin.com/Gup3nSP6

I created some logs and videos to illustrate the problem. In the videos I show pilight side by side with another application using different hardware that shows messages when I press a button.

- Learning session without filter (works): http://www.youtube.com/watch?v=051cv7lqDH8
- Learning session with filter (fails): http://www.youtube.com/watch?v=owvX7CGKEMs
- Raw output with filter, where I see some noise but also a pattern when I press a button: http://www.youtube.com/watch?v=T9LzbQXIzoo

And the logs:

- Raw output with filter : http://pastebin.com/CVx5HFma
- Raw output without filter : http://pastebin.com/P9eQJr6F

The old filter that pilight used in previous versions (with the opamp and schmitt-trigger) also didn’t work right away for me. I had to modify that circuit slightly by choosing a different resistor value. This worked, but would lead to more CPU usage than the original design, yet be far more efficient than without a filter.

I like the new filter approach better, so I hope I can somehow make that work. If needed I can flash an ATTiny myself. Any help would be greatly appreciated though Smile My knowledge of C is unfortunately not that great.
 
Reply
#8
I already have a topic about filter experiences and this should be posted in the electronics section. So please search before you post.

I don't experience any issues with the raw code you gave. When i extract the kaku pulses and feed it into pilight-send -p raw -c "..." it just gets recognized as:
Code:
{
        "code": {
                "id": 0,
                "unit": 6,
                "state": "off"
        },
        "origin": "receiver",
        "protocol": "arctech_switches_old",
        "uuid": "0338-00-00-38-000300",
        "repeats": 1
}
 
Reply
#9
Thanks for your reply CurlyMo. I'm not sure what I was doing wrong, but picking things up again today I'm getting good results, so clearly I was doing something wrong when I posted my previous comment although I'm not sure what it was unfortunately.

Your hint on sending the captured pulses is also nice and I thought I'd write a script to convert the results of pilight-receive to a send script. I came up with the following.

Capture:
Code:
pilight-receive >lamp_off

and then convert its output using this script I named convert.sh:

Code:
#!/bin/bash
if [ "$#" -ne 1 ]
then
  echo "capture a pilight-raw stream to a file and then specify its filename to me to make a send script"
  exit 1
fi

DATA=`cut -d ' ' -f 2 $1 | awk '{printf("%s ",$0) } '`
echo "pilight-send -p raw --code=\"$DATA\"" >$1.sh
chmod +x $1.sh

executed like this:

Code:
./convert.sh lamp_off

results in the following script: http://pastebin.com/mayfUVhY

but I guess the result is a bit too verbose (i.e: there's too much data), because when I execute that I get:

Code:
./lamp_off.sh: line 1:  3216 Segmentation fault      pilight-send -p raw -c "356 360 1039 367 10..............................."

My hunch is that I'm just passing too much data.

Not that it's important, because everything I'm interested in seems to work fine, but it would be kind of nice to make this work somehow.

Great work on the new filter! I've been following the project for quite a while and this is the first time my high expectations are completely met Smile Great work!
 
Reply
#10
Quote:Great work on the new filter! I've been following the project for quite a while and this is the first time my high expectations are completely met
One thing i was aiming for this whole time Wink
 
Reply
  


Possibly Related Threads...
Thread Author Replies Views Last Post
  Bandpass filter problem mayonezo 13 1,862 01-07-2018, 11:29 PM
Last Post: mayonezo
Question Passive Component Band-pass filter ripper121 8 1,250 12-16-2017, 01:38 AM
Last Post: wo_rasp
  Filter kit in 315 receiver? ceandre 1 1,024 04-14-2017, 10:10 PM
Last Post: wo_rasp
  Low Pass Filter and Receiver in Series? re-post Gustavo Woltmann 1 1,152 03-08-2017, 05:25 PM
Last Post: pilino1234
  Filter settings in config.json Saiyato 100 22,011 02-22-2017, 08:24 AM
Last Post: wo_rasp
Rainbow Lowpass filter - how DOES it work? DE8MSH 6 2,598 10-07-2016, 09:25 PM
Last Post: wo_rasp
  Low Pass Filter and Receiver in Series? Hauke 3 2,287 08-30-2016, 08:25 AM
Last Post: wo_rasp
  New filter version curlymo 0 1,203 04-02-2016, 08:05 AM
Last Post: curlymo
  Band-pass filter parts curlymo 60 55,979 01-08-2016, 12:23 AM
Last Post: gneandr
  low pass filter francois.leprieur 3 1,843 12-20-2015, 10:18 AM
Last Post: curlymo

Forum Jump:


Browsing: 1 Guest(s)