433mhz modules whats the furthest anyone has ever gotten?

demonicpicaxeguy

Senior Member
i've just had some fun with some 433mhz modules 1 transmitter and 1 reciever, using a 6 inch piece of wire as an ariel on both, all built on two breadboards,i managed to get 277m(900ft) acording to google earth.

out of shear experience how far using what type/length ariels has anyone gotten with these things and if used for serial communications what baud rate did you use and how reliable was it?
 

testerrrs

New Member
In the middle of a town, and as such, extremely built up, I have managed to get 100m out of them.

In a less built up area, the maximum I have got so far is 450m. Theoretically, they should be good for around 1km.

Why 6" aerials out of interest? You'll probably find you get better range if you use an aerial that is either a full or half (or even quarter) wavelength. I've been using bits of wire 69.1cm long, which is a full wavelength for 433.92Mhz:

100*(300,000,000 / (433.92 * 1,000,000)) = 69.14cm.

You might like to give that a try and see if your range improves, I'm guessing it will.

Jon
 

demonicpicaxeguy

Senior Member
6inch lengths of wire was just a handy quick test dodgy ariel,it's winding type wire pushed into a breadboard

the problem i had tonight was that i could only walk 277m before i ran out of room on the field so i dare say i'll get even more range when i try it again on a bigger spot tomorrow night

i'll try the proper ariel lengths tomorrow night and see how we go
 

Dippy

Moderator
The furthest I ever got (honestly) was 420m not-quite-line-of-site using (ready made) helical at both ends (according to GPS). Experiments showed a large effect with ground planes but I was limited to little ones by the enclosures. PCB layout and cabling were very important and often forgotten by people.

I was using HT12E one end and HT12D at t'other, so I'm not sure of the baud rate - but as the HT's use 4 x repeat for checking , the signal must have been pretty good. People often forget the importance of modulation and there is a good article on Radiometrix website about it an the benefits of Manchester.

The modules I used were T7G and R7G from RF Solutions and the pcb is marked 'MK COnsultants.co.uk' .

When I used serial directly from PIC my best was just over 250m using helicals, but I didn't fiddle with ground plane. I was trying to make a radio 'network', but gave up.
 

Dippy

Moderator
"Theoretically, they should be good for around 1km."

Well, with respect, it depends on the module, wiring, connections, ground-plane and the antenna.

If you buy a Beijing 5p job you'll never get a km.
Buy something decent and Yagis then , well, who knows.

Antenna length importance, absoluely. But have you included the cable length from module o/p pin to antenna? Or did you use coax and have a ground plane?

So many variables......
 

demonicpicaxeguy

Senior Member
the modules in question are just the $9.95 jobs from jaycar

personally i'm amazed i got to 277m that way it is setup, it had to be dumb luck,

i only wanted 15m out of it for getting data from a small outside weather station to my desktop

i'll probably get to 300m odd before it's out of range with the current setup i'm just interested now to see far it will go
 
Last edited:

moxhamj

New Member
Well that is pretty impressive. I got 500m out of the 1W chinese ones, but they broke into oscillations after sending a packet and I'm still trying to work out a way of keeping the RF from getting back into the picaxe. Those Jaycar ones are 10mW? I tried 10mW modules and they wouldn't even go 10 metres but they weren't Jaycar ones.
I'm still trying to find a decent module that can go a significant distance without needing yagis and without costing a packet. Any suggestions?
 

Dippy

Moderator
With your 1W devices have you tried coax from pcb to antenna, then surrounding the antenna with ground-plane (eg a metal beer-mat) , connecting coax screen to it and 'hiding' your pcb behind the ground-plane?

BTW: The RF Solutions ones I used were:
http://www.rfsolutions.co.uk/acatalog/DS307-7.pdf

I particularly liked them as a) they are made 'locally', and, b) the Rx has RSSI which hugely useful, and also, of course, the legal aspect.

PS. When I moaned to RF Sol about only getting 400m, they reckoned 800m was easily acheivable with decent wiring and gound-plane. But I guess you're looking for more?
How big is your wallet as there are some much more 'manly' transmitters out there.
You ain't going to get something for nothing.

My concern with running things 'at the edge' (and particularly with omni-directional antennae) is that they are more susceptible to interference etc. maybe from electrical sources or other RF sources - hence the preference for yagis.
 
Last edited:

manuka

Senior Member
Short Answer: ~300m Line of Sight ( LOS) using ¼ wavelength (~170mm) whips on each. With ~6dB gain "cotanga" Yagis both ends, the range may be doubled x doubled = 4 X = ~1km.

Long Answer: It depends on many factors. I've been wrangling electrons with PICAXEs & 433 MHz ISM for ~5 years ( with ~40 years prior VHF/UHF experiences) & have trialled most combos of 433MHz rx/tx/simple antenna/data rates/encoding etc. Reliable ranges with these units tends to relate to both the nature of your data (Can it be resent? Are drop outs tolerable? Error detection in use? ) & cussed local interference, as much as specs of the 433MHz electronics. As most tx are just ~10mW (the regs. limit to 25mW),often far better range boosts can be made by using more sensitive rx modules. Some cheapies run out of steam at -90dBm while others work almost down to the thermal noise at -110dBm & give 2400bps data ranges of several km LOS.

Fast data rates and noisy spectrums can thwart even superior receivers however, & a 433MHz range shootout may yield results differing from those at the same site at another time- perhaps even due to varying ground moisture (as Dr_A found) or changing interference (less 433MHz activity late at night etc). Stan
 
Last edited:

moxhamj

New Member
No Dippy, I haven't tried taking the RF away with coax, and this could well be the next experiment to do. Indeed, this may well be the simplest boost for a longer range - take the RF via some coax and get it up several metres in the air. A question as an RF newbie - does the length of the coax make much difference - is it better to go for multiples of the wavelength, or just pick a random length of several metres?
 

manuka

Senior Member
Sure - get that antenna up in the clear! But even at 433MHz coax & connectors introduce their own losses, & although these are not as bad as those typical at 2.4GHz of "1dB per metre", long cooking grade coax runs can be equivalent to a bucket brigade using leaky containers.

Impedance matching & optimum coax length is antenna related & can be rather a black art.
The common notion of using 1/2 wavelength multiples of coax arises due to the antennas feed point impedance being "mirrored" at the input of the coax when using such a length. However overall the best coax length is that which is the shortest (for the job) between the antenna & the transmitter. Achieving the best match my be futile if you introduce another 6dB of losses with longer coax runs!
 
Last edited:

Dippy

Moderator
How long is the cable from your TV aerial to your TV?

Seriously though, it will matter a bit because of the cable quality/spec/any mismatching. But many manufactured antennas have several metres of cable included so don't worry if only a couple of metres, but shorter the better.

The important things are:
1. Make sure the unscreened parts (pcb tracks / joins / joint to antenna) are as short as possible as this will add to your 'radiated' lenght. There are some nice crimp cable/pcb connectors available for just this purpose but it does mean buying a crimper so the wallet will have to be opened to get the best.

2. Make sure the screen is connected to ground on pcb AND , if possible, to a decent ground-plane around base of whip.
(There are articles about pcb track/ground design to get 50ohm impedance on pcb but I can't remember where they are).

3. If practical/possible a decent ground-plane around base of whip or helical. 6" square will improve things a helluva lot. This ground-plane connected to earth screen from coax.

4. Not to have your antenna 'hot' end near anything metal/conductive.

5. Unles the manufacturers say otherwise, don't have metal near the RF module. Most in screened cans so not really an issue.

6. Care with the RF bit on your pcb layout.

Some of the more famous manufacturer's will have info about dimensions and spacing on their websites, maybe check out Radiometrix or RF Solutions. And I daresay many others if you have the patience to Google.

PS. It is obv possible to get zillions of pcb RF connections, but for cheapness, and if a hardwired solution is OK then these are good:-
http://uk.farnell.com/1215645/connectors/product.us0?sku=ITT-SEALECTRO-055-939-9049FCD&_requestid=201269

sadly, you'll need a stripper and crimper, but lifes always tough to get it good ;)
 
Last edited:

demonicpicaxeguy

Senior Member
(Can it be resent? Are drop outs tolerable? Error detection in use? )

yep the data can be resent
drop outs are expected
error detection plently

to give you an idea of the requirements

range 15m(min) to 30m(max)
serial transfer speed 1200bps

it basically going to be used for a series of monitoring stations
mini weather station,water tank monitor, water meter logger, electricity meter logger and various temp monitors each point is going to have a tx and an rx module

i'm going to use a simple "master" and "slaves" packet based protocol
the slaves will be addressable and won't "talk" unless "spoken" to
that will avoid data collision

i'll make use of some kind of crc error checking method as well and just resend packets that contains errors
 

moxhamj

New Member
If your range is only 30m this will work fine as every module can talk to every other module in the group without having to go via repeaters. So you can simply have a unique number for each module.

How are the modules powered?
 

demonicpicaxeguy

Senior Member
i'm thinking just rechargable batteries 4 1.2v AA's
they should last a while before i have to recharge them

solar of some kind would be ideal
 

demonicpicaxeguy

Senior Member
Just 30m! You could almost use IR if LOS. Maybe hack a solar garden lamp to power things?
i was thinking about using a solar light as a power source next to the water meter and the weather station

looking at my local dick smith they have a single solar light for $4.99 and a 5 pack for $15.00 i couldn't buy a single solar panel for that!!!

might be worth looking at for all of the "slaves" the five pack might be enough for all the power requirements
 

moxhamj

New Member
Solar is well worth looking at but I'm not sure the solar lights have enough grunt. Measure the current draw of a Rx module on standby - if it is (say)1mA and you are stepping up 1.2V to 5V that will be about 5mA draw from the battery, and over 24 hours that is 120mAH. The cells give out 40mA in full sun over 12 hours plus in summer which will be ok, but maybe only 10mA for 6 hours in winter on a cloudy day. I'm still thinking of a design to use some RF Rx modules that go down to 200uA which would be 1mA from the battery which would work with just one little solar light, especially if the RF module can wake up a picaxe that is in low power mode.

Silicon Chip did a great article a few months back and they used a TL499 chip to step up the volts. Anyone know of a supplier in Australia?
 

demonicpicaxeguy

Senior Member
the datasheet for the tx module suggest 10ma(max) same for the reciever

just thinking if a pic can put out 10ma on a pin why not just run the tx module off an output pin or via a small transistor?

the reciever would have to stay turned on but it'd be good to turn that off while transmitting

the weather station is going to use a pic16f88 and i could run that at a nice low frequency for a "monitoring standby mode" and have it ducking into 4mhz or 8mhz while it transmits data and then back to standby mode again

acording the the pic16f88 datasheet it consumes 73ua @ 1mhz so i dare say it's not worth worrying about

here's what i'm planning -

the weather station - will have the usaul humidity,temp,wind speed,wind direction,light levels + uvlevels,precipitation over last 1 hour

water meter - using a reed switch to increment a 32bit counter

electricity meter X 2- funny one i'll post a photo later but it's one of those older type spinning disk meters with the black non-reflective mark

after that all works reliably i'm sure i can find more places to stick things like thermisters ,ldrs and maybe some carbondioxide sensors, carbon monoxide sensors,

ideally in the end i'll have my server hooked up to the "master" via usb and all the data graphable and logged along with a more concise version on a html page

i wonder if it's possible to get that kind of thing into the sig below
 

hippy

Technical Support
Staff member
If the modules require less than 20mA/25mA then they can probably be run directly from a PICAXE output pin without a transistor.
 

moxhamj

New Member
The Tx modules tend to only consume power when transmitting, so if you have them powered up and with a low signal line they may be using a lot less. So if they drew 15mA and the picaxe could only supply 10mA, you could switch them on with a picaxe pin, charge up a cap eg 1000uF over 1/2 a second or so, then transmit the packet, and then power down the Tx unit. But I think the picaxe pin may well be able to supply enough current.

If sensor units only transmit, you can get the power a lot lower. Run the picaxe in low power mode and wake up every now and then at random intervals and sense the data and send out the packet. If there is a checksum and the packets are infrequent then clashes with other units will be minimal.
 

demonicpicaxeguy

Senior Member
ideally i want to be able to send commands to each "slave" and retrieve data whenever i want it also completely eliminates the chance of a "packet collision" unless somone has a 433mhz garage door opener around here which i doubt because most of them are now 2.4ghz

i'm going to have the screen on my server displaying graphs based upon what the "slaves" record

so i might have the master requesting the windspeed and direction say 24 times per day and at the same time only get the meter readings 2 times a day if i want to change it i won't have to reprogram the "slaves" just change a number on the program running on the server
 

moxhamj

New Member
I'm working on the same problem and I still don't have quite the best solution so it would be great to brainstorm ideas. Battery capacity is an issue if you charge with solar and either you have sophisticated overcharge protection (temp sensors etc) or go for a current that is trickle charge. There are 2400mAH cells but I've heard quite a few are fakes and are more like 1000mAH, so trickle charge might be 40mA or so. If you have 3 cells in series you can run a picaxe and so the solar panels from 3 solar lights would work for that. But still need to regulate to 5V if sensing analog values, so either go 6-7 cells and a 5V reg or 3 cells and maybe a controllable stepup converter that produces 5V when needed and switch over from unregulated to regulated power.

Then there is the single cell solution - instead of 3 cells in series, put them in parallel. I think this is allowed. Then the charge current can be up to 120mA. Or a C or D cell rather than AA cells. Then use a single TL499 or similar to produce a regulated 5V.

I've been running trials over the last few months with a cluster of 4 nodes each with receivers and transmitters as I wanted to see how they cope with cloudy days etc. So far I am finding that currents of at least 150mA are needed to charge. Solar lights are 40mA so I think you need 4.

I've got some more ideas - but have to get back to work now so will post later...
 

Dippy

Moderator
Assuming I have read correctly, if you have a transceiver 'slave' the usual practice is to duty-cycle the receiver, then when RF is present you switch it on full (with a transistor) and keep it on until packet is finished.
Needless to say a proper module with RF sense or RSSI amakes this a piece of cake.
And, of course, with RSSI you can use ADC as well to make a 'squelch' control.
Life is so much easier when you buy good gear.
 

manuka

Senior Member
433MHz tx indeed only draw power when transmitting, & overall tx/rx current drains (as shown => www.picaxe.orcon.net.nz/all3.jpg ) are <10mA & thus well within PICAXE pin sourcing. The new BOD (brown out detect) feature can cut standby draw to a few µA, but some scheme to "awaken the dead" will be needed.

RSSI (Received Signal Strength Indication - a measurement of the power present in a received radio signal) indeed may have some mileage. It's often labeled analog. output, & even some cheapie Rx modules may have it (although you'll need a pin scrutiny & maybe a DMM). Check Jaycars KEYMARK RXB1 closeup below (thanks Stan.),especially specs. of the Himark RX4300 "engine",which surely must have RSSI on one of it's 24 pins-perhaps pin 14 "PD" (power down). See data via => http://www.himark.com.tw/rx3400_a.htm

2nd generation 433MHz units are much more sensitive,(meaning ranges to ~1km) but at a price increase. I'm still awaiting the Linx release => http://www.picaxe.orcon.net.nz/linx433.jpg ,which looks promising...
 

Attachments

Last edited:

moxhamj

New Member
Thanks to Dippy for the suggestion about coax! Have just managed 400 metres through trees with 100% reliabilty of a packet with checksum. Total cost $30. Secret is to use higher power powered modules and get the antenna 3 metres away from the board with some coax.

Writeup and photos at http://drvernacula.topcities.com/400_metre.htm

These are the modules I sent over to manuka for testing and the comment from him was that the Rx lacked sensitivity. I've found some new Rx modules and ordered some today with sensitivity of -101dbm (compared with 93dbm in this experiment). So the range might be able to go even further.
 

demonicpicaxeguy

Senior Member
i've some coax cable around here somwhere so i'll try that with the ariel,
now i've setup an interesting test

on the tx module breadboard i've got a pic16f628a sending the number 34 at 300bps and toggling an led on and off

on the rx module breadboard i got a pic16f88 recieving the byte 34
if the byte is 34 it flashes a green led
and if it's not 34 then it flashes a red one

now when running the green led stays on 99% of the time to get the red one to light up i have to take it nice and close the the tx module

so far i can walk all over the house and it's only the microwave that makes the red led light up or if i turn off the tx module

i'm going modify it again so that it sends known "packet" of data and take a walk up the road to test it again at 277m
 
Last edited:

Dippy

Moderator
I'm glad its improved Dr_ac. But have you tried a ground plane around base of your Tx helical? (i.e. make a 6" metal disk with a hole in it, poke the helical up through it, insulated with a few mm between ant and gp, then attach the coax ground to it. Old biscuit tin will do to try. Not perfect but easy to make.)

I did some measurements with modules which had RSSI. At 50m the presence and size of ground plane showed a significant change in signal level on my 'scope. Experts will say 'No ship Sherlock' , but it was proof positive of benefits of the ground-plane.

Also, whilst there seems some arguments about helical antenna...
"Helical antennas are also resonant and generally chosen for their more compact dimensions. They are more difficult to optimise than monopole antennas and are critical with regard to surrounding objects that can easily ‘de-tune’ them. They operate most efficiently when there is a substantial ground plane for them to radiate against."
Also, I thought helical antenna should be tightly wound to reduce geometry variations?

Also, that MadeinCHN 4000m module looks a bit home-made, unscreened and susceptible to detuning. "Quality" and "Made in China" in the same sentence is sometimes an oxymoron. Bet its not legal.


Also, pcbs:
"A 50&#937; transmission line can be constructed on FR4 board material by using a 3mm wide PCB track over a ground plane. This should be kept as short as possible."
(They are, of course, referring to d/s board)
 

moxhamj

New Member
Thanks Dippy - some great ideas for more experiments. I've been doing all my tests with simple 1/4 wave antennas so all the distances can be compared, but once I find one that is the best, I'll be moving up to ground planes and yagis. I have 4 yagis (and a couple of quagis) ready to go.
 

moxhamj

New Member
I doubt I'll get close to manuka - he sent a signal across Wellington bay. Perhaps to go back to the spirit of the original thread, we should stick to 1/4 waves?
 

demonicpicaxeguy

Senior Member
i just did a quick test over the same 277m as the last night,
i simply had 1 led sending the same packet of data over and over again and the rx module hooked up to a pic16f88 which was receiving them it would light up a green led if it recieved it un corrupt and a red one when it received it corrupt
packects were sent every 500ms at 300bps

i setup a 17cm piece of wire on each one

from 2m all the way to 170ish there was no loss at all i even walk behind steel fencing and an entire house didn't affect it at all

from 170ish to 200 i got the odd red flash so we're starting to get corrupt packets

for 200ish to 277 it was progressivly getting worse it started to get sensitive to my son running around me in circles and when he got into LOS it went red until he got out of LOS

so that was an interesting test

next up i'm going to pick up another pair of these things tomorrow and make a pair of transcievers

i was hoping to try out a dipole antenna as shown in the writeup but i can find my roll of coax so i might have to buy some
 

demonicpicaxeguy

Senior Member
i've just managed to pickup a pair of solar lights that use a high intensity led
looking at the battery it uses a 1.2v 600mah ni-cd battery

the problem is that the pic16f88 needs at least 1.5v for ram retention and at least 2.0v to actaully run and i dare say i won't be running it very for long periods of time

i'm thinking of modifying this for a power supply
http://www.ececs.uc.edu/~dpl/vishwa_thesis/node66.html

has anyone else got any ideas?
 

demonicpicaxeguy

Senior Member
i'll give the first one a go,

i've gotten the communication working very very nicely now and i also got around having the rx module on all the time by turning it off and turning it once every minute for 500ms if there is nothing being transmitted then it goes back into sleep mode again by setting the osccilator back to the 32khz setting

the master just has to transmit the famous "UUUUUUUUUUUUUUU" for a minute or less to get the attention
 

moxhamj

New Member
What an interesting little piezo circuit, Mycroft2152. I think I might have to breadboard that one tonight!
 

Dippy

Moderator
Elementary Mycroft.
Seriously though, neat.
I must admit I though someone had posted an even simpler White LED / 1 cell driver last year??

What this has to do with Dracs radio module though I really don't know....

... btw last night was momentus. The last of my Chinese-made CFLs failed in a puff of smoke. PHutt...pooh. Couldn't see the Mother in law. Eight (8) CFls in 16 months. My German-made Osrams all (seven) still going after 5 years.
Chinese-made CFLs saving the planet? You get what you pay for and some dumb Guvs make things mandatory without checking the facts. Pah. You get what you pay for. Wait for LEDs.
And , strictly speaking, CFLs should be carefully disposed of with the tubes.
;End Rant.
 
Top