grahamed
Posts: 277
Joined: Mon Jan 30, 2012 7:01 pm

Re: A few questions about my 433MHz automation plan

Tue Feb 09, 2016 9:14 pm

1) Never connect a 5V device to a GPIO directly. Either power the device from 3V3 and accept reduced performance maybe - how much range do you need? - or use a level shifter.

2) The tutorial correctly shows what the input looks like.

3) I use a PIC running at 5V, 3V3 reduces range somewhat and, strangely enough, so does 12V. Remember the GPIO is not an output til it is programmed as such. Include some resistance in the line at least. I have got away with it but YMMV.

Timing is an issue but generally near enough is good enough provided the error does not accumulate through the successive bits. I have two types of socket and the timings vary by a factor of two, imagine my surprise when I accidentally sent a correct code with the wrong timing and it still worked.

4) The TX is either on or off - continuous 433MHz carrier or nothing. If you send stuff it will interfere with a nearby FM radio so you can easily check that at least something is happening. An LED connected through a transistor to the 433 receiver is better, as is a 'scope, as is a logic analyser, but the Audacity thing will work just fine as it did at the earlier stage. I can't remember if the TX inverts but I do remember having the entire thing upside down at one point.

Finally you will need an aerial/antenna. 17cm or so of straight wire. Coiled looks nice but savages the range.

Good luck.

asandford
Posts: 1997
Joined: Mon Dec 31, 2012 12:54 pm
Location: Waterlooville

Re: A few questions about my 433MHz automation plan

Tue Feb 09, 2016 10:13 pm

Here's a really good tutorial. I used the sniffer to get the codes for Maplin Remote Controlled Mains sockets

grahamed
Posts: 277
Joined: Mon Jan 30, 2012 7:01 pm

Re: A few questions about my 433MHz automation plan

Tue Feb 09, 2016 11:14 pm

1) Yes, though I cannot say whether your receiver will work on 3V3 - mine do.

2) The receiver ramps up its sensitivity when there is no signal until it starts to pick any old noise. When a signal comes it is much bigger than the noise so the output becomes clean. The gain remains low for quite a few milli-seconds, longer than the inter-packet gap at least, before climbing back into noise.

3) Depends on the site, my walls have foil behind the plaster board so signal strength is always a problem. As I say I use a serial connected PIC to generate the codes so 5V is no problem, nor is timing.

The tutorial asandford mentions looks good, and there are others. I like the Audacity sniffer method - though in truth I use a logic analyser, about £10 on eBay - as you can see the signals. Magic Ninjablock methods are OK when they are OK and if they are not you haven't a clue. Horses for courses.

I use the Maplin sockets (often found on offer at half price) and ones made by Status. The Maplin ones are better as they do not forget their code whenever they powered down - code is set by small rotary switches.

asandford
Posts: 1997
Joined: Mon Dec 31, 2012 12:54 pm
Location: Waterlooville

Re: A few questions about my 433MHz automation plan

Tue Feb 09, 2016 11:42 pm

grahamed wrote: I use the Maplin sockets (often found on offer at half price) and ones made by Status. The Maplin ones are better as they do not forget their code whenever they powered down - code is set by small rotary switches.
I recently got the £30 5 socket set for ~£15. I actually contol them with the transmitter hooked up to an arduino as I found the timing more stable at the time (tested on a Pi B before the B+ and Pi2 were released), and could supply more current and voltage from the pins (I only switch it on a second before I transmit data and off after, as it swamped my weather station if left transmitting all the time).

I have no experience of Ninjablock stuff, but I found that tutorial (and the one for Arduino) useful, short and sweet; fortunately it worked (althought looking at the comments I may have been lucky).

asandford
Posts: 1997
Joined: Mon Dec 31, 2012 12:54 pm
Location: Waterlooville

Re: A few questions about my 433MHz automation plan

Sat Feb 13, 2016 11:04 pm

george7378 wrote:OK, it looks like the above page is for a slightly different design of Rx/Tx so I can't really trust it for the ones I've got. Still, I'd be really interested to know if anyone knows why the majority of tutorials are just wiring both the Tx and Rx to the Pi using 5v without any sort of protection?

It would also be brilliant if anyone's actually used a voltmeter or oscilloscope to measure the voltages across the data pins for the modules when they are powered by 5v.
As a complete untested hack, you could probably use an LED to drop some voltage and give a visual indication?

maurice1
Posts: 37
Joined: Tue Mar 05, 2013 8:55 am
Location: Dublin

Re: A few questions about my 433MHz automation plan

Tue Feb 16, 2016 6:04 pm

I tried this https://github.com/dmcg/raspberry-strogonanoff and it works fine for the maplin http://www.maplin.co.uk/p/remote-contro ... pack-n38hn


However it does need wiring pi
In the end I just used an arduino uno with ethernet shield and set it up as a webserver.
I use the cron on the pi to turn them on and off like a time clock.

boyoh
Posts: 1326
Joined: Fri Nov 23, 2012 3:30 pm
Location: Selby. North Yorkshire .UK

Re: A few questions about my 433MHz automation plan

Tue Feb 16, 2016 8:20 pm

In your power sockets switching project, I suggest you use Opto Isolators ,as voltage levelers
They are very fast switching and noise free , And will give you total isolation ,at different voltage
levels.
BoyOh ( Selby, North Yorkshire.UK)
Some Times Right Some Times Wrong

asandford
Posts: 1997
Joined: Mon Dec 31, 2012 12:54 pm
Location: Waterlooville

Re: A few questions about my 433MHz automation plan

Sun Feb 21, 2016 5:54 pm

george7378 wrote:Thanks for the further replies :) I've managed to use my components with nothing but RPi.GPIO to create a working system. I powered the receiver using 3v3 to get the codes which were easy to manually transcribe. I then used the transmitter powered by 5v to send the same codes with time.sleep(). The shortest sleep I need to do is 0.45ms, and I send each code 10 times (which I found is equivalent to holding down the button on the remote for a perfectly normal amount of time) and it seems to be working every time :) I'll try and post a write-up when I get the chance!
Glad you got it all working. The Pi's useful for reading the codes, but a bit overkill (for my use cases), so I use an ardunio (as maurice1 did, but with radios) and leave the 'heavy lifting' to the Pi.

asandford
Posts: 1997
Joined: Mon Dec 31, 2012 12:54 pm
Location: Waterlooville

Re: A few questions about my 433MHz automation plan

Mon Feb 29, 2016 11:06 pm

george7378 wrote:Here's a write-up if anyone's interested:

http://www.instructables.com/id/Super-S ... utomation/

Thanks again for the help :)
Nice and detailed write-up.

Return to “Automation, sensing and robotics”