Cheap 433mhz rf modules.

[dr pepper]

I have a couple of these, I can just about get them to work 6 feet away from each other, instead of 100m.
They both have 1/4 wave wire whips, and long wire ground planes, both battery powered.
The tx shows 433.15mhz on my freq counter, I have a rf sig gen but I cant get the rx to give an o/p from it, I guess the correct modulation is required for that.
Odd thing is the tx is deffo 433 as its marked on the can and the counter confirms that, the rx says 408, could that be 408mc, wouldnt be that surprising knowing some ebay sellers.
How can I tell what the rx is, and does anyone know if the sealed up core on the rx board is a freq adjustment?, or something else, there only looks to be 2 wires so its most likely part of an osc.

 

[MikeMl]

How do you know that the Rx is not hearing the Tx?

The Rx is a super-regen. These have only a single-tuned LC circuit for selectivity, so are as broad as a barn door. They tend to lock on the strongest signal in the vicinity, even if it is several MHz removed from the design freq. This can be a problem if there is an unrelated high-power Transmitter near by. For example, I have a ham transceiver that puts out 45W into a gain antenna between 420MHz and 450MHz. That is why I use the 315MHz version of these modules...

The "output" is usually randomly pulsing when receiving just noise (TX not keyed). It takes a few 100s of mS of continuous carrier for the receiver to adjust its detector "level", and then the RX output follows the on/off keying rate of the transmitter, but there is a max limit on the duration keyed (TX on) or unkeyed (Tx off). The only way you know if the RX is actually receiving the TX is if the data pattern out of the RX matches the OOK signal to the TX.

Roman did a very nice analysis of the timing requirements. You can read it here.

 

[dr pepper]

I wondered if'd be a poxy regen.
I tried the modules with manchester coded data with preamble, still the same.
I'm starting to think these things are for turning on lights & opening garage doors only.
OK then I'll see what roman has to say.

 

[tunedwolf]

I once made the mistake of using 30 odd so called "matched" pairs of these for a customer's project. I think they were RF Soloutions ones, saw based, and were just becoming popular at the time. They worked perfectly in my workshop, over a distance of about 30' and seemed happy to go beyond that. Less than a week after install, I had to redesign the system to use decent quality FM transceivers. These silly little 433Mhz things were great in the workshop, but wouldn't have made it across your modern electically noisy living room in practise. As for the "matched" part, I think they must have been referring to the numbers stamped on them
Every one of them required space tuned also or they went miles off frequency. In all, just a headache in the making.

 

[dr pepper]

Hmm junk then.
One last thing I'm gonna try is I saw some on the net with tants across the supply, I havent got any tants I dont think but I have some quality cans so I'll try that, if no worky I'll abandon these horrible little doofers.

 

[dr pepper]

Dr capacinator to the rescue.
I put the said 10u across both modules, zip no change, so I moved one to the other side of the living room further away from the tx which is in the garage, and it works well even at 4kbits, 10x the speed of my last effort.
The pc intereferes with the rx, no surprise, but yester when I powered the rx from a batt well away from the pc and a lot nearer to the rx it wouldn twork, so the 10u is vital, both units are now operating of wall warts, and they are ball park where they are going to be when completed.

Romans site is interesting, however virtual wire for the arduino as far as I can tell does a similar thing, it sends a preamble to adjust the rx gain, and detects pulse position not length, on top of that it does error checking.

In this page the first link is an arduino library for several rf modules, the one for these is the ask tx and ask rx, these librarys do the preamble and encoding etc for these ask rf modules.
https://www.airspayce.com/mikem/arduino/RadioHead/

 

[granddad]

I use Aurel RTX 5 volt module ASK serial at 2400 baud with ceramic antenna tested at apx 30FT OK. Not tried but these look economic !
**broken link removed**

 

[dr pepper]

Yes I'd seen those modules, I might just get a couple of those just in case, the benefit with them is they are a transceiver not single ended.

These cheapo 99p a pair ones are working well now, 20m through 6 walls and a car, I dont think a zigbee would be able to do that at 2.4 gigs.
I'd walk down the road with the tx to see how far it'll go, but its pouring down.

 

[granddad]

Only preamble I use is H' 00,55,AA, then rx ID , 20 byte message, message number and, sum check. not using any libs, C code in a PIC24FV32KA102 inverted serial from UART 24oo Baud . RTX has rather critical switching TX-RX-PowDn etc , I loose a few TX's I may try more preamble H"55 's see if better performance.

 

[dr pepper]

I get the 55 and aa, why the 00?
I wondered about the uart, however seeing as I'm playing with an arduino its no more complex for me to use a pin directly.
Romans article is interesting where he points out that its the on time that gets mashed with noisy signals, detecting the carrier off point rather than the centre of the carrier on makes an improvement, not sure about avr's but pic uarts use a majority oversample, a 1 or 0 being decided on which is the majority, if the duty cycle gets messed up then that could lead to corruption.
That said FM modules are going to be way more reliable from the start.

 

[granddad]

Dear Doc ... the H'00 results in a TX carrier burst for 9 bits ( start + 8 ) seems to wake up the listener . will have to check on PIC24 Uart sample rates always thought it was centre .... ( stopped raining here , my allotment needed it )

 

[dr pepper]

Doh, 0 is carrier on, your allowing the gain control to adjust, its Friday I'm slow.
Not sure about pic24, pic18 does multi sample and average.
Stopped raining here too, my forest flame is trying to turn red, and the teasel looks like its after the postie.

 

[granddad]

Hi Doc... these RTX modules dont have any built in com system / protocol , the pic uart TX pin is inverted ( register bit ) so my H'00 becomes H'FF at the RTX pin so carrier is ON , when I was developing ( playing with ) the interface I had a 433 RX module connected to an 74HC04 hooked up to a USB / serial chip and PC running Real-terminal so was able to monitor 433 traffic . It now looks after my greenhouse tomatoes......

https://uk.farnell.com/aurel/rtx-mid-3v/rf-mod-txrx-ask-3v-433-92mhz/dp/1699468

 

[dr pepper]

Yep that makes sense.
I thought they were a bit expensive, but then having dds, narrow band, 9k6 bitrate, and european approval you can see why.
You'd have a chance bunging a ce sticker on that, my rfdinmakers would be a nightmare to certify.
The mrs has several kinds of tomatoes in the greenhouse, and a few chillies, no electronics there though.

 

[Nigel Goodwin]

You really need to use Manchester coding (or similar) with the cheap modules, and NOT to use the hardware UART's - if you are happy with the much poorer reliability of RS232 over an unsuitable radio link, then use a software UART (as this allows you to invert the data).

I've always used the slightly more expensive FM modules, rather than the poorer AM ones.

 

[granddad]

Hi doc They were £10ish when I got them they don't seem to have the 5volt ver now , I had a go with the small separate RT / TX (ESR components ) but lost too much grey hair ! I found tomatoes did better with regular watering / humidity so my remote device pumps water as programmed reports temperature in/out and humidity, hangs on a 12volt battery / solar panel , i just have to provide water / feed. oh.... and tlc

 

[granddad]

Nigel , pic24F uart can be set for inverted TX and RX , but a 74HC04 works OK to ,

 

[Nigel Goodwin]

Nigel , pic24F uart can be set for inverted TX and RX , but a 74HC04 works OK to ,

As can various other higher spec and more modern devices, obviously adding external hardware inversion works as well - but you're adding extra hardware for little gain.

Basically standard 'RS232' isn't suitable, some type of coding is really required as accurate timing doesn't travel well via RF (or IR).

 

[dr pepper]

Hi Nige,
I'm using the arduino virtual wire library, it does a preamble. uses manchester code, and generates a checksum.
I have used uarts before with sucess with rf, you can still do a preamble and if you send the same byte twice with one inverted you still get reasonable dc average, the range is less than manc code, but if I was just looking after tomtaoes 20' away from my pc I think it'd be fine.

I'd thought about a dht device in the greenhouse to report temp/humidity, but what would I do with that data, I was gonn ahve a servo to open the window when it was hot, but I got a pneumatic opener from wilko for a tenner which works fine.

 

[Nigel Goodwin]

Hi Nige,
I'm using the arduino virtual wire library, it does a preamble. uses manchester code, and generates a checksum.
I have used uarts before with sucess with rf, you can still do a preamble and if you send the same byte twice with one inverted you still get reasonable dc average, the range is less than manc code.

My point is, why use something totally unsuitable, when you could just do it properly? - a hardware UART isn't any advantage (and has plenty of disadvantages) for running RF modules.

I suggest you try connecting a double beam scope to the input of the TX module, and the output of the RX module - and see how little they resemble each other, and how distance between the modules alters the RX signal considerably.

Apparently you can do Manchester with a UART - but I've never really understood how?