wireless transmission of alarm condition

[g2c]

Hello,

I want to protect my home with PIR modules by the windows and am looking for a mean to transmit their state (binary state) via RF. The solution should

allow the controller (Arduino board) to distinguish the various PIR location
allow the controller to know that the transmitter is healthy, e.g. the transmitter to send keep alive with minimal impact on the battery
well, of course it should be small and cheap

Any idea?

Thanks in advance,

Guy

 

[Nigel Goodwin]

Licence free radio modules, and micro-controllers to control them.

 

[Mike odom]

I use Linx RF transmitters and receivers.

 

[Mike odom]

Also, I've bought PIR sensors before, and they were like $35 from Digikey. On my last project, I got clever and bought a motion sensor lamp for like $12 and took the PIR out of it!!!

 

[g2c]

Thanks. Googling your reply returned Rf Solutions Zulu's Are there other you think of?
...and use sleep mode with the RTC running to fire the keep alive signals?
This looks promising and i was not aware of these goodies

Also, I've bought PIR sensors before, and they were like $35 from Digikey. On my last project, I got clever and bought a motion sensor lamp for like $12 and took the PIR out of it!!!

But you can have them on eBay for $1.5: look there for pir detector. Anyway, what do you mean by you took the pir out??

 

[g2c]

I use Linx RF transmitters and receivers.

Can you please give some details?

 

[Mike odom]

Can you please give some details?

https://www.linxtechnologies.com/en/home

 

[Mike odom]

But you can have them on eBay for $1.5: look there for pir detector.

I don't buy anything on ebay... I have to have a reliable source for production parts.

 

[jpanhalt]

Did yo see this sticky: https://www.electro-tech-online.com/threads/cheap-ebay-rf-modules-made-easy-and-fast.134224/

It is pretty much step by step.

John

 

[g2c]

https://www.linxtechnologies.com/en/home

Sorry, i mistakenly read Linux

 

[g2c]

Did yo see this sticky: https://www.electro-tech-online.com/threads/cheap-ebay-rf-modules-made-easy-and-fast.134224/

It is pretty much step by step.

John

Yes indeed, these look promising. I wonder if each pair has it's own distinct frequency. Otherwise, one may need to use different modulation to separate the trnsmitters

 

[jpanhalt]

The usual way to address multiple modules is to use an address code, not different frequencies. Both of these sources provide good descriptions:

https://www.eng.uwaterloo.ca/~tnaqvi/downloads/DOC/sd461_462/Stamp_RF.pdf
https://www.pololu.com/catalog/product/1337

John

 

[alec_t]

Perhaps you could hack a wireless doorbell? Mine (~£8) allows switch selection among 8 (IIRC) addresses.

 

[g2c]

The usual way to address multiple modules is to use an address code, not different frequencies. Both of these sources provide good descriptions:

**broken link removed**
https://www.pololu.com/catalog/product/1337

John

Thank you John, I am sorry but i don't understand what you call "use of address code" Do you mean the the receiver end pols the transmitters sequentially by addressing them? This supposes that the transmitters which are coupled to the IR sensors also have (an always on) receiver and hence consume quit a significant power . The u of Waterloo shows how to better synchronize Rx and Tx. The Wixel thing presentation seems to recommend different channels (they have 256) for a multi-station system.

I definitively think that with multiple transmitter system, where each can start transmitting independently, there should either use separate frequencies or a kind of CSMACD. Otherwise, the receiver may very well miss messages in case of simultaneous transmissions. Please correct me if i'm wrong

 

[g2c]

Perhaps you could hack a wireless doorbell? Mine (~£8) allows switch selection among 8 (IIRC) addresses.

Thanks dear, Excellent idea. What you call 'addresses' must be frequencies settable by a PLL variable divider. I guess its true for car remote key too(?) It's very much what i need because the it sensor just transmit 'one bit', much like pressing the bells or remote button. What kind of chip are these using?

 

[jpanhalt]

Thank you John, I am sorry but i don't understand what you call "use of address code"

Address coding is described in the links I gave. Data are transmitted digitally. Thus, one transmitter may transmit, "123" + "1" , meaning it has detected motion. Another transmitter may transmit "122" + "1" to indicate its position and detection of motion. And so forth.

Do you mean the the receiver end pols the transmitters sequentially by addressing them? This supposes that the transmitters which are coupled to the IR sensors also have (an always on) receiver and hence consume quit a significant power .

Not at all. I don't see how you draw that conclusion. Most important, how would using multiple frequencies, as you suggest, avoid the same issue. Using addresses requires that only one receiver be on. Using different frequencies requires that one receiver for each frequency be on.

The u of Waterloo shows how to better synchronize Rx and Tx. The Wixel thing presentation seems to recommend different channels (they have 256) for a multi-station system.

How is a "channel" any different than an address?

John

 

[Nigel Goodwin]

Thanks dear, Excellent idea. What you call 'addresses' must be frequencies settable by a PLL variable divider. I guess its true for car remote key too(?) It's very much what i need because the it sensor just transmit 'one bit', much like pressing the bells or remote button. What kind of chip are these using?

No, they are 'addresses' - they send a digital data string (somewhat like a TV remote) that consists of an address (to identify the particular device) and a command (obviously a pretty simple one for a bell).

They don't alter frequencies, which would be pretty pointless - and there isn't licence free space to do so.

 

[g2c]

Address coding is described in the links I gave. Data are transmitted digitally. Thus, one transmitter may transmit, "123" + "1" , meaning it has detected motion. Another transmitter may transmit "122" + "1" to indicate its position and detection of motion. And so forth.

i now understand that you meant 'source address': all transmitters use same frequency, each transmitter has it's own source address and precedes it's "message" with a source address. There is one receiver who listens and can tell who's the transmitter by listening to the source address. I believe that this can work well if there is some entity doing the "collusion detect" part of the algorithm. E.g. for car's key, if pressing once doesn't work, the human who executes the "CD" part, will press again until the door opens. In this project i don't see who'll be doing the CD: If a burglar and his colleague activates simultaneously IR sensors 1 and 2, both transmitter will transmit simultaneously the message <srce_addr(i)><infraction>, i=1,2 but the colliding messages will be corrupted. Similarity if in absence of alarm both happen to transmit simultaneously <srce_addr><keep_alive(i)>i=1,2, the receiver might not receive either and consider that the receivers had been broken by an aggressor. Again, please correct me if i'm wrong

How is a "channel" any different than an address?

According to the CC2511F32 datasheet, these are specral channel, ~250KHz appart

 

[g2c]

No, they are 'addresses' - they send a digital data string (somewhat like a TV remote) that consists of an address (to identify the particular device) and a command (obviously a pretty simple one for a bell).

They don't alter frequencies, which would be pretty pointless - and there isn't licence free space to do so.

ok, in the case of a doorbell i see no problem in distinguishing the transmitters by a source address and not by frequency: In case of collision (two neighbors pressing their respective bell simultaneously corrupting each other's transmission), all one has to do is press the button again

 

[jpanhalt]

Yes, there are multiple frequencies in the 2.4 GHz band. Let's assume you have an installation with 20 sensors and each is on a different frequency. You will need 20 fixed-frequency Wixel's to detect them, as there is no scanning ability built in. Now, since you are dealing with smart robbers, they bring a blocker that transmits on each frequency. Therefore, you need an acknowledgement sent by the main receiver to the triggered sensor to say alarm received. You will need 20 transmitters to do that (or a system to scan the transmitter). Of course, the robbers don't worry about complying with the FCC, so they broadcast continuously and block the acknowledgement too.

How is that operationally different than having all transmitters and receivers on the same frequency using different addresses. For security, you use a typical transmitter/receiver combination that scans and binds to each other and frequency hopping (See: Hedy Lamarr and composer George Antheil, "Secret Communications System" , https://en.wikipedia.org/wiki/Frequency-hopping_spread_spectrum ).

Allowing each station to have its own coded identifier (maybe even 64-bit) will be more effective at preventing a would be robber from sending a fake "all is OK disregard alarm" than simply using 250 distinct frequencies. Of course, there will need to be a handshake between central receiver/transmiter and each sensor receiver/transmitter combination. But, you will not avoid that by using different frequencies either.

Perhaps the most productive next step would be for you to sketch out your proposal with a little more detail than given in your first post, and let us contribute to improving that design.

Edit: I would also consider a "blocking detector" so the alarm would sound if an expected ack signal is not received. My simple alarm system has that ability. The battery-operated senders operate for >3 years without needing to replace the batteries.

John