andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sun Nov 20, 2016 5:19 pm

I have found something. I have measured the voltage of the 433-receiver. If no signal is present the receiver (measured to the ground) has stable 1,2V, approx. The needle is not moving. If I push the button on my garage opener the voltage drops to about 0V (I cannot measure that exactly because the needle is moving, but the decline is obvious - no increase). Can this be a reason why I cannot see anything? And how can I utilize this?
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
davidcoton
Posts: 5130
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Noise with 433MHz receiver

Sun Nov 20, 2016 10:40 pm

You mean the receiver output pin? It sounds like you need a voltage amplifier to interface to the Pi (GPIO expects 3V3 logic signal). This could be a single transistor with two resistors, or something like an op-amp or comparator.
Location: 345th cell on the right of the 210th row of L2 cache

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 5:57 am

davidcoton wrote:You mean the receiver output pin?
Yes.
davidcoton wrote: It sounds like you need a voltage amplifier
Can you point me to a website where I can find more information? Or tell me some keywords to look that up? Last time I was playing with transistors was thirty years ago...
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
joan
Posts: 14985
Joined: Thu Jul 05, 2012 5:09 pm
Location: UK

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 9:05 am

I think it is highly unlikely that any receiver marketed for the Arduino or Raspberry Pi needs a voltage boost. If the website has reviews have a look through for user comments.

User avatar
davidcoton
Posts: 5130
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 10:05 am

If the amp is needed (and I agree =with Joan it really shouldn't be if the receiver was intended for Pi), you just need a single transistor stage. But note it will invert the signal!

I don't have time to do a web search for you, sorry. But the basics haven't changed in thirty years -- I learnt about transistors over forty years ago!
Location: 345th cell on the right of the 210th row of L2 cache

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 10:15 am

joan wrote:I think it is highly unlikely that any receiver marketed for the Arduino or Raspberry Pi needs a voltage boost.
Well, the receiver was not especially produced for the Raspberry, maybe it is faulty (but how come that three of them have the wrong voltage: China?!). I have seen other people struggling with my setup and maybe this is the reason why! Still, that does not answer either Joans' nor davidcoton's remarks. On the other hand, I have never seen anything like a datasheet or whatever from the producer.

I will try to do my best and comment back. Maybe I start with a pull-up resistor first. If the voltage is between 0 and 1V, pulling up close to 3.3V should be easy.
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
davidcoton
Posts: 5130
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 2:07 pm

You can only pull it up if it is an open collector output. With anything else you risk damaging the receiver -- do you have a circuit diagram (probably a silly question given the origin)?
Location: 345th cell on the right of the 210th row of L2 cache

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Mon Nov 21, 2016 4:29 pm

I have nothing ("special offer, $2 incl shipping"...).

I was told I might have to pay tax, though :lol:
With anything else you risk damaging the receiver
I just read something about op am. This is too complicated for me. If I am damaging my receiver I will loose about 2$. That is worth the effort. Is a 10k resistor a good start? Or shall I take 100k?
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Thu Nov 24, 2016 6:08 pm

I managed to get a signal using a 38k resistor as pull-up, I checked several resistors and this seemed to be the best. Below is a sketch of my construction, also a picture of the signal (using piscope). Unfortunately, neither

Code: Select all

sudo ./RFSniffer
nor

Code: Select all

sudo ./_433D r2
will give me any code. Any idea what I can do?
Attachments
Bildschirmfoto 2016-11-24 um 18.54.27.png
Bildschirmfoto 2016-11-24 um 18.54.27.png (31.7 KiB) Viewed 7238 times
Kids.jpg
Kids.jpg (58.99 KiB) Viewed 7238 times
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sat Nov 26, 2016 12:49 pm

I managed to get a bit further. What I found out using piscope was that there is a lot of noise coming in (we had this before, but at that time it came from actually installing rpimonitor using the same port - rpimonitor is not installed this time).

I then lowered the noise using

Code: Select all

pigs fg 27 300
and I had to go to 300 to achieve this. 250 was not enough. But if I know

Code: Select all

sudo ./_433D -r 27 -d 5 -l 300
the noise comes back (I can see this on piscope) so that I cannot read the code. The code itself is "visible" in piscope, but looking closer reveals only "high" during the time I press the button. Help from experts is appreciated.
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sat Nov 26, 2016 2:38 pm

Maybe I have to correct myself again. I removed the resistor (joan told me...) and indeed got a signal that is more informative than the one above. See the screenshot below. There are some symmetries in there, but how come that this cannot be decoded? I am pretty sure this is Manchester. I also measured the time between zero and one and came up with something regular (second picture). Still a long way to go...
Attachments
time_between_zero_and_one.png
time_between_zero_and_one.png (54.38 KiB) Viewed 7166 times
signal.png
signal.png (15.95 KiB) Viewed 7166 times
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sat Nov 26, 2016 2:52 pm

maybe it is not Manchester? As far as I understand it there must to be only two "pulse-length" if it is Manchester decoded. Irrespective whether it is 00, 01, 10 or 11 only one short and one long pulselength is possible (see https://www.youtube.com/watch?v=XKtxxZ327UM). But I have several length. Here are the measured differences (I assume microsec, directly from piscope) together with the actual signal:

Code: Select all

0	211
1	720
0	210
1	260
0	210
1	65
0	100
1	175
0	135
1	715
0	185
1	750
0	215
1	725
0	215
1	250
0	200
1	70
0	10
1	20
0	105
1	15635
0	220
1	715
0	215
1	720
0	205
1	730
0	210
1	260
0	201
1	1199
0	210
1	725
0	195
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sun Nov 27, 2016 12:11 pm

New approach.


I installed pilight to observe the noise. It turned out that my XY-MV-5 receiver is not able to distinguish between noise and the signal from my garage opener. Therefore, this receiver is useless for me. I turned to the RBX12.

There, I do see something (see above) but I am not able to distinguish what is noise and what is signal. Pilight helped me to look at noise first. I recorded about five seconds of not-doing-anything with

Code: Select all

sudo pilight-raw -L

(the website is slightly confusing, it took me one hour to figure out that config.json has to be adapted) and the picture of those numbers can be seen below (running time of five seconds = x-axes, number of microseconds between HIGH-LOW on y-axes). Pure noise. :shock:
NoiseWithPilight-raw2.jpg
NoiseWithPilight-raw2.jpg (60.3 KiB) Viewed 7081 times
Another evaluation shows which numbers (in microseconds between HIGH and LOW) appear how often:

Code: Select all

micro-sec	distr.
50	32,30%
75	53,90%
100	66,53%
200	90,15%
300	97,51%
400	99,42%
500	99,87%
600	99,98%
700	100,00%
800	100,00%
This table is to be read as follows: Numbers between HIGH and LOW between 0 and 50 make up 32.30% of all data. Numbers between 0 and 75 make up 53.90% of all data. etc.

From this I guess the following: If I want an error that is less than 0.02% I should erase all numbers below 600 (of course, I might be able to erase the correct signal as well!). Erasing everything below 500 gives me an error of 0.13% (again, I might erase the signal as well). Maybe I am able to figure out the signal using this information. (Maybe not.)
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sun Nov 27, 2016 8:17 pm

Using the above I am now able to look at all time length that are 500mico-sec and above using (I was not able to formulate a more elegant code)

Code: Select all

sudo pilight-raw | sed -n -e 's/433gpio: //p' | sed -n -e '/[5-9][0-9]\{2,4\}/p'
Still a lot.
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Sun Dec 04, 2016 9:13 pm

I installed a low-pass filter (ATTiny as in the pilight forum, https://wiki.pilight.org/doku.php/low-pass_filter) but I was not successful. I could flash the processor, wiring is ok but still there is too much noise. I enclose picture with the pilight-raw signal received with the same superheterodyn receiver and the low-pass filter, it looks very similar to the picture above
NoiseWithATTiny Kopie.jpg
NoiseWithATTiny Kopie.jpg (59.97 KiB) Viewed 6944 times
I also did the same calculation as above about the distribution of noise. Here are the results:

Code: Select all

microsec	distr
50	42%
75	61%
100	72%
200	92%
300	98%
400	100%
500	100%
I do not see that this filter is better than the construction I had above. Given those numbers it even performed worse.

Does anyone has an idea how I can read the signal without noise? I do not see many choices anymore.
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 1:00 pm

Just learned from another post:

"Purpose of the filter is the reduction of interrupts and CPU load.

433 MHz transmission is based on OOK (ON-OFF-Keying) of a 433 MHz RF carrier, the result is a square signal with sharp rising/falling edges.

You cannot apply any kind filter to this kind of signals, you can only suppress rising/falling edges based on the knowledge that the minimum distance between rising/falling edges is appr. 100µS and simply ignore transitions with shorter durations.

Thus it is no surprise that you still have plenty of pulses."
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
joan
Posts: 14985
Joined: Thu Jul 05, 2012 5:09 pm
Location: UK

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 1:18 pm

I have lost track a bit about what you are trying to do.

I do not remember seeing anything which suggests you are actually detecting a signal from your transmitter(s).

Have you confirmed the transmitter frequency?

Have you connected a receiver tuned to the same frequency to the Pi?

Have you attached an aerial to the receiver?

Even if you are picking up lots of radio noise the difference in signal will be obvious during transmission. If you can see no difference you have done something wrong.

User avatar
pi-anazazi
Posts: 776
Joined: Fri Feb 13, 2015 9:22 pm
Location: EU

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 1:46 pm

https://wiki.pilight.org/doku.php/low-pass_filter

Basic 433MHz software (pilight-receive), how to build a low-pass filter and simple projects to test your setup :-)
Kind regards

anazazi

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 4:05 pm

joan wrote:Have you confirmed the transmitter frequency?
Maybe I start with the entire story from the beginning. I have two different senders that control the garage opener (see below the picture of a very cheap copy, the original is from Came) and a somfy roller shutter (second picture). My idea was to copy the signal of both senders and then send the raw signal using a raspberry that can be controlled via Wifi from outside my house (I do know of the risks). I am stuck with recording the signal, in particular because the Came sender does not send Manchester encoding, as far as I understand.
joan wrote:Have you connected a receiver tuned to the same frequency to the Pi?
I bought a cheap XY-MV-5V and got rid of that receiver because there was only noise. I can bought a more expensive (ok, 2 Euro) superheterodyn receiver and still saw some noise. Then, I read about the low-pass filter from pilight, got a ATTiny85, wired it according to pilight-CurlyMo's-website and connected it to my raspberry. I do receive some signal using

Code: Select all

sudo pilight-raw -L

now.
joan wrote:Have you attached an aerial to the receiver?
No, but in order to record any signal the distance between receiver and sender is less then 1 cm.
joan wrote:Even if you are picking up lots of radio noise the difference in signal will be obvious during transmission. If you can see no difference you have done something wrong.
Before I added the lpf I saw a permanent signal, see the above pictures in piscope. I do not see those signals any more. There are only several instances when a LOW-HIGH or a HIGH-LOW occures (maybe every second or so one change from LOW to HIGH or from HIGH to LOW). What surprises me is that Piscope seems to be freezing: Usually the clock below left is running in real time. It does not anymore. Also, using pilight-draw does not give me anything if I push the button on any of both remotes even if the distance is less than 1cm to the receiver.
Attachments
somfy.jpg
somfy remote
somfy.jpg (33.7 KiB) Viewed 6786 times
garage opener.jpg
garage opener remote (using a cheap chines copy)
garage opener.jpg (54.61 KiB) Viewed 6786 times
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
joan
Posts: 14985
Joined: Thu Jul 05, 2012 5:09 pm
Location: UK

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 4:12 pm

"What surprises me is that Piscope seems to be freezing".

That's just the way it works. If there are no level changes the display will stop updating.

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 5:13 pm

joan wrote:f there are no level changes the display will stop updating.
Again something learning, great - thank you!
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 5:27 pm

To wrap up: I just checked the signal from the cheap Chinese sender using pilight-raw and received a clear and unique signal. So in the end there are two lessons learned for me:
  • Get not the cheapest receiver (in particular, XY-MV-5 was bad for me). At least "superheterodyn" is necessary.
  • I also needed a low-pass filter (see above), this eliminated the noise. I burned one ATTiny before I got the wiring correct, so do not buy one, but three or so.
I cannot close the thread, but with your help I was successful. Thank you.
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

User avatar
davidcoton
Posts: 5130
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Noise with 433MHz receiver

Tue Dec 06, 2016 8:59 pm

andies wrote: I cannot close the thread, but with your help I was successful. Thank you.
Edit your original post and add the tag [Solved]. That's the nearest to closing the thread.
Location: 345th cell on the right of the 210th row of L2 cache

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver [Solved]

Sat Dec 10, 2016 6:21 pm

I noticed that using the recoding feature of Piscope gives me three instead of two signals:

Code: Select all

2000C1FF
2800C1FF
2000C0FF


Any idea what that is?
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

andies
Posts: 144
Joined: Mon Nov 11, 2013 8:12 pm
Location: Berlin

Re: Noise with 433MHz receiver [Opened again]

Thu Dec 15, 2016 9:26 pm

Unfortunately, I have to open the thread again. For some reason I was unable to get any signal. Learned from the posts above I did the following:
  • started

    Code: Select all

    sudo pigpiod
    and also watched PISCOPE
  • checked gpiotest, but

    Code: Select all

    Tested user gpios: 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 
    Failed user gpios: None
  • Measured the voltage directly at the RXB2-superheterodyn-receiver. Everything fine, although the voltage is not that distinctive, still 0.8V with a high signal and 1.7V with a low signal. As long as nothing happens (no button pushed) the voltage is stable at 1.7V and if I press the button it goes down to 0.8 and moves a little bit. So I suspect the receiver is working. I double-checked with another receiver: I got the same behavior.

Now comes the strange part. At the moment when I connect Data from the receiver with GPIO 17 or GPIO27 the voltage between GND and Data drops to zero (0.0V), regardless whether I push a button or make a cup of coffee (voltage meter impedance is 20kOhm)?! Disconnecting Receiver-Data from GPIO17/GPIO27 pushes the voltage back to 1.7V. Can anyone explain that to me?

The reason for not getting a signal seems to be based on this strange issue. I have no clue what is going on here. Is the RPi (3, newest jessie) broken?
raspberry B, Noir camera, Mac Book Air, iPhone, Bezzera

Return to “Troubleshooting”